Economía

Highlights The global 6-month credit impulse is likely to turn up in the fourth quarter. This warrants profit-taking in some pro-defensive equity sector, regional, and country allocation... ...for example, in the 35 percent outperformance of European healthcare versus banks in just seven months. But do not become aggressively pro-cyclical until the 10-year yield on the Italian BTP (now at 3.2) moves closer to 3... ...and the sum of the 10-year yields on the U.S. T-bond, German bund and JGB (now at 3.4) also moves closer to 3. Chart Of The WeekThe Cycle Is About To Turn Feature One of the most common questions we get is, when will the cycle turn? And our response is always, which cycle? The cycle that most people focus on is the so-called business cycle, which describes multi-year economic expansions punctuated by recessions. However, the business cycle - to the extent that it is a cycle - is very irregular. Its upswings and downswings vary greatly in length (Chart I-2). This irregularity is one reason why economists are useless at calling the turns. Nevertheless, investors still obsess with calling the business cycle because they think this is the only cycle that drives the financial markets. Chart I-2The Business Cycle Is Very Irregular We disagree. Nature bestows us with a multitude of cycles with different periodicities: the daily tides, the monthly phases of the moon, the annual seasons, and the multi-year climate cycles. So it would be unnatural, and somewhat arrogant, to assume the economy and financial markets possess only one cycle. In fact, just as in nature, the economy and financial markets experience a multitude of cycles with different periodicities. There Is Not One Cycle In The Economy, There Are Many If you plotted yearly changes in temperature, you would get a flat line and you would think there were no seasons! The point being that you cannot see a yearly cycle if you look at yearly changes. To see the cyclicality of the seasons, you must plot 6-month changes in temperature. Likewise, you cannot see the shorter-term cycles in the economy and financial markets using analysis, such as yearly changes, designed to see longer-term cycles. Once you grasp this basic maths, the mini-cycles in the economy and financial markets will stare you in the face (Chart I-3), and a whole new world of investment opportunities will open up. Chart I-3The Mini-Cycle Is Very Regular As we advised on January 4: "Global growth experiences remarkably consistent - and therefore predictable - 'mini-cycles', with half-cycle lengths averaging eight months. As the current mini-upswing started in May 2017 we can infer that it is likely to end at some point in early 2018. So one surprise could be that global growth will lose steam in the first half of 2018 rather than in the second half, contrary to what the consensus is expecting... Pare back exposure to cyclicals and redeploy to defensives" The advice proved to be very prescient. The global economy did enter a mini-downswing sourced in the emerging markets (Charts I-4 - I-6). Chart I-4The U.S. Mini-Downswing Was Muted... Chart I-5...The Euro Area Mini-Downswing Was Also Muted... Chart I-6...But The China Mini-Downswing Was Severe Nevertheless, the global nature of financial markets meant that the German 10-year bund yield declined by 40 bps, while European healthcare equities outperformed banks by a mouth-watering 35 percent, and materials by 15 percent (Chart I-7 and Chart I-8). Some of these performances are as large as can be gained in a full business cycle begging the question: Why obsess with the impossible-to-predict business cycle when there are equally rich pickings in the easier-to-predict mini-cycle? Chart I-7Banks Vs. Healthcare Tracks The Mini-Cycle Chart I-8Materials Vs. Healthcare Tracks The Mini-Cycle Furthermore, if you get the equity sector calls right, you will get the equity regional and country calls right too. As cyclicals have underperformed, the less cyclically-exposed S&P500 has been the star performer of the major regional indexes. And cyclical-heavy stock markets like Italy's MIB have strongly underperformed defensive-heavy stock markets like Denmark's OMX (Chart I-9). Chart I-9Italy Vs. Denmark = Banks Vs. Healthcare It follows that the evolution of the global economic mini-cycle is pivotal in every investment decision (Box 1). BOX 1 The Theory Of Economic And Market Mini-Cycles The academic foundation of the global economic mini-cycles is a model called the Cobweb Theorem.1 When bond yields rise, interest rate sensitive sectors in the economy feel a headwind, but with a lag. Similarly, when bond yields decline, interest rate sensitive sectors feel a tailwind, but again with a lag. The lag occurs because credit demand leads credit supply by several months. As credit demand leads credit supply, the turning point in the price of credit (the bond yield) always leads the quantity of credit supplied (the credit impulse). The result is a perpetual mini-cycle oscillation in both economic variables. And because the quantity of credit supplied is a marginal driver of economic activity, this also creates mini-cycles in economic activity. These mini-cycles are remarkably regular with half-cycle lengths averaging around eight months and the regularity creates predictability. Moreover, as most investors are unaware of this predictability, the next turning point is not discounted in financial market prices - providing a compelling investment opportunity for those who do recognise the existence and predictability of these cycles. The Mini-Cycle Will Soon Turn Up The global 6-month credit impulse entered its current mini-downswing in January. Given that mini-downswings tend to last around eight months, we should expect the global economy to exit its mini-downswing in September, the escape valve being the recent decline in bond yields (Chart Of The Week). The caveat is that bond yields were slow to react to the mini-downswing and the decline in 10-year yields, averaging around 40 bps from the peak, has been pretty shallow. It follows that the next mini-upswing could be delayed to October/November, and be somewhat muted. Nevertheless, the surprise could be that global growth will stabilise in the fourth quarter of 2018, contrary to what the consensus is expecting. And this would suggest taking some of the most mouth-watering profits in pro-defensive equity sector, regional, and country allocation - for example, in the 35 percent outperformance of European healthcare versus banks (Chart I-10). Chart I-10Banks Have Severely Underperformed Healthcare Would we go a step further and become pro-cyclical? Not yet. One reason is that there is a limit to how far bond yields can rise before destabilising the very rich valuations of all risk-assets. This is captured in our 'rule of 4' which says that when the sum of the 10-year yields on the U.S. T-bond, German bund, and Japanese government bond (JGB) exceeds 4 - which broadly equates to the global 10-year yield exceeding 2 percent - it is time to go underweight equities. With the sum now equal to 3.4, yields can rise by only 25-30 bps before hurting risk-assets. Another reason for circumspection is that the investment landscape is still scattered with a large number of landmines, one of which has its own rule of 4. The Other 'Rule Of 4': The Italian 10-Year Bond Yield When Italian bond prices decline, it erodes the value of Italian banks' €350 billion portfolio of BTPs and weakens the banks' balance sheets. Investors start to get nervous about a bank's solvency when equity capital no longer covers net non-performing loans (NPLs). On this basis, the largest Italian banks now have €160 billion of equity capital against €130 billion of net NPLs, implying excess capital of €30 billion (Chart I-11). It follows that the markets would start to worry about Italian banks' mark-to-market solvency if their bond valuations sustained a drop of around a tenth from the recent peak. We estimate this equates to the 10-year BTP yield breaching and remaining above 4 percent (Chart I-12).2 Chart I-11Italian Banks' Equity Capital Exceeds Net NPLs By 30 Bn Euro Chart I-12Italian Banks' Solvency Would Be In Question If The 10-Year Yield Breached 4% Today the 10-year BTP yield stands just shy of 3.2 percent, but it is about to enter a testing period. The Italian government must agree its 2019 budget by September and present a draft to the European Commission by mid-October. The budget must tread a fine line. Cutting the structural deficit to appease the Commission would diminish the credibility of the populist government. It would also be terrible economics, making it harder for Italy to escape its decade-long stagnation.3 On the other hand, locking horns with Brussels and aggressively increasing the structural deficit might panic the bond market. The optimal outcome would be to leave the structural deficit broadly where it is now. To sum up, the global 6-month credit impulse is likely to turn up in the fourth quarter, warranting some profit-taking in pro-defensive positions. But we do not advise aggressive pro-cyclical sector, regional, and country allocation until the 10-year yield on the Italian BTP (now at 3.2) - and the sum of the 10-year yields on the U.S. T-bond, German bund and JGB (now at 3.4) - both move closer to 3. Dhaval Joshi, Senior Vice President Chief European Investment Strategist dhaval@bcaresearch.com 1 Please see the European Investment Strategy Special Report 'The Cobweb Theory And Market Cycles' published on January 11 2018 and available at eis.bcaresearch.com. 2 Assuming that the average maturity of Italian banks' BTPs is around 5 years. 3 Please see the European Investment Strategy Special Report 'Monetarists Vs Keynesians: The 21st Century Battle' July 12 2018 available at eis.bcaresearch.com. Fractal trading Model* In support of the preceding fundamental analysis, the outperformance of healthcare versus banks is technically extended. Its 130-day fractal dimension is at the lower bound which has reliably signalled previous trend exhaustions. On this basis we would position for a 10% reversal with a symmetrical stop-loss. In other trades, long PLN/USD reached the end of its 65-day holding period comfortably in profit, and is now closed. This leaves six open positions. For any investment, excessive trend following and groupthink can reach a natural point of instability, at which point the established trend is highly likely to break down with or without an external catalyst. An early warning sign is the investment's fractal dimension approaching its natural lower bound. Encouragingly, this trigger has consistently identified countertrend moves of various magnitudes across all asset classes. Chart I-13 * For more details please see the European Investment Strategy Special Report "Fractals, Liquidity & A Trading Model," dated December 11, 2014, available at eis.bcaresearch.com Fractal Trading Model The post-June 9, 2016 fractal trading model rules are: When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. Use the position size multiple to control risk. The position size will be smaller for more risky positions. Recommendations Equities Bond & Interest Rates Currency & Other Positions Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields Interest Rate Chart II-5Indicators To Watch - Interest Rate Expectations Chart II-6Indicators To Watch - Interest Rate Expectations Chart II-7Indicators To Watch - Interest Rate Expectations Chart II-8Indicators To Watch - Interest Rate Expectations

Bank Credit Analyst

Highlights Two key issues will remain important drivers of global financial markets in the coming quarters: the direction of the dollar and Chinese policy stimulus. Policy and growth divergences will remain tailwinds for the dollar and there is little the Trump Administration can do to reverse the upward trend. Dollar strength is exposing poor macro fundamentals in many emerging market economies. The problems facing EM economies run deep, and will not disappear anytime soon. Expect more EM fireworks. EM market turmoil could pause the Fed's tightening campaign, but this would require evidence that the U.S. economy and/or financial markets are being negatively affected. Chinese stimulus is a risk to our base-case outlook. A growth impulse might keep the RMB from weakening further, boost commodity prices and support EM exports. However, we believe that Chinese stimulus will not be a 'game changer', and might even cause more problems if the authorities push the RMB lower. The U.S. economy and financial system are less exposed to emerging markets than in the Eurozone. An excellent profit backdrop also provides U.S. risk assets with a strong tailwind. Nonetheless, the U.S. is not immune to EM woes. Poor valuation implies a meaningful correction in U.S. risk assets on any flight-to-quality event. Stay cautious on asset allocation. Fed Chair Powell is willing to wait for the "whites of the eyes" of inflation before becoming alarmed, almost ensuring that the FOMC will fall behind the inflation curve. Evidence of labor market overheating is accumulating. Bond yields will rise as the FOMC tries to catch up and long-term inflation expectations bounce. We believe that investors are underestimating the upside in U.S. inflation risks over the medium term. We recommend below-benchmark duration, although government bonds would temporarily rally if EM turbulence sparks a flight-to-quality. We still expect the supply/demand balance in the world oil market to tighten later this year. Stay positioned for higher oil prices. Japanese corporate profits have been stellar, but that will soon change. EPS growth is likely to soften in the Eurozone too. Favor the U.S. market in unhedged terms. Feature There are numerous key issues on the investment landscape, but two stand out at the moment because they both have wide-ranging global implications: (1) Will the U.S. dollar continue to appreciate; and (2) Will Chinese policymakers place structural reform on the back burner and 'go for growth' in the near term? The latest U.S. economic and profit data provide a strong tailwind for American risk assets. Nonetheless, the mighty U.S. dollar is casting a dark shadow over the heavily-indebted emerging market economies, sparking comparisons with the late 1990s. Could Turkey be the start of a 'domino' effect, similar to Thailand's plunge into financial crisis in 1997 that eventually spread to Brazil and Russia, and finally contributed to the demise of Long-Term Capital Management in the fall of 1998? On the global growth front, the story has not changed much from our assessment last month. Growth is solid, but slowing, in part due to a deceleration in developed-economy capital spending. The global expansion has become less synchronized and relative growth dynamics are pointing to more upside for the greenback (Chart I-1). Chart I-1Cyclical Divergence Is Still Dollar Bullish As in the late 1990s, the Fed is likely to ignore turbulence in EM financial markets and will continue on its tightening path until it begins to affect the U.S. economy or asset prices. The path of least resistance for the dollar is up until something breaks. A major policy impulse from China could alter the feedback loop between the strengthening dollar and EM asset prices. A growth pickup would lift China's imports and commodity prices, both of which would support emerging market economies and asset prices. There is plenty of uncertainty regarding the size of the recently-announced Chinese stimulus measures, but our take is that they are likely to underwhelm because a major growth push would undermine the authorities' structural initiatives. The implication is that the global backdrop will remain unfriendly to emerging market assets at a time when they are more vulnerable than the consensus believes. The risk of a financial accident is escalating. The good news is that the U.S. earnings picture remains excellent, which precludes us from being underweight on risk assets. Nonetheless, investors should have no more than a benchmark allocation to equities and corporate bonds in the major advanced economies. We are upgrading government bonds to neutral at the expense of cash on a tactical basis, to reflect the rising possibility of a global flight-to-quality. The First Domino Turkey has had all the hallmarks of a crisis for a long while. Erdogan's slim hold on power has motivated several populist policy decisions that have stretched Turkey's macro fundamentals. The central bank has been forced to provide large injections of liquidity into the banking system, despite double-digit inflation readings. The country suffers from a classic "twin deficit" problem. Turkish private sector external debt stands at 40% of GDP, of which 13% of GDP is short-term, the highest among EM countries. Erdogan wants economic growth at all costs, but has done little in terms of the structural reforms necessary to lift the country's growth potential. The Lira has lost almost 26% of its value versus the dollar since August 1 and Turkish spreads have blown out. It appears that a lot of bad news has been discounted, but our EM strategists do not see this as a buying opportunity. One risk is that Erdogan imposes capital controls next. Our emerging market team's long held caution on EM is rooted in concern for failing fundamentals.1 They emphasize that Turkey was the catalyst, not the main cause, for the broader financial stress observed across EM assets in August. BCA has highlighted for some time that EM debt is a ticking time bomb. Chart I-2 shows that EM dollar-denominated debt is now as high as it was in the late 1990s as a share of both GDP and exports. Chart I-3 highlights the most vulnerable EM economies in terms of the foreign currency funding requirement, and the foreign debt-servicing obligation relative to total exports. Turkey stands out as the most vulnerable, along with Argentina, Brazil, Indonesia, Chile, and Colombia. Chart I-2Debt Makes EM Vulnerable Chart I-3EM Debt Exposure In all previous major EM selloffs, any decoupling between different EM regions proved to be unsustainable. And it certainly does not help that the Fed remains on its tightening path; EM equities usually fall when U.S. financial conditions tighten (Chart I-4). The combination of a strong dollar and weak RMB is a deadly combination for highly-indebted emerging market economies. Chart I-4EM Highly Sensitive To U.S. Financial Conditions... Investors should expect contagion to intensify. China To The Rescue? Some investors are hoping that China will 'save the day' by providing a major dose of policy stimulus, as it did in 2015, the last time that EM was close to a tipping point. We doubt China will be able to play the same stabilizing role. The Chinese authorities are committed to their long-term structural goals. They have been trying to reorient the economy toward consumption and away from investment and exports, as well as undertake other reforms to reduce financial risk, pollution, poverty and corruption. China kept policy on the tight side until recently, which resulted in a gradual growth slowdown. The Li Keqiang index (LKI) is a good coincident indicator for economic growth (Chart I-5). This index has ticked up in recent months, along with imports, but this likely reflects industrial activity designed to fill foreign orders before the new U.S. tariffs take effect. Our LKI model, based on money and credit, points to further economic weakness ahead. Chart I-5China: Watch Credit And Fiscal Impulse The escalation of the trade war with the U.S. is forcing the Chinese authorities to provide some short-term policy stimulus in order to pre-empt any resulting economic damage. A flurry of policy announcements over the past month has given investors the impression that Beijing has cranked up the policy dial, including cuts to short-term interest rates, a decrease in reserve requirements, liquidity provision to the banking system, and promises of various forms of fiscal stimulus. Chinese stimulus has historically been positive for commodity prices and EM assets. However, we are less sanguine this time. First, the authorities are not abandoning structural reforms, which means that the associated growth headwinds will not disappear. Second, our China experts believe that Chinese policy is only turning moderately reflationary; this is not the 'big bang' that followed the Great Recession in the late 2000s, or the same level of stimulus provided following the 2015-16 global manufacturing downturn. There will no doubt be some fiscal stimulus, but we do not expect a major expansion in bank credit to the private sector because of the government's crackdown on shadow banking, excessive leverage and growing non-performing loans. The change in the policy stance amounts to 'taking the foot off the brake' rather than pressing firmly on the accelerator.2 Third, and perhaps most importantly, the authorities may rely even more on the currency lever to do the heavy lifting if the economy continues to slow and/or the tariff war escalates further. This would be negative for commodity demand because a weaker RMB will make commodities dearer for Chinese producers. Metals prices are particularly at risk. China's competitors will also feel the sting of a cheaper RMB. It will be critical to watch the Chinese money and credit data in the coming months to gauge whether our view on the policy stimulus is correct. We will also be watching the combined credit and fiscal impulse which, at the moment, points to continued weakening in import growth in the near term (Chart I-5, bottom panel). Slower EM growth and/or more financial market turbulence is likely to take a larger toll on the euro area than the United States. Exports to emerging markets account for only 3.6% of GDP for the U.S., compared to 9.7% of GDP for the euro area. Euro area banks also have more exposure to emerging markets than U.S. banks (Chart I-6). Notably, Spanish banks - BBVA in particular - has sizable exposure to Turkey. Meanwhile, Italian assets have come under pressure as the rift between the European Commission and the new populist government widens and Italian banks become increasingly wary of financing their government. Chart I-6DM Bank Exposure To EM European growth will therefore likely continue to trail that of the U.S. Our base case does not see euro area growth falling below a trend pace in the coming quarters, but relative growth momentum and the ongoing policy divergence will favor the dollar over the euro. FOMC: No Urgency The key message from the latest FOMC Minutes and Chairman Powell's Jackson Hole speech is that policymakers are sticking with the "gradual" approach to tightening, despite the late-cycle acceleration in economic growth. The blowout second-quarter GDP report supports the view that fiscal stimulus is stoking the economy at a time when there is little slack. Evidence that the labor market is overheating is not simply anecdotal anymore. In past cycles, an acceleration in growth at a time when inflation is already at target and unemployment is below estimates of full employment would have sparked aggressive Fed action. But the Minutes and Powell's speech revealed no sense of urgency. Powell made the case that the Fed must proceed carefully in an environment where there is much uncertainty about the level of the neutral policy rate, the natural rate of unemployment and the slope of the Phillips curve. Moreover, long-term inflation expectations are still hovering below a level that is consistent with meeting the 2% target over the medium term. Some FOMC policymakers believe that this fact justifies taking chances with an inflation overshoot in the coming quarters. Another reason for the FOMC to proceed cautiously is the wage picture, which is confusing even to economic experts because the official measures paint a mixed picture (Chart I-7). The Employment Cost Index for private sector workers continues to march higher. However, growth in compensation per hour, average hourly earnings (AHE) and unit labor costs have all eased a little this year. The Atlanta Fed Wage Tracker, one of the cleanest measures of wages, reveals an even more significant pullback. The softening in wage growth has been fairly widespread across age cohorts, educational attainment and regions, according to the Atlanta Fed data (Chart I-8). Part-time workers appear to be the only segment that has bucked the trend. It is not clear why workers in the 16-24 age group, as well as those with bachelor's degrees (of any age), have seen the most pronounced softening in wage growth this year. Chart I-7Mixed U.S. Wage Data Chart I-8U.S. Wage Slowdown Broadly-Based Which measure is telling the correct story: the ECI or the Atlanta Wage Tracker? Both are a relatively clean measure of wages and it is difficult to tell based on the relative merits of each index alone. Nonetheless, there is little doubt that the labor market is now very tight by historical standards. Small business owners' compensation plans remained near record levels in July, while concerns about the "quality of labor" have never been higher (Chart I-9). Chart I-10 shows that the ratio of the level of job openings to unemployed workers has surpassed the pre-recession level in all but one sector according to the Jolts survey. Indeed, in most cases this ratio is well above the previous peak. Unemployment is now below the estimated level of full-employment in more than 80% of U.S. states. Chart I-9U.S. Labor Shortage Is Growing Chart I-10JOLTS Signals Very Tight Jobs Market No Evidence Of U.S. Overheating? Labor shortages first appeared for skilled workers, helping to explain why highly-skilled workers have enjoyed the fastest wage gains in recent years. But this year's Fed Beige Books have noted that many businesses are now having trouble finding low- and middle-skilled workers, as listed in Table I-1. These industries roughly line up with the ones that reveal above-average growth in average hourly earnings, and with the ones where labor market tightness is the most acute according to the Jolts survey (second and third columns in the table). The shortages appear to be broadly based, ranging from truck transportation to financial services, manufacturing and construction. This makes it all the more curious that Chairman Powell finds that there is no evidence of overheating in the labor market. The evidence seems pretty conclusive to us and it even features in the Fed's own Beige Book. Keep in mind that inflation is not always the 'cost push' type, beginning in the labor market and traveling to consumer prices. Sometimes inflation can begin in the market for goods and services, and then affect wage demands. U.S. consumer price inflation appears to be headed higher based on the New York Fed's Underlying Inflation Gauge (Chart I-11). Our CPI diffusion index shows that inflation is accelerating in a majority of categories. Other measures of underlying inflation, such as the Sticky Price Index, the Trimmed Mean, and the Median inflation rate are all in a solid uptrend. Dollar strength this year will eventually put downward pressure on core goods inflation, but that will take some time; non-energy goods inflation is more likely to rise in the near term as it catches up to the previous acceleration in imported goods prices (Chart I-11, bottom panel). Table I-1Labor 'Shortages' Identified In The Beige Book Chart I-11U.S. Underlying Inflation Is Rising U.S. Inflation To Surprise On Upside We believe that the market is underestimating the risk of a meaningful inflation overshoot over the medium term. Investors still do not believe that the Fed will be able to consistently meet the 2% target over the long-term, based on CPI swaps and TIPS breakeven rates. BCA's Chief Global Strategist, Peter Berezin, penned a two-part Special Report in August on the potential for upside inflation surprises over the coming years.3 First, increasing political pressure on the major central banks is worrying. Second, policymakers are coming around to the idea that there may be an exploitable trade-off between higher inflation and lower unemployment. This was a mistake last made in the inflationary 1970s. Finally, the pressure to keep monetary policy accommodative until the "whites of the eyes" of inflation are visible will remain strong. Bonds are in for some trouble if we are correct on the inflation outlook. We recommend that investors with a 6-12 month investor horizon remain short in duration and overweight TIPS versus conventional Treasurys. That said, we cannot rule out a flight-to-quality episode at some point, possibly reflecting trade tensions and/or EM turmoil, which would send Treasury yields temporarily lower. The Fed may be forced to place rate hikes on hold if financial conditions tighten too quickly. No Margin Peak Yet In The U.S.... The S&P 500 was unfazed by the turmoil in emerging markets and the re-widening in Italian bond spreads in August, likely because of continuing good news on the profit front. Corporate earnings remained in a sweet spot in the second quarter. Nominal GDP grew by a whopping 5.4% from a year ago, helping to boost the top line for the corporate sector. The lagged effect of previous dollar depreciation is still flattering earnings, although this only accounts for about two percentage points according to our model (Chart I-12). Meanwhile, equity buybacks have kicked into overdrive (Chart I-13). Chart I-12U.S. Dollar Impact On EPS Growth Chart I-13U.S. Equity Buyback In Overdrive Margins continued their impressive ascent in the second quarter to well above the pre-Lehman peak (Chart I-14). A lot of the increase is related to the tax cuts; EBITDA margins are still substantially below the 2007 peak according to the S&P data. It is disconcerting that all of the surge in S&P 500 margins is due to the Tech sector (Chart I-14, bottom panel). Excluding Tech, S&P after-tax margins have simply moved sideways since 2010. Looking ahead, the tailwind from previous dollar depreciation will shift to a headwind by mid-2019. Chart I-12 shows that the contribution from changes in the dollar to EPS growth will shift from a positive two percentage points to a drag of 1½ percentage points if the dollar is flat from today's level in broad trade-weighted terms. If the dollar rises by another 5% this year, then next year's drag on EPS growth will reach three percentage points. Moreover, the impact of the tax cuts on after-tax profits will fade next year. Wage pressures are building and this should eventually squeeze profit margins. That said, a margin peak does not appear to be imminent. Last month we introduced some macro indicators for profit margins (Chart I-15). Most appeared to be rolling over a month ago, but they have all since ticked up. Chart I-14Tech And Taxes Driving Profit Margins Chart I-15U.S. Margin Indicators Have Turned Up The bottom line is that we continue to expect a mean reversion in U.S. profit margins in the coming years, but this is not a risk for at least the rest of 2018. ...But Profit Outlook Darkening In Japan Second quarter earnings season was also a good one for Japanese companies. Twelve-month forward earnings estimates have been in a steep incline and margins have been rising (Chart I-16). Despite this, the Nikkei has only managed to move sideways this year in local currency terms. Concerns over trade and global growth have perhaps weighed on Japanese stock performance. Company profits have a high beta with respect to global growth. Things are looking shaky on the domestic front too. Domestic demand growth is decelerating, consistent with a weakening Economy Watcher's Survey. Some of the weakness may be related to poor weather, but the LEI suggests that this trend will continue in the coming quarters (Chart I-17, bottom panel). Chart I-16Japan: Trailing Earnings Are Solid... Chart I-17...But Profit Margins Will Narrow Chart I-17 presents some of the variables that have helped to explain historical trends in Japanese EPS. Industrial production growth, a good proxy for top line growth, is decelerating. Nominal GDP growth has fallen to just 1.1% year-over-year, at a time when total labor compensation has surged by more than 4%. The difference between these two, a proxy for profit margins, has therefore plunged. Previous shifts in the yen have not had a large impact on EPS growth over the past year and we do not expect that to change much in 2019. On a positive note, Japanese stocks are attractively valued now that the 12-month forward P/E ratio has fallen below 13 (Chart I-16, bottom panel). It is also constructive that the Bank of Japan is the only central bank that is not backing away from monetary stimulus. The recent widening of the trading band for the 10-year JGB yield was a technical change meant to give the central bank more flexibility, not a signal that policymakers are planning to change tack. Nonetheless, we believe that earnings growth and margins will disappoint market expectations over the next year. The story is much the same for the Eurozone. Both trailing and forward profit margins have been in a strong uptrend. Twelve-month forward EPS growth has been holding at a solid 9%. Nonetheless, the data that feed into our Eurozone profit model point to some softening ahead, including industrial production and the difference between nominal GDP and the aggregate wage bill (not shown). The Eurozone's credit impulse turned negative even before concerns about EM and Italian politics exploded onto the scene. Thus, home-grown profit generation is likely to moderate along with foreign-sourced earnings. For the moment, the BCA House View remains at benchmark on Japanese and Eurozone stocks in currency-hedged terms. In unhedged terms, we prefer the U.S. market to these other bourses because of our bullish dollar bias. Investment Conclusions: Two key issues will remain important drivers of global financial markets in the coming months and quarters: the direction of the dollar and Chinese policy stimulus. We believe that the U.S. dollar has additional upside potential due to growth and policy divergences. There is some speculation in the financial community that President Trump might resort to currency intervention. However, any intervention would be sterilized by the Fed. The only way to shift currencies on a sustained basis would be to organize a coordinated change in monetary or fiscal policies among the U.S. and its main trading partners. This is highly unlikely. Thus, the path of least resistance is up for the U.S. dollar. Dollar strength is exposing poor macro fundamentals in many emerging market economies. The problems facing EM economies run deep, and will not disappear anytime soon because high debt levels make these economies vulnerable to any weakness in global growth, commodity prices or global liquidity conditions. EM financial market turmoil could cause the Fed tightening campaign to go on hold, but this would require evidence that the former is negatively affecting the U.S. economy and/or financial markets. In other words, we need to see some pain before the Fed blinks. Chinese stimulus is a risk to our base-case EM outlook. Policy stimulus might keep the RMB from weakening further, boost commodity prices and support EM exports. This would not change the EM debt situation, but would at least give emerging economies a temporary reprieve. Careful analysis suggests that Chinese stimulus will not be a 'game changer', and might even cause problems if the authorities push the RMB lower. But it will be critical to monitor the next couple of money and credit reports. The U.S. economy and financial system are less exposed to further EM turmoil than in the Eurozone. But as the LTCM event demonstrated in 1998, the U.S. is not immune. Moreover, U.S. equity prices are more expensive than they were during previous EM selloffs that have occurred since the Great Recession. This could mean a larger equity re-rating on any flight-to-quality. This is not to say that we expect a bear market in DM risk assets to get underway in the near future. A U.S./global recession before 2020 is unlikely. Nonetheless, the risk of a meaningful correction is elevated enough that caution is warranted, especially at a time when all risk assets appear expensive. Chart I-18 updates our valuation measures for some major asset classes. All appear to be expensive, especially U.S. equities, raw materials and gold. EM sovereigns and equities are at the cheaper end of the spectrum, but are still not cheap in absolute terms even after the recent selloff. Chart I-18Major Asset Valuation Comparison Treasurys rallied briefly after Chairman Powell signaled that he is not willing to accelerate the pace of rate hikes in light of the U.S. economy's growth acceleration. He is willing to wait until he sees the "whites of the eyes" of inflation before becoming alarmed, almost ensuring that the FOMC will fall behind the inflation curve. Bond yields will rise as the FOMC tries to catch up and long-term inflation expectations bounce. Over the medium term, we believe that investors are underestimating the upside in U.S. inflation risks. We recommend below-benchmark duration, although bonds may temporarily rally if EM turbulence sparks a flight-to-quality. We still expect the supply/demand balance in the world oil market to tighten later this year. Stay positioned for higher oil prices. Finally, as we go to press, the U.S. is trying to force Canada to sign on to the U.S./Mexico 'agreement in principal' by August 31. A framework deal with Canada would likely leave many tough issues unresolved. There is also a chance that Canada misses the deadline and that the existing trilateral deal will not survive. It is technically possible that Canada's refusal to join the U.S.-Mexico bilateral deal will delay its ratification well into next year. In the meantime, Trump could raise the stakes for Canada by boosting tariffs on Canadian autos and/or by suspending NAFTA altogether. As a result, we decided to go ahead and publish our Special Report on U.S. equity sector implications if NAFTA is not ratified and tariffs rise to WTO levels. The report begins on page 20. Mark McClellan Senior Vice President The Bank Credit Analyst August 30, 2018 Next Report: September 27, 2018 1 Please see BCA Emerging Market Strategy Weekly Report "What's Really Driving The EM Selloff?"dated June 28, 2018, available on ems.bcaresearch.com 2 Please see BCA China Investment Strategy Weekly Report "China is Easing Up On The Brake, Not Pressing The Accelerator," dated July 26, 2018, available on cis.bcaresearch.com 3 Please see BCA Global Investment Strategy Special Reports: "1970s-Style Inflation: Could It Happen Again? Parts I and II," dated August 10 and 24, 2018, available on gis.bcaresearch.com II. What If NAFTA Is Not A Done Deal? U.S. Equity Implications This Special Report examines the impact of a NAFTA cancelation on 21 level-three GICs industries. While the latest news on the NAFTA renegotiation with Mexico is positive as we go to press, there is still a non-negligible risk that the existing trilateral deal will not survive. The U.S.-Mexico bilateral deal is an "agreement in principle" and will take time to ratify. Meanwhile, a framework deal with Canada would leave many thorny issues to be resolved. President Trump can still revert to his tough tactics on Canada ahead of the U.S. mid-term elections. If the President does not gain major concessions that can be presented as "victories" to voters, he is likely to take an aggressive stand in order to fire up his political base. The probability of Trump triggering Article 2205 and threatening to walk away from the suspended U.S.-Canada free trade agreement is still not trivial, despite the deal with Mexico. By itself, the cancelation of NAFTA would not be devastating for any particular U.S. industry because the size of the tariff increases would be fairly small as long as all parties stick with MFN tariff levels. That said, the impact would not be trivial, especially for those industries that have extensive supply lines that run between the three countries involved (especially Autos). We approached the issue from four different perspectives; international supply chains, a model-based approach, and an analysis of foreign revenue exposure and input cost exposure. The broad conclusion is that there are no winners from a NAFTA cancelation for the U.S. manufacturing GICs industries. Pharmaceuticals, Health Care Equipment & Supplies, Personal Products and Construction Materials are lower on the risk scale, but cannot be considered beneficiaries of a NAFTA collapse. The remaining industries are all moderately-to-highly exposed. Considering the four perspectives as a group, the most vulnerable industries are Automobiles, Automobile Components, Metals & Mining, Food Products, Beverages, and Textiles & Apparel. Our U.S. equity sector specialists recommend overweight positions in Defense and Financials; while neither stands to benefit from a NAFTA abrogation, they should at least be relative outperformers. They recommend underweight positions on Auto Components, Steel and Electrical Components & Equipment as relative (and probably absolute) underperformers should NAFTA disappear. While the latest news on the renegotiation of the North American Free Trade Agreement (NAFTA) is positive as we go to press, there is still a non-negligible risk that President Trump could revert to his tough tactics ahead of the U.S. mid-term elections.1 Even if Canada signs on to a framework deal, a lot of thorny details will have to be worked out. A presidential proclamation triggering Article 2205 of the NAFTA agreement (as opposed to tweeting that the U.S. will withdraw) would initiate a six-month "exit" period. Trump could use this deadline, and the threat of canceling the underlying U.S.-Canada FTA, to put pressure on Canada (if not Mexico) to concede to U.S. demands, just as he could revoke his exit announcement anytime within the six-month period. While some market volatility would ensue upon any exit announcement, even a total withdrawal at the end of the six months would have a limited macro-economic impact as long as the U.S. continued to respect its WTO commitments and lifted tariffs only to Most Favored Nation (MFN) levels. Nonetheless, a modest tariff hike is not assured given the Administration's "America First" policy, its looming threat of Section 232 tariffs on auto imports, its warnings against the WTO itself, and the steep tariffs it has already imposed on Canada, including a 20% tariff on softwood lumber and the 300% tariff on Bombardier CSeries jets. Moreover, even a small rise in tariffs to MFN levels would have a significant negative impact on industries that are heavily integrated across borders. Our first report on the evolving U.S. trade situation analyzed the implications of the U.S.-China trade war for the 24 level two U.S. GICs equity sectors. This Special Report examines the impact of a NAFTA cancelation on 21 level three GICs industries (finer detail is required since NAFTA covers mostly goods industries). We find that there are no "winners" among the U.S. equity sectors because the negative impact would outweigh any positive effects. The hardest hit U.S. industries would be Autos, Metals & Mining, Food Products, Beverages, and Textiles and Apparel, but many others are heavily exposed to a failure of the free trade agreement. Out Of Time President Trump is seeking a new NAFTA deal ahead of the U.S. midterms in November. While this timing may yet prove too ambitious, the U.S. has made progress in recent bilateral negotiations with Mexico, raising the potential that Trump will be able to tout a new NAFTA framework deal by November 6. Yet, investors should be prepared for additional volatility. There are technical issues with the bilateral U.S.-Mexico deal that could delay ratification in Congress until mid-2019. The new Mexican Congress must ratify the deal by December 1 if outgoing President Enrique Peña Nieto is to sign off. Otherwise, the incoming Mexican President Andrés Manuel López Obrador may still want to revise any deal he signs, prolonging the process. Meanwhile, it would be surprising if the Canadians signed onto a U.S.-Mexico deal they had no part in negotiating without insisting on any adjustments.2 The important point is that President Trump's economic and legal constraints on withdrawing from NAFTA have fallen even further with the Mexican deal. If Trump does not get major concessions that can be presented as "victories" to voters, he is likely to take an aggressive stand in order to fire up his political base, as a gray area of "continuing talks" will not inspire voters. This could mean imposing the threatened auto tariffs or threatening to cancel the existing trade agreements with Canada. Thus, the risk of Trump triggering Article 2205 is still not trivial. A bilateral Mexican trade deal is not the same as NAFTA. Announcing withdrawal automatically nullifies much of the 1993 NAFTA Implementation Act. Some provisions of NAFTA under this act may continue, but the bulk would cease to have effect, and the White House could refuse to enforce the rest. The potential saving grace for trade with Canada was that the Canada-U.S. Free Trade Agreement (CUSFTA), which took effect in 1989, was incorporated into NAFTA. The U.S. and Canada agreed to suspend CUSFTA's operation when NAFTA was created, but the suspension only lasts as long as NAFTA is in effect. However, Trump may walk away from both CUSFTA and NAFTA in the same proclamation. In that event, WTO rules for preferential trade would require the U.S. and Canada to raise tariffs on trade with each other to Most Favored Nation (MFN) levels. These tariff levels are shown in Charts II-1A and II-1B. The Charts also show the maximum tariff that could potentially be applied under WTO rules. The latter are much higher than the MFN levels, underscoring that the situation could get really ugly if a full trade war scenario somehow still emerged among these three trading partners. Chart II-1AU.S.: MFN Tariff Rates By GICS Industry (2017) Chart II-1BMexico & Canada: MFN Tariff Rates By GICS Industry (2017) Current tariffs are set at zero for virtually all of these GICs industries, which means that the MFN levels also indicate how much tariffs will rise at a minimum if NAFTA is cancelled. Tariffs would rise the most for Automobiles, Textiles & Apparel, and Food Products (especially agricultural products), and Beverages. U.S. tariffs under the WTO are not significantly higher than NAFTA's rates; the average MFN tariff in 2016 was 3½%, which compares to 4.1% for the average Canadian MFN tariff. Would MFN Tariffs Be Painful? An increase in tariff rates of 3-4 percentage points may seem like small potatoes. Nonetheless, even this could have an outsized impact on some industries because tariffs are levied on trade flows, not on production. A substantial amount of trade today is in intermediate goods due to well-integrated supply chains. Charts II-2A and II-2B present a measure of integration. Exports and imports are quite large relative to total production in some industries. The most integrated U.S. GICs sectors include Automobiles & Components, Materials, Capital Goods and Electrical & Optical Equipment. Higher tariffs would slam those intermediate goods that cross the border multiple times at different stages of production. For example, studies of particular automobile models have found that "parts and components may cross the NAFTA countries' borders as many as eight times before being installed in a final assembly in one of the three partner countries."3 Tariffs would apply each time these parts cross the border if NAFTA fails. Chart II-2AU.S./Canada Supply Chain Integration Chart II-2BU.S./Mexico Supply Chain Integration Appendix Tables II-1 to II-4 show bilateral trade by product between the U.S. and Canada, and the U.S. and Mexico. In 2017, the U.S. imported almost $300b in goods from Canada, and exported $282b to that country, resulting in a small U.S. bilateral trade deficit. The bilateral deficit with Mexico is larger, with $314b in U.S. imports and $243b in exports. The largest trade categories include motor vehicles, machinery, and petroleum products. Telecom equipment and food products also rank highly. As mentioned above, the impact of rising tariffs is outsized to the extent that a substantial portion of trade in North America is in intermediate goods. Box II-1 reviews the five main channels through which rising tariffs can affect U.S. industry. Box II-1 Trade Channels There are at least five channels through which rising tariffs can affect U.S. industry: (1) The Direct Effect: This can be positive or negative. The impact is positive for those industries that do not export much but are provided relief from stiff import competition via higher import tariffs. The impact is negative for those firms facing higher tariffs on their exports, as well as for those firms facing higher costs for imported inputs to their production process. These firms would be forced to absorb some of import tariffs via lower profit margins. Some industries will fall into both positive and negative camps. U.S. washing machines are a good example. Whirlpool's stock price jumped after President Trump announced an import tariff on washing machines, but it subsequently fell back when the Administration imposed an import tariff on steel and aluminum (that are used in the production of washing machines). NAFTA also eliminated many non-tariff barriers, especially in service industries. Cancelling the agreement could thus see a return of these barriers to trade; (2) Indirect Effect: The higher costs for imported goods are passed along the supply chain within an industry and to other industries that are not directly affected by rising tariffs. This will undermine profit margins in these indirectly-affected industries to the extent that they cannot fully pass along the higher input costs. There would also be a loss of economies-of-scale and comparative advantage to the extent that firms are no longer able to use an "optimal" supply network that crosses borders, further raising the cost of doing business; (3) Foreign Direct Investment: Some U.S. imports emanate from U.S. multinationals' subsidiaries outside the U.S., or by foreign OEM suppliers for U.S. firms. NAFTA eliminated many national barriers to FDI, expanded basic protections for companies' FDI in other member nations, and established a dispute-settlement procedure. The Canadian and Mexican authorities could make life more difficult for those U.S. firms that have undertaken significant FDI in retaliation for NAFTA's cancellation; (4) Macro Effect: The end of NAFTA, especially if it were to lead to a trade war that results in tariffs in excess of the MFN levels, would take a toll on North American trade and reduce GDP growth across the three countries. Besides the negative effect of uncertainty on business confidence and, thus, capital spending, rising prices for both consumer and capital goods will reduce the volume of spending in both cases. Moreover, corporate profits have a high beta with respect to economic activity. The macro effect would probably not be large to the extent that tariffs only rise to MFN levels; (5) Currency Effect: To the extent that a trade war pushes up the dollar relative to the Canadian dollar and Mexican peso, it would undermine export-oriented industries and benefit those that import. However, while we are bullish the dollar due to diverging monetary policy, the dollar may not benefit much from trade friction given that tariffs would rise for all three countries. Chart II-3 is a scatter chart of GICs industries that compares the average MFN tariff on U.S. imports to the average MFN tariff on Canadian and Mexican imports from the U.S. A U.S. industry may benefit if it garners significant import protection but does not face a higher tariff on its exports to the other two countries. Unfortunately, there are no industries that fall into the north-west portion of the chart. The opposite corner, signifying low import protection but high tariffs on exports, includes Beverages, Household Durables, Household Products, Personal Products and Machinery. Chart II-3Import And Export Tariffs Faced By U.S. GICS Industries Model-Based Approach The C.D. Howe Institute has employed a general equilibrium model to estimate the impact of a NAFTA failure at the industrial level.4 The model is able to capture the impact on trade conducted through foreign affiliates. The study captures the direct implications of higher tariffs, but also includes a negative shock to business investment that would stem from heightened uncertainty about the future of market access for cross-border trade. It also takes into consideration non-tariff barriers affecting services. Table II-1Impact Of NAFTA Cancellation By Industry As with most studies of this type, the Howe report finds that the level of GDP falls by a relatively small amount relative to the baseline in all three countries - i.e. there are no winners if NAFTA goes down. Moreover, the U.S. is not even able to reduce its external deficit. While the trade barriers trim U.S. imports from NAFTA parties by $60b, exports to Canada and Mexico fall by $62b. At the industry level, the model sums the impacts of the NAFTA shock on imports, exports and domestic market share to arrive at the estimated change in total shipments (Table II-1). It is possible that an industry will enjoy a boost to total shipments if a larger domestic market share outweighs the damage to exports. However, the vast majority of U.S. industries would suffer a decline in total shipments according to this study, because the estimated gain in domestic market share is simply not large enough. Beef, Pork & Poultry and Dairy would see a 1-2% drop in total shipments relative to the baseline forecast. Next on the list are textiles & apparel, food products and automotive products. Even some service industries suffer a small decline in business, due to indirect income effects. Foreign-Sourced Revenue And Input Cost Approach Another way to approach this issue is to identify the U.S. industries that garner the largest proportion of total revenues from Mexico and Canada. Unfortunately, few companies provide much country detail on where their foreign revenues are derived. Many simply split U.S. and non-U.S. revenues, or North American and non-North American revenues. Table II-2 presents the proportion of total revenues that is generated from operations outside the U.S. for the top five companies in the industry by market cap (in some cases the proportion that is generated outside of North America was used as a proxy for foreign- sourced revenues). While this approach is not perfect, it does provide a good indication of how exposed a U.S. industry is to Canada and Mexico. This is because any company that has "gone global" will very likely be doing substantial business in these two countries. Table II-2Foreign Revenue Exposure At the top of the list are the Metals & Mining, Personal Products, and Auto Component industries. Between 62% and 81% of revenues in these three industries is derived from foreign sources. Following that is Household Durables, Leisure Products, Chemicals and Tobacco. Indeed, all of the level three GICs industries we are analyzing are moderately-to-highly globally-oriented, with the sole exception of Construction Materials. Table II-3Import Tariff Exposure U.S. companies are also exposed to U.S. tariffs that boost the price of imported inputs to the production process. This can occur directly when firm A imports a good from abroad, and indirectly, when firm A then sells its intermediate good to firm B at a higher price, and then on to firm C. In order to capture the entire process, we used the information contained in the Bureau of Economic Analysis' Input/Output tables. We estimated the proportion of each industry's total inputs that would be affected by a rise in tariffs to MFN levels. We then allocated the industries contained in the input/output tables to the 21 GICs level 3 industries we are considering, in order to obtain an import exposure ranking in S&P industry space (Table II-3). All 21 industries are significantly vulnerable to rising input costs, which is not surprising given that we are focusing on the manufacturing-based GICs industries and NAFTA focused on trade in goods. The vast majority of the industries could face a cost increase on 50% or more of their intermediate inputs to the production process. The Automobile industry is at the top of the list, with 72% of its intermediate inputs potentially affected by the shift up in tariffs (Automobile Components is down the list, at 56%). Containers & Packaging, Oil & Gas, Aerospace & Defense, Textiles and Food Products are also highly exposed to tariff increases. The automobile industry is a special case because of the safeguards built into NAFTA regarding rules-of-origin and the associated tracing list. The U.S. is seeking significant changes in both in order to tilt the playing field toward U.S. production, but this could severely undermine the intricate supply chain linking the three countries. Box II-2 provides more details. Box II-2 Automotive Production In NAFTA; Update Required We are focused on two key aspects to the renegotiation of the NAFTA rules that could have far reaching implications for automakers and the auto component maker supply base: the tracing list and country of origin rules. Regarding the first of these, the Trump administration has a legitimate gripe when it comes to automotive production. A tracing list was written in the early-1990's to define automotive components such that the rules of origin (ROO) could be easily met; anything not on the list is deemed originating in North America. As anyone who has driven a vehicle of early-1990's vintage and one of late-2010's vintage can attest, high tech components (largely not included on the tracing list) have grown exponentially as a percentage of the cost of the vehicle and, at least with respect to electronic and display components, are sourced mostly from overseas. Updating the tracing list would force auto makers to source a significantly greater amount of components domestically, almost certainly raising the cost of the vehicle and either hurting margins or hurting competitiveness through higher prices. The current NAFTA ROO require that 62.5% of the content of a vehicle must be sourced in North America, with no distinction between any of the member nations. The result of this legislation has been the creation of a highly integrated supply base that sees components move back and forth across borders through each stage of the manufacturing process. Early proposals from the Trump administration for a NAFTA rework included a country of origin provision for as much as 50% U.S. content. Such a provision would certainly cause a massive disruption in the automotive supply chain with components manufacturers forced to relocate or automakers electing to source overseas and pay the 2.5% MFN tariff on exports within North America. Either scenario presents a headwind to the tightly woven auto components base, underscoring BCA's U.S. Equity Strategy's underweight recommendation on the sector. The recently announced bilateral trade deal with Mexico raises the ROO content requirements to 75% from the 62.5% contemplated under NAFTA but, importantly, no country of origin provisions appear in the new deal. Still, given how quickly this is evolving, a final NAFTA deal could be significantly different. Chart II-4 presents a scatter diagram that compares import tariff exposure (horizontal axis) with foreign revenue exposure (vertical axis). The industries in the north-east corner of the diagram are the most exposed to NAFTA failure. The problem is that there are so many in this region that it is difficult to choose the top two or three, although Metals & Mining stands out from the rest. It is easier to identify the industries that face less risk in relative terms: Pharmaceuticals, Construction Materials, Health Care & Supplies, Leisure Products and, perhaps, Machinery. The rest rank highly in terms of both foreign revenue exposure and import tariff exposure. Chart II-4Foreign Revenue And Import Tariff Exposure Conclusions: By itself, a total cancelation of NAFTA would not be devastating for any particular U.S. industry because the size of the tariff increases would be fairly small as long as all parties stick with MFN tariff levels. That said, the impact would not be trivial, especially for those industries that have extensive supply lines that run between the three countries involved. The negative impact on GDP growth would likely be worse for Canada (and Mexico if its bilateral somehow fell through), but U.S. exporters would see some loss of business. We approached the issue from four different perspectives; international supply chains, a model-based approach, and an analysis of foreign revenue exposure and import tariff exposure. The broad conclusion is that there are no winners from a NAFTA cancelation for the U.S. manufacturing GICs industries. Pharmaceuticals, Health Care Equipment & Supplies, Personal Products and Construction Materials are lower on the risk scale, but cannot be considered beneficiaries of a NAFTA collapse. The remaining industries are all moderately-to-highly exposed. Considering the four perspectives as a group, the most vulnerable industries are Automobiles, Automobile Components, Metals & Mining, Food Products, Beverages, and Textiles & Apparel. Our U.S. equity sector specialists recommend overweight positions in Defense and Financials; while neither stands to benefit from a NAFTA abrogation, they should at least be relative outperformers. They recommend underweight positions on Auto Components, Steel and Electrical Components & Equipment as relative (and probably absolute) underperformers should NAFTA disappear. As we go to press, rapid developments are taking place in the NAFTA negotiations. The U.S. and Mexico have completed a bilateral agreement in principle and a Canadian team is looking into whether to sign onto the agreement by a U.S.-imposed August 31 deadline. This deadline would enable the current U.S. Congress to proceed to ratification before turning over its seats in January, though it is not a hard deadline. It is possible that the negotiations will conclude this week and the crisis will be averted. But the lack of constraints on President Trump's trade authority gives reason for pause. If Canada demurs, Trump could move to raise the cost through auto tariffs or announcements that he intends to withdraw from existing U.S.-Canada agreements in advance of November 6. While Mexico has now tentatively secured bilaterals with both countries through the new U.S. deal and the Trans-Pacific Partnership (which includes Canada), it still stands to suffer if a trilateral agreement is not in place. Moreover it is technically possible that Canada's refusal to join the U.S.-Mexico bilateral could delay the latter's ratification well into next year. Therefore, we treat Mexico the same as Canada in our analysis, despite the fact that Mexican assets stand to benefit in relative terms from having a floor put under them by the Trump Administration's more constructive posture and this week's framework deal. If Trump does not pursue a hard line with Canada, then it will be an important sign that he is adjusting his trade policy to contain the degree of confrontation with the developed nations and allies and instead focus squarely on China, where we expect trade risks to increase in the coming months. Mark McClellan Senior Vice President The Bank Credit Analyst Matt Gertken Associate Vice President Geopolitical Strategy Chris Bowes Associate Editor U.S. Equity Strategy APPENDIX TABLE II-1 U.S. Imports From Canada (2017) APPENDIX TABLE II-2 U.S. Exports To Canada (2017) APPENDIX TABLE II-3 U.S. Imports From Mexico (2017) APPENDIX TABLE II-4 U.S. Exports To Mexico (2017) 1 Please see BCA Geopolitical Strategy Special Report, "A Mexican Standoff - Markets Vs. AMLO," dated June 28, 2018, and Weekly Report, "Are You 'Sick Of Winning' Yet?" dated June 20, 2018, available at gps.bcaresearch.com 2 Please see BCA Geopolitical Strategy and Global Investment Strategy Special Report, "NAFTA - Populism Vs. Pluto-Populism," dated November 10, 2017, available at gps.bcaresearch.com 3 Working Together: Economic Ties Between the United States and Mexico. Christopher E. Wilson, November 2011. 4 The NAFTA Renegotiation: What if the U.S. Walks Away? The C.D. Howe Institute Working Paper. November 2017. III. Indicators And Reference Charts Our equity indicators continue to signal that caution is warranted, but U.S. profits have been so strong recently as to dominate any negative market forces. Our Monetary Indicator is hovering at a low level by historical standards, suggesting that liquidity conditions have tightened. It is constructive that our Composite Technical Indicator has hooked up, narrowly avoiding a technical break below the zero line. It is also positive that our Composite Sentiment Indicator is rising, but not yet to a level that would signal trouble for stocks from a contrary perspective. However, our U.S. Willingness-to-Pay (WTP) indicator continues to erode, and the Japanese WTP appears to be rolling over. The WTP indicators track flows, and thus provide information on what investors are actually doing, as opposed to sentiment indexes that track how investors are feeling. Flows into the U.S. stock market are waning, and those into the Japanese market are wavering. Flows into European stocks have flattened off. Moreover, our Revealed Preference Indicator (RPI) for stocks remained on a 'sell' signal in August. The RPI combines the idea of market momentum with valuation and policy measures. It provides a powerful bullish signal if positive market momentum lines up with constructive signals from the policy and valuation measures. Conversely, if constructive market momentum is not supported by valuation and policy, investors should lean against the market trend. These indicators are not aligned at the moment, further supporting the view that caution is warranted. Our indicators thus suggest that the underlying health of the U.S. equity bull market is fraying at the edges. Nonetheless, robust U.S. profits figures have sparked a euphoric late-cycle blow-off phase. The net revisions ratio is still in positive territory, and the net earnings surprises index has surged to an all-time high. Not much has changed on the U.S. Treasury front. The 10-year bond is slightly on the cheap side according to our model, and oversold conditions have not yet been worked off. This month's Overview section discusses the potential for upside inflation surprises in the U.S. that will place the FOMC "behind the curve". The term premium and long-term inflation expectations are still too low. This year's dollar rally has taken it to very expensive levels according to our purchasing power parity estimate. The long-term trend in the dollar is down, but economic and policy divergences vis-à-vis the U.S. and the other major economies suggests that the dollar is likely to continue moving higher in the near term. EQUITIES: Chart III-1U.S. Equity Indicators Chart III-2Willingness To Pay For Risk Chart III-3U.S. Equity Sentiment Indicators Chart III-4Revealed Preference Indicator Chart III-5U.S. Stock Market Valuation Chart III-6U.S. Earnings Chart III-7Global Stock Market And Earnings: ##br##Relative Performance Chart III-8Global Stock Market And Earnings: ##br##Relative Performance FIXED INCOME: Chart III-9U.S. Treasurys And Valuations Chart III-10Yield Curve Slopes Chart III-11Selected U.S. Bond Yields Chart III-1210-Year Treasury Yield ComponentsChart III-13U.S. Corporate Bonds And Health Monitor Chart III-14Global Bonds: Developed Markets Chart III-15Global Bonds: Emerging Markets CURRENCIES: Chart III-16U.S. Dollar And PPP Chart III-17U.S. Dollar And Indicator Chart III-18U.S. Dollar Fundamentals Chart III-19Japanese Yen Technicals Chart III-20Euro Technicals Chart III-21Euro/Yen Technicals Chart III-22Euro/Pound Technicals COMMODITIES: Chart III-23Broad Commodity Indicators Chart III-24Commodity Prices Chart III-25Commodity Prices Chart III-26Commodity Sentiment Chart III-27Speculative Positioning ECONOMY: Chart III-28U.S. And Global Macro Backdrop Chart III-29U.S. Macro Snapshot Chart III-30U.S. Growth Outlook Chart III-31U.S. Cyclical Spending Chart III-32U.S. Labor Market Chart III-33U.S. Consumption Chart III-34U.S. Housing Chart III-35U.S. Debt And Deleveraging Chart III-36U.S. Financial Conditions Chart III-37Global Economic Snapshot: Europe Chart III-38Global Economic Snapshot: China Mark McClellan Senior Vice President The Bank Credit Analyst

Informe especial

Highlights "When you come to a fork in the road, take it." - Yogi Berra The last time we invoked the great American philosopher Yogi Berra was in September 2015. Back then, the oil market was at a critical juncture, as the market-share war initiated by OPEC in November 2014 approached its dénouement.1 The signal feature of the oil market in September 2015 was a massive 1.5mm b/d oversupply that was rapidly filling storage globally. We noted this surplus "... either will be cleared gradually or convulsively. ... (H)igh-cost supply either will exit the market rationally ... or via sharp lurches toward cash-breakeven costs, as global inventories fill on the back of slowing demand in an oversupplied market. Either way, markets will balance." In the event, prices lurched sharply into the left tail of the distribution toward cash-breakevens, with Brent approaching $25/bbl in 1Q16 (vs. more than $100/bbl in mid-2014). Oil's at a critical juncture again. Only this time, prices are poised to push higher into the right tail of the distribution, ahead of the likely loss of 2mm b/d or more of exports on the back of U.S.-imposed sanctions against Iran, and the all-but-certain collapse of Venezuela's economy. In our modelling, these events - along with constrained U.S. shale oil output due to pipeline bottlenecks in the Permian basin, and still-strong demand assumptions - could send prices above $120/bbl.2 This is not a foregone conclusion, however. Downside risks to global oil demand - largely from tariffs and non-tariff barriers to trade, and the Fed's monetary policy - are building. In this Special Report, we expand our examination of downside risks to oil prices arising from divergent monetary policies at systematically important central banks, particularly their impacts on currency markets, which we began last week. Feature Chart of the WeekOil Prices And USD TWIB Share##BR##Long-Term Trend, Equilibrium We strongly believe Fed policy will, once again, become a key variable in the evolution of oil prices, mostly via the FX markets. As a result, our regular monthly oil price forecast will be complemented by an additional component: our U.S. trade-weighted dollar (USD TWIB) forecast. In the current market, this is a downside scenario not a revised expectation. The FX simulation we describe below for prices hugs the lower boundary of the 95-percent confidence interval we use to situate our scenarios within. This will allow us to judge our expectation against market-cleared expectations. Our thesis that the USD's appreciation earlier this year would have a moderate effect on the evolution of oil prices - i.e., that supply-demand fundamentals would dominate this evolution - has been spot-on so far in 2018.3 This is largely due to OPEC 2.0's remarkable production discipline, and strong demand, particularly out of EM economies, which caused global inventories to draw, and kept the forward curves for Brent and WTI backwardated. 4 However, with the U.S. economy powering ahead - growing at a 3.1% rate in 1H18 - and, per our House view, the Fed continuing to lean into its rates-normalization policy, the USD will rise ~ 5% over the next year.5 We have shown in the past how important the USD can be for oil prices. Our oil financial model uses the USD as its main explanatory variable, and shows these variables are cointegrated in the long run - i.e. they share a common long-term trend and are in an equilibrium relationship (Chart of the Week). Consequently, forecasting the U.S. dollar is crucial step in our oil-price forecasting process. The Fed And Oil Prices As the Iran sanctions approach, OPEC 2.0 has indicated - not in a particularly clear manner - that it will be increasing production. While it appears the producer coalition will raise output slower than it previously led the market to believe, it is raising output.6 In addition, the U.S. Strategic Petroleum Reserve (SPR) also will be releasing 11mm barrels of oil over the October - November period. This short-term measure will help keep gasoline prices down going into the U.S. mid-term elections. While OPEC 2.0 calibrates the output required to offset the Iran-Venezuela supply-side risks, demand growth is being threatened by tariff and non-tariff barriers to trade, and the Fed's monetary policy.7 Between tariffs and U.S. monetary policy, we believe Fed policy trumps U.S. trade policy ... at least for now. Fed Policy Trumps Tariffs A lot of ink has been spilled on the Sino - U.S. trade war, but so far, the actual damage done to the $17 trillion global trading markets is trivial (Chart 2). Of course, this could quickly change if the U.S. and China step up their tit-for-tat tariffs and both plunge into an all-out trade war. Fed policy is neither trivial nor local: It is a global macro variable, largely because it impacts the U.S. dollar directly. This is important for EM economies, especially as it pertains to trade. We have shown EM imports and exports are exquisitely sensitive to the USD TWIB.8 This makes the USD TWIB particularly important to commodity markets, since most of the world's traded commodities are priced in USD and EM demand dominates global demand. When the Fed is tightening, the dollar appreciates, and commodities priced in USD become more costly ex-U.S. at the margin. This lowers demand for goods priced in USD. In addition, a stronger USD lowers the cost of production ex-U.S., which, again, at the margin, incentivizes supply growth, since commodity producers effectively arbitrage their local currency weakness by selling their output for USD. This supply-side effect is tempered somewhat by the degree to which commodity producers ex-U.S. are exposed to dollar strength in their input markets. For example, if a producer's production inputs are priced in USD - e.g., drilling services - its margins suffer, and output increases are constrained or nullified. The Fed is the only systematically important central bank we follow - the others being the ECB, BoJ and PBoC - implementing and executing an interest-rate normalization policy. This has supported USD strength against the systematically important currencies we follow, as well (Chart 3). Chart 2Tariffs Are A Less Threat To Global Growth ... Chart 3Important Central Banks Keeping Policies Accommodative The IMF is encouraging the ECB to maintain its accommodative policy, and the BoJ also is keeping its policy relatively loose.9 The BoJ is keeping policy on hold for now, and is guiding to no rate hikes until 2020. Our colleagues in BCA's FX and Fixed Income desks expect the BoJ to continue with its Yield Curve Control Strategy for the remainder of the year, and most of next year. The absence of monetary tightening will keep Japanese yields lower than other major central banks. The PBoC appears to have moved toward a more accommodative mode, in the wake of the Sino - U.S. trade war. We believe the PBoC will remain accommodative in terms of official lending rates to avoid too-fast a deceleration of the economy, largely because of high private debt levels.10 EM Trade Volumes And Oil Prices Against a largely accommodative backdrop ex-U.S., the USD TWIB appreciated ~5% y/y, while the JP Morgan Emerging Markets FX index dropped ~11% (Chart 4). In the wake of USD TWIB strength, EM trade volumes have held up reasonably well; but growth rates have been under pressure particularly in Central and Eastern Europe (Chart 5, bottom panel). This is being offset by a turn-around in the Middle East and Africa (third panel). Chart 4Fed Policy Drives USD Strength Chart 5EM Trade Slowing, But Still Holding Up Assuming the Fed maintains its existing course re policy-rate normalization, our Fed-policy models indicate the USD TWIB will continue to strengthen (Chart 6).11 On the flip side of that, EM currencies will continue to weaken (Chart 7). This will keep pressure on EM trade volumes, particularly the important import volumes. Over the next year, we expect continued slowing in trade volumes, although, on average, we still expect y/y growth (Chart 8). While growth is slowing in EM trade, the levels of trade will remain high, unless a full-blown global trade war erupts that literally forces trade to contract. Chart 6Fed Policy Will Propel USD TWIB Higher... Chart 7... And Keep EM Currencies Weaker Chart 8Downward Trend In EM Trade Will Continue As USD Strengthens ... Strong USD, Weak EM Trade := Bearish Fed policy will strengthen the USD TWIB and weaken EM trade. These factors will tend to pull crude oil prices down, in and of themselves (Chart 9). We do not think these factors will dominate the evolution of crude oil prices over the next six months, however. That said, the current environment forces us to adapt our modelling procedure in order to account for the possible re-emergence of the USD as a key driver of oil prices. Going forward, our regular monthly oil price forecast will be complemented by our U.S. trade weighted dollar forecast.12 We will be rolling out our new oil-price forecasting models next month, when we update our supply-demand balances and price forecasts. For the immediate future, we continue to believe upside price risk dominates the oil market: The approaching U.S. sanctions against Iran and the all-but-certain collapse of Venezuela's economy could remove as much as 2mm b/d of exports from oil markets by next year, if not sooner. This would constitute an oil-price shock, pushing prices into the right tail of the price distribution, consistent with the modelling we've done for the past several months (Chart 10). Chart 9... Adding A Bearish Factor To##BR##The Evolution Of Brent, WTI Prices Chart 10Upside Risks##BR##Still Dominate We reiterate our conclusion from last week, however, that an oil-supply shock, coupled with slower EM trade growth ultimately will produce strong deflationary impulses. If markets avoid an oil supply shock, and if the Fed maintains its rates-normalization policy while the rest of the world's systemically important central banks remain accommodative, pressure will build on EM trade - and incomes - that reduces commodity demand, in line with lower aggregate demand from the EM economies. In either event, the Fed's rates-normalization policy most likely will have to turn accommodative to counter this. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com 1 Please see our Special Report entitled "Oil Volatility To Stay Higher Longer," published September 17, 2015. It is available at ces.bcaresearch.com. 2 We have written at length regarding this possible price evolution. Please see, e.g., BCA's Commodity & Energy Strategy Weekly Reports from August 16 and August 2, 2018, entitled "OPEC 2.0 Sailing Close To The Wind," and "Calm Before The Storm In Oil Markets." Both are available at ces.bcaresearch.com. 3 For more details, please see Commodity & Energy Strategy Weekly Report published February 8, 2018, "OPEC 2.0 Vs. The Fed." It is available at ces.bcaresearch.com. 4 OPEC 2.0 is the name we coined for the producer coalition lead by the Kingdom of Saudi Arabia (KSA) and Russia. 5 In Jackson Hole last week, Fed Chair Jerome Powell gave a strong endorsement of the Fed's rates-normalization. Please see "Fed Chair Powell: further rate hikes best way to protect recovery," published by reuters.com August 24, 2018. 6 On a 4Q19 vs 4Q18 basis, we expect global oil supply to increase just over 1mm b/d, and for demand to rise 1.8mm b/d, leaving the market in a physical deficit in 2H18 and 2019. We expect Brent to average $70/bbl in 2H18 and $80/bbl in 2019. Please see our updated balances estimates and price forecasts in "OPEC 2.0 Sailing Close To The Wind," published August 16, 2018, by BCA Research's Commodity & Energy Strategy. It is available at ces.bcaresearch.com. 7 Our $80/bbl forecast for Brent crude next year - and the physical deficit we expect - implicitly assumes OPEC 2.0 either wants to keep the market relatively tight, which will force inventories to draw and backwardate the forward curves, or that it is pushing up against the limits of the production it can readily bring to market. 8 We most recently discussed this in our Commodity & Energy Strategy Weekly Report published August 23, 2018, "Trade, Dollars, Oil & Metals ... Assessing Downside Risk." It is available at ces.bcaresearch.com. 9 Please see Abdih, Yasser, Li Lin, and Anne-Charlotte Paret (2018), "Understanding Euro Area Inflation Dynamics: Why So Low for So Long?" published by the IMF this month. 10 Please see BCA Research's Global Fixed Income Strategy Weekly Report, "An R-Star Is Born," dated August 7, 2018, and Foreign Exchange Strategy Weekly Report, "Rhetoric Is Not Always Policy", dated July 27, 2018, available at gfis.bcaresearch.com and fes.bcaresearch.com. 11 We have a suite of models we use to forecast the USD TWIB, many of which use proxies for the Fed's Congressionally mandated policy goals - i.e., maximum employment, stable prices and moderate long-term interest rates. We use cointegrating regressions to estimate these policy-driven models. The R2 coefficients of determination for the models are clustered around 0.95. The out-of-sample results are strong; we use a weighted-average of the five forecasts based on root-mean-square errors to come up with our USD TWIB forecast. We presented our policy-variables USD TWIB models in last week's Commodity & Energy Strategy Weekly Report. Please see "Trade, Dollars, Oil & Metals ... Assessing Downside Risk." It is available at ces.bcaresearch.com. 12 With the introduction of these financial and macro variables, our oil price forecast will be a weighted average of our core Fundamental model and the new Financial model - i.e. the final forecast will look like [aFundamental+(1-a)Financial]. The weights - a and (1-a) - are time-varying, and will reflect our Bayesian probabilities for the relative importance of each model's contribution to price action every month. These weights are crucial. We allow them to vary in order to capture periods in which our analysis tells us we should expect the USD effect to be muted by idiosyncratic supply, demand or inventory dynamics. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed in 2018 Summary of Trades Closed in 2017

Informe especial

Highlights Since 2010, China's private sector has accounted for the majority of the country's increase in the debt-to-GDP ratio, most of which has been on the balance sheets of state-owned enterprises (SOEs) and the household sector. While policymakers achieved their goal of maintaining aggregate demand in the decade following the global financial crisis, the financial condition of SOEs has been greatly sacrificed as a result. An analysis of SOE return on equity highlights a sharp decline in return on assets, which has occurred due to both declining profit margins and a falling asset turnover ratio. Even worse, a comparison of adjusted SOE ROA to borrowing costs suggests that the marginal operating gain from debt has become negative. This has profound implications for policymakers, as it suggests that further leveraging of SOEs could push them into a debt trap and/or shackle the monetary authority's ability to meaningfully raise interest rates. We can envision a modest releveraging scenario over the coming 12-18 months, but even that scenario is not consistent with a surge in investment-driven economic activity. Policymakers face a clear choice between growth and leveraging, and our bet is that they will choose just enough of the latter to prevent the former from decelerating significantly. This implies that the typical beneficiaries of Chinese reflation are not likely to outperform global risk assets, and that China's contribution to global growth is not set to rise sharply. However, over the coming 6-12 months, we acknowledge that domestic stocks are significantly oversold, and we are watching closely for an opportunity to time a reversal. Feature Global investors have paid considerable attention to China over the past month, focusing on the likely stimulative response of policymakers to an upcoming, tariff-induced export shock. We recently presented our view of the likely character and magnitude of upcoming Chinese stimulus in a two-part joint special report with our geopolitical team,1 and concluded that an acceleration in fiscal spending was far more likely than a sharp pickup in credit growth. In this report, we further examine the constraints facing Chinese policymakers and again conclude that they are likely to remain committed to preventing a significant releveraging of the economy. The financial condition of Chinese state-owned enterprises features prominently in our argument, and we highlight how the damage caused by China's post-2008 "business model" is a serious roadblock to further credit excesses. Whereas most modern central banks characterize their monetary policy decisions within the context of a trade-off between growth and inflation, Chinese policymakers now appear to understand that they face a trade-off between growth and leveraging. While we agree that economic stability will always remain the paramount objective of policymakers and a major policy mistake is not likely in the cards, reflationary efforts are likely to be carefully calibrated to avoid a dramatic overshoot of credit growth. This means that there is both limited downside and upside to Chinese economic activity, implying that expectations of a material, credit-driven reacceleration in growth are not likely to be met. A Brief Review Of Chinese Private Sector Debt Chart 1A Now Familiar Concern After several years of intense concern about China's elevated debt, Chart 1 should be familiar to most investors. It highlights the significant rise in Chinese credit to the non-financial sector (i.e. total credit to governments, households, and non-financial corporations) based on data from the Bank for International Settlements (BIS), most of which has occurred in the private sector (non-financial firms and households). But Charts 2-4 presents a different breakdown of credit to the non-financial sector, based on IMF data, that includes a separation of corporate debt into private and state-owned enterprises (SOEs). The data shown in Charts 2-4 covers the 2010-2016 period; for reference, private non-financial sector debt continued to rise relative to GDP in 2017, in large part due to households (see Table A1 in Appendix 1 for the most recent IMF estimate of China's non-financial sector debt, absent the breakdown in corporate debt by ownership that the fund previously provided). Chart 2 presents the IMF's version of the rise in total non-financial debt (akin to Chart 1 from the BIS), and Charts 3 and 4 attribute the rise in debt to different sectors. Chart 3 shows that the increase in private sector debt accounts for 70% of the increase in leverage since 2010, and Chart 4 shows that the rise in SOE debt has accounted for nearly half of the rise in private sector debt. Within the private sector, household leverage has also risen substantially, accounting for roughly 40% of the rise from 2010-2016. Non-SOE corporates accounted for only 12% of the total rise in private leverage, the smallest of all sectors. Chart 2Another Perspective On Chinese Leveraging, With A Breakdown Of Corporate Debt By Ownership   Chart 3The Private Sector Has Accounted For ##br##Most Of Chinese Leveraging... Chart 4...Due Mostly To State-Owned ##br##Enterprises And Households     When considering the potential economic impact of a sharp rise in leverage, BCA's view is that the focus should usually be on the increase in private sector debt rather than government debt. Public sector deleveraging is fundamentally a political choice in countries that have control over their own monetary policy, and simply will not occur in China over the coming year given the headwinds facing the economy. Given this, Chart 4 suggests that to understand any constraints facing policymakers from excessive leverage, investors should primarily devote their attention towards China's SOEs. China's State-Owned Enterprises: The Sacrifice Of Profitability For Stability Chart 5Within SOEs, Industrial And Construction Firms ##br##Account For Half Of The Increase In Debt When assessing the risk of a potential private sector debt crisis in China, many investors have a sanguine view. The common refrain is that Chinese corporations, particularly state-owned enterprises, will be bailed out by the government if debt problems arise. Ultimately, we agree with this view, although we would note that the market pressure required to force the government to act could be quite severe. Still, there is a more pressing concern for investors: an analysis of the financial condition of China's state-owned enterprises suggests that the country may have reached the limit of how much SOEs can be further leveraged by policymakers in an attempt to rescue the economy, without significantly increasing the ultimate cost to the public. Our sense is that the campaign to control debt growth over the past two years reflects this economic reality, suggesting that the motivation behind the campaign will not be easily abandoned. Chart 2 showed that non-financial SOE debt-to-GDP rose by 20 percentage points from 2010-2016, a change in the stock of debt of roughly RMB33 trillion. Chart 5 shows that roughly half of this amount can be accounted for by the change in liabilities of state-owned industrial and construction enterprises over the same period. To the extent that they broadly reflect the condition of all non-financial SOEs, the availability of income statement and balance sheet data for these two industries allows us to make some inferences about the debt sustainability of China's state-owned firms.   Table 1 presents a breakdown of return on equity (ROE) for state-owned/state-holding companies in these industries, using the DuPont approach. Several points are noteworthy: Industrial & construction SOEs are highly leveraged entities, with an assets to equity ratio of 2.7. This explains the substantial difference between return on equity, which has been decently high, and a low single-digit return on assets (ROA). From 2010-2016, the ROE for industrial & construction SOEs fell from 14% to 8%, entirely because of a substantial decline in ROA. The decline in ROA occurred because of a roughly equal combination of declining profit margins and a falling asset turnover ratio. Based on the DuPont approach to expressing leverage,2 SOEs in the industrial and construction industries increased their leverage only very modestly during the period. But when leverage is expressed as liabilities relative to net income, a considerably more relevant measure when considering the potential to service debt, leverage nearly doubled. Table 1A Meaningful Decline In SOE Efficiency And Profitability We presented Chart 6 in our last weekly report of 2017,3 and used it to represent a stylized timeline of China's economic history over the past 15 years. The chart describes how China's extremely rapid growth phase from 2002-2008 was followed by the global financial crisis and a normal, counter-cyclical rise in the debt-to-GDP ratio from 2008 to 2010. Chart 6A Stylized Timeline Of China's Recent Economic History However, amidst the Great Recession, it became clear that China's export-enabled catchup growth phase was durably over, and policymakers were faced with a hard choice. They could either replace exports with debt-fueled domestic demand as a growth driver in order to buy time to transition to a services-led economy (the "reflate" path), or allow the labor market to suffer the consequences of a sharp slowdown in export growth while preserving fiscal and state-owned firepower for some uncertain future opportunity (the "stagnate" path). The picture that emerges from the combination of this narrative and our analysis of the evolution of SOE financial health is straightforward, but sobering. State-owned enterprises, already highly indebted at the onset of the global economic recovery, were levered even further in order to pursue the "reflate" path described above. While policymakers achieved their goal of maintaining aggregate demand, the consequence of their choice is that both the profitability and efficiency of SOEs have declined significantly. Avoiding An SOE Debt Trap A significant deterioration in SOE efficiency against the backdrop of a sharp rise in leverage speaks to the existence of capital misallocation, i.e. investment that has been funded by debt but cannot produce sufficient income to repay the debt. This suggests that SOEs are likely to have a bad debt problem at some point that will need to be resolved with government support. But in our view, the decline in profitability is a more immediate problem for policymakers, because it does not appear that SOEs can be leveraged any further without pushing them dangerously towards a self-reinforcing debt trap. Chart 7 illustrates why. The chart shows SOE ROA adjusted for interest expenses (a proxy for EBIT/Assets) versus a market-based proxy for SOE borrowing rates.4 Adjusted ROA fell below borrowing rates in 2013, suggesting that some of the observed decline in SOE profitability has occurred because the marginal operating gain from debt for Chinese state-owned enterprises has become negative. If so, this has profound implications for Chinese policymakers. Chart 8 illustrates how the process of perpetually leveraging an entity with a negative marginal operating gain from new borrowing eventually leads to a debt trap. An initial increase in debt causes interest costs to rise and profits to fall, as the return on new assets fails to exceed the interest rate on the debt used to acquire the assets. The process repeats itself as the entity is directed to leverage further, although management may choose to raise the entity's debt in this situation regardless of policy objectives (e.g. to cover a working capital deficit) if they mistakenly believe that the decline in ROA below debt costs is temporary. In addition, the existence of a negative marginal gain from new borrowing for a significant portion of the private sector would imply that China's natural rate of interest may have fallen. Chart 9 shows some evidence in support of this notion: the rise in the weighted average lending rate since late-2016 was relatively minor compared with levels that have prevailed over the past decade, and yet it is clear that it succeeded in materially slowing the investment-driven sectors of China's economy. This suggests that further leveraging of SOEs could tighten the shackles on the PBOC in terms of its ability to meaningfully raise interest rates, potentially fueling credit excesses in other sectors of the economy Chart 7SOEs Now Appear To Have A Negative ##br##Financial Gain From Debt Chart 8A Stylized Example Of ##br## Debt Trap Dynamics Chart 9Has SOE Leveraging Caused China's ##br##Natural Rate Of Interest To Fall?         In short, the financial condition of China's state-owned enterprises appears to represent a proximate constraint preventing policymakers from responding to economic weakness with a significant acceleration in credit growth. It is not just that SOEs are highly levered and there is "a lot of debt in the system"; material further leveraging of these entities risks deteriorating what is already very poor profitability, which may push SOEs into an outright debt trap. That would precipitate a crisis and necessitate a bailout from the government, the cost of which will increase directly in line with the amount of additional debt taken. We agree that economic stability will always remain the paramount objective of policymakers, and we fully expect a policy response to address the upcoming export shock from the U.S. But whereas most modern central banks characterize their monetary policy decisions within the context of a trade-off between growth and inflation, our analysis of China's state-owned enterprises suggests that Chinese policymakers now seem to understand that they face a trade-off between growth and leveraging. This implies that current reflationary efforts from policymakers are likely to be carefully calibrated to avoid a dramatic overshoot of credit growth. Envisioning Modest Releveraging Chart 10Modest Releveraging Is Ok, As Long As ##br##Its Pace Continues To Slow What is a carefully calibrated credit response likely to look like, and what does it mean for private sector debt growth? As noted above, my colleague Matt Gertken addressed this question by presenting three scenarios in part 1 of the recent joint special report with our geopolitical team.5 His base-case view, to which he assigned 70% odds, implied that there would be a very modest reacceleration in total social financing (on the order of 1% or so). In this report we take a second approach to estimating the potential magnitude of a modest reacceleration scenario using the BIS private sector credit data, primarily to incorporate different growth rates for the corporate and household sectors. Using the BIS data, Chart 10 shows the growth rate in Chinese total private sector debt, nominal GDP, and the difference between the two. The significant leveraging period from 2010-2016 is evidenced by the persistently positive gap between credit and GDP growth (it was only briefly negative in 2011).   But the chart also shows that there has been a downtrend in the gap since 2013, with 2017 representing a major overshoot (to the downside). Given that the trend shown in Chart 10 points downward and reflects policy efforts to control debt growth, we could envision Chinese policymakers tolerating some acceleration in credit growth relative to GDP, as long as it does not materially overshoot the trendline to the upside. Using this framework as a guide, we can calculate what modest releveraging might mean for corporate sector debt, assuming the following: Chinese policymakers, through a combination of fiscal spending and modest releveraging, succeed in stabilizing nominal GDP growth at current levels. Policymakers tolerate total non-financial private sector credit growth that is 4% in excess of nominal GDP growth. Household credit growth remains well in excess of GDP growth, in-line with its average of the past 5 years. Given the significant leveraging of the household sector and the recent uptick in home sales, this appears to be a reasonable assumption barring a major crackdown on the property market by Chinese officials. Chart 11 presents the result of these assumptions, which shows non-financial corporation credit growth accelerating to roughly 12% by the end of 2019. At first blush, the chart appears to show a meaningful acceleration, as the annual change in year-over-year credit growth based on this measure would meet or exceed that of the past two credit cycles. But there are two important caveats for investors: Even as depicted in Chart 11, non-financial corporate credit growth would still be extremely weak relative to its recent history. At the end of 2019, the chart shows that corporate credit growth would be almost two percentage points lower than its weakest point in 2015. Chart 11 illustrates a scenario where the level of credit to the total private non-financial sector grows by RMB36 trillion by the end of 2019. Chart 12 shows that when compared to our estimate of the stock of adjusted total social financing, this rise barely even registers as an acceleration. Chart 11A Rebound, But Weak Relative To History Chart 12Barely Even Registers As An Acceleration In Adjusted TSF In short, while the degree of acceleration in credit growth as implied in our scenario varies depending on the definition of credit employed, the bottom line for investors is that a modest releveraging scenario is not consistent with a surge in investment-driven economic activity. Policymakers face a clear choice between growth and leveraging, and our bet is that they will choose just enough of the latter to prevent the former from decelerating significantly. This cautious, contingent attitude towards an acceleration in private sector credit growth would be in marked contrast to previous episodes of reflation, suggesting that investors who are following China's "old stimulus rulebook" are likely to be disappointed. Implications For Investment Strategy Chart 13No Signs Yet Of A Heavy, Credit-Based Response There are two clear implications of our analysis for investment strategy. First, in ironic reference to Reinhart & Rogoff's book that coined the term, "this time" is likely to be different for China because policymakers seem resolute in their intention to prevent a financial crisis (as opposed to the term having been used in the past by those who have ended up contributing to one). Our analysis shows that the debt burden for state-owned enterprises is already extreme, and that further, material, forced leveraging of the sector risks a possible debt trap. This implies that the typical beneficiaries of Chinese reflation are not likely to outperform global risk assets, and that China's contribution to global growth is not set to rise sharply. For now, our BCA China Play Index and the relative performance of infrastructure stocks seem to support our conclusion (Chart 13). Second, if this time is not different, i.e. if policymakers allow a significant further releveraging of the private sector, either intentionally or by accident, investors should recognize that the longer-term outlook for China may darken considerably if the country is not capable of quickly shifting away from its old growth model over the next few years. Unfortunately for officials in China, the reality of economics is that positive NPV projects for SOEs to invest in cannot simply be willed into existence. The significant decline in profitability and asset turnover that we have observed in state-owned enterprises since 2010 speaks to the poor use of credit, and policymaker reliance on the traditional methods of stimulus is likely to achieve the country's short-term goals at the expense of making the already large debt problem (and the cost of the eventual bailout by the public sector) much worse. This would raise both the political and economic risks facing the country, at a time when a U.S. and/or global recession appears likely within the next 2-3 years. As a final point, despite our caution against over-optimism concerning China's stimulative response, we acknowledge that policymakers are likely to succeed in preventing a significant deceleration in their economy over the coming 6-12 months. Given how materially Chinese stock prices have declined, it remains a debate whether a mere stabilization of economic activity at a modest pace will be enough for domestic or investable equities to meaningfully rally in absolute or relative terms. For now, we have highlighted that the relative selloff in domestic stocks appears to be quite late, particularly in common currency terms, and we are watching closely for an opportunity to time a reversal.   Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com   Appendix 1 Appendix A-1Chinese Non-Financial Sector Debt 1 Pease see China Investment Strategy Special Reports "China: How Stimulating Is The Stimulus?", dated August 8, 2018, and "China: How Stimulating Is The Stimulus? Part Two", dated August 15, 2018 available at cis.bcaresearch.com 2 The DuPont approach breaks down return on equity into the product of profit margins (profits / revenue), asset turnover (revenue / assets), and financial leverage (assets / equity). 3 Pease see China Investment Strategy Weekly Report "Legacies Of 2017", dated December 21, 2017, available at cis.bcaresearch.com 4 We use the yield-to-maturity of the ChinaBond Corporate Bond Index as our proxy for the interest rate paid by state-owned firms, given that the index includes bonds issued by central and local government SOEs. Importantly, our proxy is closely aligned with the weighted average bank loan borrowing rate paid by SOEs from 2014-2016, as per a 2017 report from the China Academy of Fiscal Science ("Cost reduction: 2017 survey and analysis", August 28, 2017). 5 Pease see China Investment Strategy Special Reports "China: How Stimulating Is The Stimulus?", dated August 8, 2018, and "China: How Stimulating Is The Stimulus? Part Two", dated August 15, 2018 available at cis.bcaresearch.com Cyclical Investment Stance Equity Sector Recommendations

Informe especial

Highlights Xi Jinping is trying to do two things at once: ease policy while cracking down on systemic financial risk; The trade war with the U.S. is a genuine crisis for China and is eliciting fiscal stimulus; Credit growth is far more likely to "hold the line" than it is to explode upward or collapse downward; The 30% chance of a policy mistake from financial tightening has fallen to 20% only, as bad loan recognition is underway and a critical risk to monitor; Hedge against the risk of a stimulus overshoot. China's policy headwinds have begun to recede, but Beijing is not riding to the rescue for emerging markets; While monetary policy has eased substantively, credit growth will be hampered by the government's financial crackdown; Potential changes to China's Macro-Prudential Assessment framework could be significant, but the impact on credit growth is overestimated at present; The recognition of non-performing loans (NPLs) and cleansing of China's banking system is still in early innings and will weigh on banks' risk appetite; The anti-corruption campaign is another reason to be cautious on EM. Geopolitical Strategy recommends clients stay overweight China (ex-tech) relative to EM. Feature "We have upheld the underlying principle of pursuing progress while ensuring stability." - Xi Jinping, General Secretary of the Communist Party of China, October 18, 2017 "Any form of external pressure can eventually be transformed into impetus for growth, and objectively speaking will accelerate supply-side structural reforms." - Guo Shuqing, Secretary of the China Banking and Insurance Regulatory Commission, July 5 PART I Last year we made the case that China's General Secretary Xi Jinping would double down on his reform agenda in 2018, specifically the bid to control financial risk, and that this would bring negative surprises to global financial markets as policymakers demonstrated a higher pain threshold.1 This view has largely played out, with economic policy uncertainty spiking and a bear market in equities developing alongside an increase in corporate and even sovereign credit default risk (Chart 1). We also argued, however, that Xi's "deleveraging campaign" would be constrained by the Communist Party's need for overall stability. Trade tensions with the U.S., and Beijing's perennial fear of unemployment, would impose limits on how much pain Beijing would ultimately tolerate: The Xi administration will renew its reform drive - particularly by curbing leverage, shadow banking, and local government debt. Growth risks are to the downside. But Beijing will eventually backtrack and re-stimulate, even as early as 2018, leaving the reform agenda in limbo once again.2 Over the past month, China has clearly reached its pain threshold: authorities have announced a series of easing measures in the face of a slowing economy, a trade war, and a still-negative broad money impulse (Chart 2). Chart 1Policy Uncertainty Up, Stocks Down Chart 2PMI Falling, Money Impulse Still Negative How stimulating is the stimulus? Will it lead to a material reacceleration of the Chinese economy? What will it mean for global and China-dedicated investors? We expect policy to be modestly reflationary. A substantial boost to fiscal thrust, and at least stable credit growth, is in the works. Yet Xi's reform agenda will remain a drag on the economy. While this new stimulus will not have as dramatic an effect as the stimulus in 2015-16, it will have a positive impact relative to expectations based on China's performance in the first half of the year. We advise hedging our negative EM view against a rally in China plays and upgrading expectations for Chinese growth in 2019. The policy headwind is receding for now. Xi Jinping's "Three Tough Battles" Chart 3Xi Jinping Caps Government Spending And Credit Xi will not entirely abandon the "Reform Reboot" that began last October. From the moment he came to power in 2012-13, he pursued relatively tight monetary and fiscal policy. Total government spending growth has dropped substantially under his administration, while private credit growth has been capped at around 12% (Chart 3). Xi partly inherited these trends, as China's credit growth and nominal GDP growth dropped after the massive 2008 stimulus. But he also embraced tighter policy as a way of rebalancing the economy away from debt-fueled, resource-intensive, investment-led growth. A comparison of government spending priorities between Xi and his predecessor makes Xi's policy preferences crystal clear: the Xi administration has increased spending on financial and environmental regulation, while minimizing subsidies for housing and railways to nowhere (Table 1 and 2). Table 1Central Government Spending Preferences (Under Leader's Immediate Control) Table 2Total Government Spending Preferences (Under Leader's General Control) These policies are "correct" insofar as they are driven not merely by Xi's preferences but by long-term constraints: The middle class: Pollution and environmental degradation threaten the living standards of the country's middle class. Broadly defined, this group has grown to almost 51% of the population, a level that EM politicians ignore only at their peril (Chart 4). Asset bubbles: The rapid increase in China's gross debt-to-GDP ratio since 2008 is a major financial imbalance that threatens to undermine economic stability and productivity as well as Beijing's global aspirations (Chart 5). The constraint is clear when one observes that "debt servicing" is the third-fastest category of fiscal spending growth since Xi came to power (Table 2). Chart 4Emerging Middle Class A Latent Political Risk Chart 5The Rise And Plateau Of Macro Leverage The problem is that Xi also faces a different, shorter-term set of constraints arising from China's declining potential GDP, "the Middle-Income Trap," and the threat of unemployment.3 The interplay of these short- and long-term constraints has forced Xi to vacillate in his policies. In 2015, the threat of an economic "hard landing," ahead of the all-important mid-term party congress in 2017, forced him to stimulate the "old" industrial economy and sideline his reforms. Only when he had consolidated power over the Communist Party in 2016-17 could he resume pushing the reform agenda.4 In July 2017, Xi announced the so-called "Three Critical Battles" against systemic financial risk, pollution, and poverty. The three battles are interdependent: continuing on the capital-intensive economic model will overwhelm any efforts to cut excessive debt or pollution (Chart 6), yet sudden deleveraging could derail the Communist Party's basic claim to legitimacy through improving the lot of poor Chinese. The macroeconomic impact of the three battles is broadly deflationary, as credit growth falls and industries restructure. The first battle - the financial battle - will determine the outcome of the other two battles as well as the growth rate of China's investment-driven economy, Chinese import volumes, and emerging market stability (Chart 7). Chart 6Credit Stimulus Correlates With Pollution Chart 7Credit Determines Growth And Imports On July 31, in the midst of worldwide speculation about China's willingness to stimulate, Xi reaffirmed this "Three Battles" framework. Remarkably, despite a general slowdown, a sharp drop in the foreign exchange rate, the revival of capital flight, and a bear market, he announced that the battle against systemic financial risk would continue in the second half of 2018. However, he also admitted that domestic demand needed a boost in the short term. Hence there should be no doubt in investors' minds about the overarching policy framework or Xi Jinping's intentions in the long run. The question driving the markets today is what China will do in the short term and whether it will initiate a material reacceleration in economic activity. Bottom Line: Xi Jinping remains committed to the reform agenda that he has pursued since coming to power in 2012. But he is forced by circumstances to vary the pace and intensity. At the top of the agenda is the control of systemic financial risk. This is a policy driven by the belief that China's economic and financial imbalances threaten to undermine its overall stability and global rise. Why The Shift Toward Easier Policy? The gist of the July 31 Politburo statement was that policy will get more dovish in the short term. It mentioned "stability" five times. The Politburo pledged to make fiscal policy "more proactive" and to find a better balance between preventing financial risks and "serving the real economy." This direct promise from Xi Jinping of more demand-side support gives weight to the State Council's similar statement on July 23 and will have reflationary consequences above and beyond the central bank's marginal liquidity easing thus far. What is motivating this shift in policy, which apparently flies in the face of Xi's high-profile deleveraging campaign? If we had to name a single trigger for China's change of tack, it is not the economic slowdown so much as the trade war with the United States. The war began when the U.S. imposed sanctions on Chinese firm ZTE in April and China depreciated the RMB, but it escalated dramatically when the U.S. posted the Section 301 tariff list in June (Chart 8).5 This is a sea change in American policy that is extremely menacing to China. China runs a large trade surplus and has benefited more than any other country from the past three decades of U.S.-led globalization. Its embrace of globalization is what enabled the Communist Party to survive the fall of global communism! Chart 8More Than Market Dynamics At Work Chart 9China Is Less Export-Dependent True, China has already seen its export dependency decline (Chart 9). But Beijing has so far managed this transition gradually and carefully, whereas a not-unlikely 25% tariff on $250-$500 billion of Chinese exports will hasten the restructuring beyond its control (Chart 10). A very large share of China's population is employed in manufacturing (Chart 11). To the extent that the tariffs actually succeed in reducing external demand for Chinese goods, these jobs will be affected. Chart 10Tariffs Will Add More Pain To Factory Workers Chart 11Manufacturing Unemployment A Huge Threat Unemployment is anathema to the Communist Party. And China is simply not as experienced as the U.S. in dealing with large fluctuations in unemployment (Chart 12). While Chinese workers will blame "foreign imperialists" and rally around the flag, the pain of unemployment will eventually cause trouble for the regime. Domestic demand as well as exports will suffer. It is even possible that worker protests could evolve into anti-government protests. Chart 12China Not Experienced With Layoffs Given that Chinese and global growth are already slowing, it is no surprise that the Politburo statement prioritized employment.6 China's leaders will prepare for social instability as the worst possible outcome of the showdown with America - and that will push them toward stimulus. In addition, there will be no short-term political cost to Xi Jinping for erring on the side of stimulus, as there is no opposition party and the public is not demanding fiscal and monetary austerity. Moreover, the main macro implication of Xi's decision last year to remove term limits - enabling himself to be "president for life" in China - is that his reforms do not have to be achieved by any set date. They can be continually procrastinated on the basis that he will return to them later when conditions are better.7 The policy response to tariffs from the Trump administration also signals another policy preference: perseverance. Xi would not be straying from his reform priorities if not for a desire to counter American protectionism. China is not interested in kowtowing but would rather gird itself for a trade war. Still, our baseline view is that the Xi administration will stimulate without abandoning the crackdown on shadow lending or launching a massive "irrigation-style" credit surge that exacerbates systemic risk.8 Policy will be mixed, as Xi is trying to do two things at once. Bottom Line: China's slowdown and the outbreak of a real trade war with the United States is forcing Xi Jinping to ease policy and downgrade the urgency of his attempt to tackle systemic financial risk this year. Can Fiscal Easing Overshoot? Yes. How far will China's policy easing go? China has a low level of public debt, and fiscal policy has been tight, so we fully expect fiscal thrust to surprise to the upside in the second half of the year, easily by 1%-2% of GDP, possibly by 4% of GDP. Chart 13Fiscal Tightening Was The Plan For 2018 A remarkable thing happened this summer when researchers at the People's Bank of China and the Ministry of Finance began debating fiscal policy openly. Such debates usually occur during times of abnormal stress. The root of the debate lay in the national budget blueprint laid out in March at the National People's Congress. There, without changing official rhetoric about "proactive fiscal policy," the authorities revealed that they would tighten policy this year, with the aim of shrinking the budget deficit from 3% of GDP target in 2017 to 2.6% in 2018. The IMF, which publishes a more realistic "augmented" deficit, estimates that the deficit will contract from 13.4% of GDP to 13% (Chart 13). This fiscal tightening coincided with Xi's battle against systemic financial risk. Hence both monetary and fiscal policy were set to tighten this year, along with tougher regulatory and anti-corruption enforcement.9 Thus it made sense on May 8 when the Ministry of Finance revealed that the quota for net new local government bond issuance this year would increase by 34% to 2.18 trillion RMB. This quota governs new bonds that go to brand new spending (i.e. it is not to be confused with the local government debt swap program, which eases repayment burdens but does not involve a net expansion of debt). Local government spending is the key because it makes up the vast majority (85%) of total government spending, which itself is about the same size as new private credit each year. In June, local governments took full advantage of this opportunity, issuing 316 billion RMB in brand new bonds (up from a mere 17 billion in May - an 11.8% increase year-on-year) (Table 3). This spike in issuance is later than in previous years. Combined with the Politburo and State Council pledging to boost fiscal policy and domestic demand, it suggests that net new issuance will pick up sharply in H2 2018 (Chart 14).10 Table 3Local Government Bond Issuance And Quota Chart 14Local Government Debt Can Surprise In H2 At the same time, the risk that special infrastructure spending will fall short this year is receding. About 1.4 trillion RMB of the year's new bond allowance consists of special purpose bonds to fund projects. The State Council said on July 23 it would accelerate the issuance of these bonds, since, at most, only 27% of the quota was issued in the first half of the year (Chart 15). The risk of a shortfall - due to stricter government regulations over the quality of projects - is thereby reduced. What is the overall impact of these moves? The Chinese government provides an annual "debt limit" that applies to the grand total of explicit, on-balance-sheet, local government debt. The limit increased by 11.6% for 2018, to 21 trillion RMB (Table 4), which, theoretically, enables local governments to splurge on a 4.5 trillion RMB debt blowout. Should that occur, 2.6 trillion RMB of that amount, or 3% of GDP, would be completely unexpected new government spending in 2018 (creating a positive fiscal thrust).11 Chart 15June Issuance Surged, Special Bonds To Pick Up Table 4Local Government Debt Quota Is Not A Constraint Such a blowout may not be likely, but it is legally allowed - and the political constraints on new issuance have fallen with the central government's change of stance. This means that local governments' net new bond issuance can move up toward this number. More feasibly, local governments could increase their explicit debt to 19.3 trillion RMB, a 920 billion RMB increase on what is expected, which would imply 1% of GDP in new spending or "stimulus" in 2018.12 The above only considers explicit, on-balance-sheet debt. Local governments also notoriously borrow and spend off the balance sheet. The total of such borrowing was 8.6 trillion RMB at the end of 2014, but there is no recent data and the stock and flow are completely opaque.13 The battle against systemic risk is supposed to curtail such activity this year. But the newly relaxed supervision from Beijing will result in less deleveraging at minimum, and possibly re-leveraging. Similarly, the government has said it is willing to help local governments issue refinancing bonds to deal with the spike in bonds maturing this year.14 This frees them up to actually spend or invest the money they raise from brand new bonds. In short, our constraints-based methodology suggests that the risk lies to the upside for local government debt in 2018, given that it is legal for debt to increase by as much as 2.5 trillion RMB, 3% of GDP, over the 1.9 trillion RMB increase that is already expected in the IMF's budget deficit projections for 2018. What about the central government? Its policy stance has clearly shifted. The central government could quite reasonably expand the official budget deficit beyond the 2.6% target. Indeed, that target is already outdated given that new individual tax cuts have been proposed, which would decrease revenues (add to the deficit) by, we estimate, a minimum of 0.44% of GDP over a 12-month period starting in October.15 Other fiscal boosts have also been proposed that would add an uncertain sum to this amount.16 The total of these measures can quite easily add up to 1% of GDP, albeit with the impact mostly in 2019. Finally, the strongest reason to err on the side of an upward fiscal surprise is that an expansion of fiscal policy will allow the Xi administration to boost demand without entirely relying on credit growth. First, local governments are actually flush with revenues due to strong land sales (Chart 16), which comprise around a third of their revenues. This enables them to increase spending even before they tap the larger debt allowance. Second, China's primary concern about financial risk is due to excessive corporate (and some household) leverage, particularly by state-owned enterprises (SOEs) and shadow banking. It is not due to public debt per se. It is entirely sensible that China would boost public debt as it attempts to limit leverage. In fact, this would be the Zhu Rongji playbook from 1998-2001. This was the last time that China announced a momentous three-year plan to crack down on profligate lending, hidden debts, and credit misallocation. The authorities deliberately expanded fiscal policy to compensate for the anticipate credit crunch and its drag on GDP growth (Chart 17).17 Chart 16Land Sales Enable Non-Debt Fiscal Spending Chart 17China Boosted Fiscal During Last Bad Debt Purge As for the impact on the economy, the money multiplier will be meaningful because the economy is slowing and fiscal policy has been tight. But fiscal spending does operate with a six-to-ten month lag, meaning that China/EM-linked risk assets will move long before the economic data fully shows the impact. Our sense, judging by the unenthusiastic response of copper prices thus far, is that the market does not anticipate the fiscal overshoot that we now do. Bottom Line: The political constraints on local government spending have fallen. Fiscal policy could add as much as 1%-3% of GDP to the budget deficit in H2 2018, namely if local government spending is unleashed by the recently announced policy shift. This is comparable to the 4% of GDP fiscal boost in 2008-09 and 3% in 2015-16. Can Monetary Easing Overshoot? Yes, But Less Likely. Credit is China's primary means of stimulating the economy, especially during crisis moments, and it has a much shorter lag period than fiscal spending (about three months). But Xi's agenda makes the use of rapid, credit-fueled stimulus more problematic. Based on the sharp drop in the interbank rate - in particular, the three-month interbank repo rate that BCA's Emerging Markets Strategy and China Investment Strategy use as a proxy for China's benchmark rate - it is entirely possible that credit growth will increase to some degree in H2 2018. Interbank rates have now fallen almost to 2016 levels, while the central bank never hiked the official 1-year policy rate during the recent upswing (Chart 18). In other words, the monetary setting has now almost entirely reversed the financial crackdown that began in 2017. The sharp drop in the interbank rate is partly a consequence of the three cuts to required reserve ratios (RRRs) this year, which amounts to 2.8 trillion RMB in new base money from which banks can lend.18 One or two more RRR cuts are expected in H2 2018, which could free up another roughly 800 billion-to-1.6 trillion RMB in new base money. With China accumulating forex reserves at a slower pace than in the past, and facing a future of economic rebalancing away from exports and growing trade protectionism, RRRs can continue to decline over the long run (Chart 19). China will not need to sterilize as large of inflows of foreign exchange.19 If China's banks and borrowers respond as they have almost always done, then credit growth should rise. The risk to this assumption is that the banks may be afraid to lend as long as the Xi administration remains even partially committed to its financial crackdown. Moreover, the anti-corruption campaign is continuing to probe the financial sector. While this has only produced a handful of anecdotes so far, they are significant and may have helped cause the decline in loan approvals since early 2017. Critically, China has begun the process of recognizing non-performing loans (NPLs), by requiring that "special mention loans" be reclassified as NPLs, thus implying that NPL ratios will spike, especially among small and regional lenders (Chart 20). This is part of the deleveraging process we expect to continue, but it can take on a life of its own and will almost certainly weigh on credit growth to some extent for as long as it continues. Chart 18Monetary Settings Back To Easy Levels Chart 19RRR Cuts Can Continue Chart 20NPL Recognition Underway (!) What will be the prevailing trend: monetary easing or the financial crackdown? In Chart 21 we consider three scenarios for the path of overall private credit growth (total social financing, ex-equity) for the rest of the year, with our subjective probabilities: Chart 21Three Scenarios For Private Credit In H2 2018 In Scenario A, 10% probability, we present an extreme case in which Beijing panics over the trade war and the banks engage in a 2009-style lending extravaganza. Credit skyrockets up to the 2010-17 average growth rate. This would mark a massive 11.9 trillion RMB or 13.8% of GDP increase in excess of the amount implied by the H1 2018 data. This size of credit spike would be comparable to the huge spikes that occurred during past crises, such as the 22% of GDP increase in 2008-09 or the 9% of GDP increase in 2015-16. Needless to say, this is not our baseline case, but it could materialize if the trade war causes a global panic. In Scenario B, 70% probability, we assume, more reasonably, that traditional yuan bank loans are allowed to rise toward their average 2010-17 growth rate as a result of policy easing, yet Xi maintains the crackdown on non-bank credit in accordance with this "Three Battles" framework. Credit growth would still decelerate in year-on-year terms, but only just: it would fall from 12.3% in 2017 to 11.5% in 2018. Additional policy measures could easily bump this up to a modest year-on-year acceleration, of course. This scenario would result in a credit increase worth 2.9 trillion RMB or 3.4% of GDP on top of the level implied by H1 2018. In Scenario C, 20% probability, we assume that the 2018 YTD status quo persists: bank credit and non-bank credit continue growing at the bleak H1 2018 rate. The administration's attempt to maintain the crackdown on financial risk could frighten banks out of lending. This would mean no credit increase in 2018 beyond what is naturally extrapolated from the H1 2018 data. Credit growth would slow from 12.3% to 10.7% in 2018. This scenario would be surprising, but not entirely implausible given that the Politburo is insisting on continuing the Three Battles. The collapse in interbank rates and the easing measures already undertaken - such as reports that the Macro-Prudential Assessments will lighten up, and that the People's Bank is explicitly softening banks' annual loan quotas20 - lead us to believe that Scenario B is most likely, and possibly too conservative. This is the scenario most consistent with the latest Politburo statement: that authorities will continue the campaign against systemic risk, namely through the policy of "opening the front door" (traditional bank loans go up) and "closing the back door" (shadow lending goes down), which began in January. The Chinese government has always considered control of financial intermediation to be essential. The only way to reinforce the dominance of the state-controlled banks, while preventing a sharp drop in aggregate demand, is to allow them to grow their loan books while regulators tie the hands of their shadow-bank rivals (Chart 22). Chart 22Opening The Front Door, Closing The Back One factor that could evolve beyond authorities' control is the velocity of money. Money velocity is essentially a gauge of animal spirits. If a single yuan changes hands multiple times, it will drive more economic activity, but if it is deposited away for a rainy day, then the bear spirit is in full force. Thus, if credit growth accelerates, but money in circulation changes hands more slowly, then nominal GDP can still decelerate - and vice versa.21 China's money velocity suffered a sharp drop during the tumult of 2015, recovered along with the policy stimulus in 2016, and has tapered a bit in 2018 in the face of Xi's deleveraging campaign. Yet it remains elevated relative to 2012-16 and clearly responds at least somewhat to policy easing. The implication is that money velocity should remain elevated or even pick up in H2. Again, the risk to this view is that Xi's ongoing battle against financial risk, and anti-corruption campaign in the financial sector, could suppress money velocity as well as credit growth. Bottom Line: We see a subjective 70% chance that the drop in credit growth will be halted or reversed in H2 as a result of the central bank's liquidity easing and the Politburo's willingness to let traditional bank lending grow while it discourages shadow lending. Our baseline case says the impact could amount to new credit worth 3.4% of GDP in H2 2018 that markets do not yet expect. Investment Conclusions Beijing's shift in policy suggests that our subjective probability of a policy mistake this year, leading to a sharp economic deceleration, should be reduced from 30% to 20% (Credit Scenario C above).22 Why is this dire scenario still carrying one-to-five odds? Because we fear that the financial crackdown and rising NPLs could take on a life of their own. Meanwhile the risk of aggressive re-leveraging has risen from 0% to 10% (Credit Scenario A above). Summing up, Table 5 provides a simple, back-of-the-envelope estimate of the size of both fiscal and monetary policy measures as a share of GDP. Table 5Potential Magnitude Of Easing/Stimulus Our bias is to expect a strong fiscal response combined with a weak-to-moderate credit response. This would reflect the Xi administration's desire to prevent asset bubbles while supporting growth. A more proactive fiscal policy harkens back to China's handling of its last financial purge in 1998-2001. If banks prove unable or unwilling to lend sufficiently, additional fiscal expansion will pick up the slack. New local government debt can surprise by 1% of GDP or more, while formal bank lending amidst an ongoing crackdown on shadow lending could add new credit of around 3.4% of GDP and hence mitigate or halt the slowdown in credit growth. The combined effect would be an unexpected boost to demand worth 4.4% of GDP in H2 2018, which would exert an unknown, but positive, multiplier effect. We are replacing our "Reform Reboot" checklist, which has seen every item checked off, with a new "Stimulus Checklist" that we will monitor going forward (Appendix). Chart 23How To Monitor The Stimulus Impact Neither the size of this stimulus, nor the composition of fiscal spending, will be quite as positive for EM/commodities as were past stimulus efforts. China's investment profile is changing as the reform agenda seeks to reduce industrial overcapacity and build the foundations for stronger household demand and a consumer society. Increases in fiscal spending today will involve more "soft infrastructure" than in the past. We recommend reinstituting our long China / short EM equity trade, using MSCI China ex-tech equities. We also recommend reinitiating our long China Big Five Banks / short other banks trade, to capture the disparity of the financial crackdown's impact. To capture the new upside risk for global risk assets, our colleague Mathieu Savary at BCA's Foreign Exchange Strategy has devised a "China Play" index that is highly sensitive to Chinese growth - it includes iron ore prices, Swedish industrial stocks, Brazilian stocks, and EM junk bonds (all in USD terms), as well as the Aussie dollar-Japanese yen cross. BCA Geopolitical Strategy also recommends this trade as a portfolio hedge to our negative EM view (Chart 23).23 A major risk to the "modest reflation" argument in this report will materialize if the RMB depreciates excessively in response to the escalating trade war (Trump will likely post a new tariff list on $200 billion worth of goods in September).24 This could result in renewed capital outflows breaking through China's capital controls, the PBC appearing to lose control, EM currencies and capital markets getting roiled, EM financial conditions tightening sharply, and global trade and growth slowing sharply. China would ultimately have to stimulate more (moving in the direction of Credit Scenario A above), but a market selloff would occur first and much economic damage would be done. PART II In the first part of this two-part Special Report, we concluded that policy headwinds to China's economic growth have begun to recede, but recent easing measures will likely disappoint the markets. Chart 24Money Growth Bottomed, Credit Still Weak In essence, China is girding for a trade war with the United States, which favors stimulus. But it is still attempting to reduce systemic financial risk. As a result, fiscal stimulus may surprise to the upside, but credit growth will be lackluster. The problem for investors - especially for emerging market (EM) assets and the commodity complex - is that Chinese fiscal stimulus typically operates with a six-to-ten month lag, as opposed to credit stimulus which only takes about three months to kick in.25 July statistics confirm our suspicion that credit stimulus will be hampered by the government's crackdown on shadow banking. Total credit growth remains weak, although broad money (M2) does appear to be bottoming (Chart 24). Thus far, BCA's China Investment Strategy has been correct in characterizing the latest developments as "taking the foot off the brake" rather than "pressing down on the accelerator."26 In this part of the report we take a deeper dive into the policy factors that cause us to limit our "stimulus overshoot" scenario to a 10% subjective probability. The three chief reasons are: overstated easing of macro-prudential controls; the continuing process of cleansing the banking sector of non-performing loans; and the anti-corruption campaign in the financial sector. A Preemptive Dodd-Frank Since the Xi administration redoubled its efforts to tackle systemic financial risk last year, we have urged investors to be cautious about Chinese growth.27 The creation of new institutions and new regulatory requirements set in motion processes that would be hard to reverse quickly. While these institutions are now making several compromises for the sake of stability, their operations will continue to weigh on credit growth. In July 2017, China's government held the National Financial Work Conference to address the major issues facing the country's financial system. This conference takes place once every five years and has often occasioned significant shakeups in financial regulation. In 1997, it initiated a sweeping purge of the banking system, and in 2002, it saw the creation of three financial watchdogs that would become critical institutional players throughout the 2000s.28 One of the skeletons in the closet from 2002 was the debate over whether financial regulation should be heavily centralized or divided among different, specialized, state agencies. Former Premier Wen Jiabao won the argument with the creation of the three watchdogs covering banking, securities, and insurance. After a series of controversies and conflicts, the Xi administration decided that these agencies had failed in their primary purpose of curbing systemic risk and ordered a reorganization with greater centralization. At the 2017 financial conference, Xi announced the creation of the Financial Stability and Development Committee (FSDC) to act as a centralized watchdog over the entire financial system. The FSDC would coordinate with the central bank, oversee macro-prudential regulation, and prevent systemic risk. Liu He, Xi's right-hand man on the economy and a policymaker with a hawkish reputation, was soon promoted to the Politburo and given the top job at the FSDC.29 As a second step, the Xi administration announced that it would combine the banking and insurance regulators into a single entity - the China Banking and Insurance Regulatory Commission (CBIRC). The CBIRC, to be headed by Xi ally, and notable hawk, Guo Shuqing, would continue and escalate the crackdown on shadow lending that Guo had begun at the helm of the bank watchdog in 2017 (Chart 25). The merging of the agencies would also close the regulatory gap that had seen the insurance regulator increase its dominion and rent-seeking by encouraging "excessive" financial innovation and risky pseudo-insurance products.30 Chart 25Crackdown On Informal Credit Continues The FSDC was expected, rightly, to bring a more hawkish tilt to Chinese macro-prudential regulation. In reference to the U.S.'s Financial Stability Oversight Council, we dubbed these moves a "Preemptive Dodd-Frank."31 We also argued, however, that the purpose was to bring unified command and control to financial regulation and that China would continue to prize stability above all. Therefore the degree of tightening or loosening should vary in accordance this goal.32 After a series of announcements in July and August, it is clear that China's government has shifted to a more accommodative posture (please refer back to Chart 18 and Chart 19). As usual, there are rumors of high-level political intrigue to go along with the policy shift: some argue that Premier Li Keqiang is making a comeback while Xi's golden boy, Liu He, has been sidelined due to his failure to forestall tariffs during his trade talks with Donald Trump this spring.33 Such rumors are valuable only in revealing the intensity of the policy debate in Beijing. What is certain, however, is that the FSDC, with Liu He as chairman, only met for the first time as a fully assembled group in early July, just before the major easing measures were taken. This implies that any initial conclusions were pragmatic (i.e. not excessively hawkish). Moreover, Guo Shuqing is not only the CBIRC head but also the party secretary of the PBOC, meaning that central bank chief Yi Gang cannot have adopted easing measures without Guo's at least condoning it. Chinese policymakers see the recent easing measures as "fine-tuning" even as they continue the rollout of new regulatory institutions and systems. It is thus too soon to claim that Xi Jinping or any of these government bodies have thrown in the towel on their attempts to contain excessive leverage. Both the Politburo and the State Council - the highest party and state decision-makers - have made clear that they do not intend to endorse a massive stimulus on the magnitude of 2008-09 or 2015-16.34 They have also insisted that the "Tough Battle" against systemic financial risk, and the campaign to "deleverage" the corporate sector, will continue. What does this mean in practical terms? While new regulations will be compromised, they will also continue to be implemented. For example, authorities have watered down new regulations governing the $15 trillion asset management industry, yet the regulations are still expected to go into force by 2020. These rules will weigh on shadow banking activity (e.g. wealth management products) as banks prepare to meet the requirements.35 Two other examples are critical and will be discussed below: first, the potential easing of rules under the Macro Prudential Assessment (MPA) framework for stress-testing banks; second, this year's changes to rules governing non-performing loans (NPLs). In the former case, the degree of financial easing is potentially significant but at present overestimated by investors; in the latter case, the degree of tightening is already significant and widely underestimated. Bottom Line: New financial regulatory institutions will inherently suppress credit growth, especially by dragging on informal or non-bank credit growth. Macro-Prudential Assessments: Less Easing Than Meets The Eye A key factor in determining China's credit growth going forward will be banks' responses to any softening of the Macro Prudential Assessment (MPA) requirements. News reports have suggested that a relaxation of these rules may occur, but authorities have not finalized such a move. Furthermore, the impact on credit growth may be far less than the astronomical sums being floated around the investment community. The MPA framework began in 2016. It is an evaluative system of "stress-testing" China's banks each quarter. As such it is part of the upgrade of macro-prudential systems across the world in the aftermath of the global financial crisis, comparable to the American Financial Stability Oversight Committee or the European Systemic Risk Board.36 It is managed by the PBOC and the FSDC. The MPA divides banks into systemically important financial institutions and common institutions, and subdivides the former into those of national and regional importance. The evaluation method contains seven major criteria for assessing bank stability: Capital adequacy and leverage ratios; Bank assets and liabilities; Liquidity conditions; Pricing behavior for interest rates; Quality of assets; Cross-border financing; Execution of credit policy. The first and fourth of these criteria (capital adequacy and leverage ratios, and pricing behavior for interest rates) are in bold font because they result in a "veto" over the entire assessment: if a bank fails to maintain a sufficient capital buffer, or deviates too far from policy interest rates, it can fail the entire stress-test. Otherwise, failure of any two of the other five categories results in overall failure. A system of rewards and punishments awaits banks depending on how they perform (Diagram 1). Diagram 1China's Macro Prudential Assessment Framework Explained On July 20, the PBOC published a document saying that "in order to better regulate assets of financial institutions, during Macro Prudential Assessment (MPA), relevant parameters can be reasonably adjusted." Subsequently Reuters reported that the PBOC would reduce the "structural parameter" and the "pro-cyclical contribution parameter" of the capital adequacy ratio (CAR) requirements, thereby easing rules on one of the veto items. The structural parameter would fall from 1.0 to 0.5. Rumors suggest that the pro-cyclical parameter could fall from 0.4-0.8 to 0.3. No such changes have been finalized - only a few banks actually claim to have received notification of a change and there are regional differences. Clearly a general change of the rule would reduce regulatory constraints on bank credit. But how big would the impact be? Under the MPA, banks' CARs are not allowed to fall too far below the "neutral CAR," or C*, a variable that is calculated using the formula outlined in Diagram 2. Most of the variables in this formula will not change often: for instance, the minimum legal CAR will be slow to change, as will the capital reserve buffer and the bonus buffer for systemically important institutions. The one factor that can change frequently is the "discretionary counter-cyclical buffer," as it responds to the country's current place in the business cycle. Diagram 2China's Macro-Prudential Assessment Framework: Capital Adequacy Ratios The key input to this factor is broad credit growth. Thus, if authorities should reduce the CAR's cyclical parameter from a simple average of 0.6 to 0.3, broad credit growth could go higher without creating an excessive increase in the pro-cyclical buffer. In other words, at present about 60% of bank credit expansion in excess of nominal GDP growth counts toward a counter-cyclical capital buffer, which is added to other capital buffers. A tweak to this parameter could decrease that proportion to 30%, meaning that bank lending could go twice as high with the same impact on the counter-cyclical buffer. More significantly, if authorities should reduce the CAR's structural parameter from 1.0 to 0.5, any increase in credit growth would have a less dramatic impact on C*. Hence banks would be able to lend more while still keeping their neutral CAR within the appropriate range relative to their actual CAR. Banks could theoretically lend twice as much with the same impact on the assessment.37 On paper these changes could result in unleashing as much as 41.4 trillion RMB in new lending in 2018, or 28 trillion (33% of GDP) on top of what could have been expected without any adjustment to the macro-prudential rules. This is because broad credit growth would theoretically be allowed to grow as fast as 30% instead of 17%.38 But in reality this growth rate is extremely unlikely. Why? Because it assumes that banks will grow their lending books as rapidly as they are allowed. In fact, banks are currently increasing broad credit at a rate of about 10%, which is considerably lower than either today's or tomorrow's permitted rate of growth under the MPA framework (Chart 26). Chart 26Banks Are Not Lending To The Regulatory Maximum If tweaks to the MPA increase this speed limit to 30%, it does not mean that banks will drive any faster than they are already driving. They are lending at the current pace for self-interested reasons (and there is fear of excessive debt, default, or insolvency due to the government's ongoing regulatory and anti-corruption crackdown).39 Chart 27Regulators Can Deprive Banks Of MLF Access Still, if the MPA rules are tweaked, then it will send a signal that macro-prudential scrutiny is abating and banks can lend more aggressively - this would have some positive effect on credit growth, at least for major banks that are secure in meeting their CARs. Moreover, there will be a practical consequence in that fewer banks will be punished for having insufficient CARs. At present, only rarely do banks fail the evaluations. But a strict CAR requirement during an economic downturn could change that. The proposed MPA adjustment would show that banks are graded on a sliding rule: the authorities would slide the grading scale downward to enable more banks to pass the test. This means fewer failures, which means fewer punitive measures that could upset liquidity or stability in the banking system. Ultimately, in order for the new system to have any credibility at all, punishment will have to be meted out to banks that fail the stress tests. A key punishment within the MPA system is exclusion from medium-term lending facility (MLF) loans from the PBOC. This is a regulatory action with teeth, as this is one of the PBOC's major means of injecting liquidity (Chart 27). A misbehaving bank could face short-term liquidity shortage or even insolvency. Therefore the authorities are opting to soften the rules so that the new regulatory system is preserved yet the harshest implications are avoided (for now). This would be short-term gain for long-term pain, the opposite of what China needs from the standpoint of an investor looking for improvements to productivity and potential GDP growth. But it would not necessarily be a great boon for global risk assets in the near term. While it could help stabilize expectations for China's domestic growth, it is not clear that it would unleash a mass wave of new bank loans that would reaccelerate China's economy and put wings beneath EM assets and commodity prices. Bottom Line: Tweaking the MPA parameters is a clear example of policy easing. Yet the MPA system itself is a fairly rigorous means of stress-testing banks that is part of a much larger expansion of financial sector regulation. The results of the easier rules - if implemented - will not be as reflationary as might be expected from the headline 41 trillion RMB in new loans that could legally be created. Banks are already expanding loans more slowly than they are allowed to do, so increasing the speed limit will have little effect. The real purpose of the macro-prudential tweaks is to make it more difficult for banks to fail their stress tests in a downturn. As such, any tweaks would actually reveal that Chinese policymakers are expecting a more painful downturn, not that they are asking for a credit splurge. NPL Recognition Will Weigh On Credit Growth Another factor that we have highlighted that separates today's easing measures from outright stimulus: the growing recognition of non-performing loans (NPLs) in China's banks and the financial cleansing process. The government's reform push has already led to two trends that are relatively rare and notable in the Chinese context: rising corporate defaults (Chart 28) and rising bankruptcies (Chart 29). While the impact may be small relative to China's economic size, the direction of change is significant in a country that has been extremely averse to recognizing losses. Chart 28Defaults Are Rising Chart 29Creative Destruction In China These changes reflect the tightening of financial conditions and restructurings of various industries and as such are evidence of Xi's attempt to make progress on reforms while maintaining stability. They also reflect a general environment that is conducive to the realization of bad loans. Two recent policy decisions are affecting banks' accounting of bad loans. First, the CBIRC issued new guidance that eases NPL provision requirements for "responsible" banks (banks with good credit quality) while maintaining the existing requirements for "irresponsible" banks.40 Since the major state-controlled banks will largely meet the standards, they will be able to lend somewhat more (we estimate around 600 billion RMB or 0.7% of GDP). This would support the recent trend in which traditional bank lending rises as a share of total credit growth. Second, however, the CBIRC is requiring banks to reclassify all loans that are 90-or-more-days delinquent as NPLs, resulting in upward revisions of bank NPL ratios. This will send the official rate on an upward march toward 5%, from current extremely low 1.9% (Chart 30). It is the direction of change that matters, as NPL recognition can take on a life of its own. While many state banks may already have recognized the 90-day delinquent loans, many small and regional banks probably have not. Anecdotally, a number of small banks are reporting large NPL ratios as a result of the regulatory clampdown and definition change. Rural commercial banks, in particular, are in trouble with several showing NPLs in double digits (Chart 31). These small and regional banks will have until an unspecified date in 2019 to reclassify these loans and raise provisions against them. The result will hamper credit growth. Chart 30Bad Loan Ratios Set To Rise Chart 31City And Rural Commercial Banks Most At Risk Of Rising Bad Loans To get a more detailed picture of the NPL recognition process, we have updated our survey of 16 commercial banks listed on the A-share market.41 This research reveals that banks have continued to increase the amount of bad loans they have written off. While the NPL ratio has remained roughly the same, cumulative loan-loss write-offs combined with NPLs have reached 7% of total loans and are still rising (Chart 32). This shows that a cleansing process is well underway. It is concerning that write-offs have reached nearly 50% of pre-tax profits. And even as losses mount, the proportion of each year's losses to the previous year's NPLs has fallen, implying that the previous year's NPLs had grown bigger (Chart 33). Chart 32The Bank Cleansing Process Continues Chart 33Write-Offs Almost 50% Of Bank Profits Furthermore, while loan losses grow, the surveyed banks' profit growth has been reduced to virtually zero (Chart 34). Chart 34Write-Offs Almost 50% Of Bank Profits Our updated "stress test" for Chinese banks, which is based on the same sample of 16 commercial banks, suggests that if total NPLs rise to a pessimistic, but still quite realistic, ratio of 13% (a weighted average of NPL ratio assumptions per sector, ranging from 10%-30%), then total losses could amount to 10.4 trillion RMB, or 12% of GDP (Table 6). Table 6Pessimistic Scenario Analysis##br## For Commercial Bank NPLs In this scenario, banks' net equity would be impacted by 38% as this amount surpasses the buffer of net profits (1.75 trillion RMB) and NPL provisions (3 trillion). China's banks are well provisioned, but they would be less so after a hit of this nature. A similar stress-test by BCA's Emerging Markets Strategy found that equity impairment could range from 33%-49%, implying that Chinese banks were roughly 29% overvalued on a fair price-to-book-value basis.42 Looking at different economic sectors, it is apparent that domestic trade, manufacturing, and mining have seen the highest incidence of loans going sour (Table 7). In all three cases, it is reasonable to conjecture that the NPL ratio can continue to expand - and not only because of the definitional change. First, wholesale and retail (4.7%) consists largely of SMEs, and the government is publicly concerned about their ability to get credit. Second, manufacturing (3.9%) has been hit by changing trade patterns and rising labor costs and has not yet suffered the impact from recently imposed U.S. trade tariffs. Third, mining (3.6%) has felt the first wave of the impact from the government's cuts to overcapacity in recent years, but has seen very extensive restructuring and the fallout may continue. Table 7China: Troubled Sectors Can Produce More Bad Loans More realistic NPL recognition is an important and positive development for China over the long run. Over the short run, banks' efforts to write-off NPL losses will weigh on their willingness to lend and could pose a risk to overall economic activity. Bottom Line: The government's reform and restructuring efforts are initiating a process of creative destruction in the Chinese economy. This is most notable in the government's willingness to recognize NPLs, which will continue to weigh on credit growth. The government is trying to control the pace and intensity of this process, but we expect credit stimulus to be disappointing relative to fiscal stimulus as long as the financial regulatory crackdown is at least half-heartedly implemented. Anti-Corruption Campaign Is Market-Negative Another reason to expect total credit growth to remain subdued comes from the anti-corruption campaign and its probes into local government finances and the financial sector. Chart 35Anti-Corruption Campaign Trudges Onward One of the new institutions created in China's 2017-18 leadership reshuffle was the National Supervisory Commission (NSC). This is a powerful new commission that is capable of overseeing the highest state authority (the National People's Congress). It is also ranked above the formal legal system, the Supreme Court and the public prosecutor's office. It is charged with formalizing the anti-corruption campaign and extending it from the Communist Party into the state bureaucracy, including state-owned enterprises.43 Having operated for less than a year, it is not possible to draw firm conclusions about the doings of the NSC, let alone any macro impact. Tentatively, the commission has focused on financial and economic crimes that have the potential to create a "chilling effect" among government officials and bank executives.44 Notably, the NSC has investigated Lai Xiaomin, former chief executive of Huarong, the largest of the big four Asset Management Corporations (AMCs), i.e. China's "bad banks." There is more than one reason for Huarong to attract the attention of investigators, but it is notable that it had extensive investments in areas outside its official duty of acquiring and disposing of NPLs. The implication could be that the government wants the AMCs to focus on their core competency: cleaning up the coming deluge of NPLs. The anti-corruption is also targeting local government officials for misappropriating state funds. These investigations involve punishment of provincial officials for false accounting as well as embezzlement and other crimes. We have noted before that the provinces that revised down their GDP growth targets most aggressively this year were also some of the hardest hit with anti-corruption probes into falsifying data and misallocating capital.45 On several occasions it has appeared as if the anti-corruption campaign was losing steam, but the broadest tally of cases under investigation suggest that it is still going strong despite hitting a peak at the beginning of the year (Chart 35). The campaign remains a potential source of disruption among the very officials whose risk appetite will determine whether central government policy easing actually results in additional bank lending and local government borrowing. Bottom Line: While difficult to quantify, the anti-corruption campaign will dampen animal spirits within local governments and the financial sector as long as the new NSC is seeking to establish itself and the Xi administration remains committed to prosecuting the campaign aggressively. Investment Conclusions Table 8Estimates Of Hidden Local Government Debt We would be surprised if credit growth did not perk up at least somewhat as a result of the past month's easing measures. But as outlined above, these measures may disappoint the markets as a result of the ongoing financial regulatory drive, the baggage of NPL recognition, and any negative impact on risk appetite due to the anti-corruption campaign. And this is not even to mention the dampening effects of ongoing property sector and pollution curbs.46 In lieu of a credit surge, Beijing is likely to rely more on fiscal spending to stabilize growth. Fiscal spending also faces complications, of course. In recent years, China's local governments have built up a potentially massive pool of off-balance-sheet debt due to structural factors limiting local government revenue generation (Table 8). Beijing is now attempting to force this debt into the light. The local government debt maturity schedule suggests a persistent headwind in coming years as hidden debt is brought onto the balance sheet and governments scramble to meet payment deadlines (Chart 36). In addition, the local government debt swap program launched in 2014-15 will wrap up this month. Chart 36Local Governments Face Rising Debt Payments Nevertheless Beijing has introduced a new class of "refinancing bonds" in 2018 to help stabilize the fiscal situation. These bonds are separate from brand new bonds that have the potential to increase significantly over the second half of this year. China's Finance Ministry has also reportedly asked local governments to issue 80 percent of net new special purpose bonds by the end of September. Since only about a quarter of the year's 1.35 trillion RMB quota was issued in H1, this order would mean that about half of the quota (675 billion RMB out of 1.35 trillion RMB) would be issued in August and September alone - implying a significant surge to Chinese demand, albeit with a lag of six months or so.47 The latest data releases from July suggest that Beijing is trying to do two things at once: ease liquidity conditions while cracking down on excess leverage. Until we see a spike in credit growth, we will continue to expect the policy turn to be only moderately reflationary, with the ability to offset existing headwinds but not spark a broad-based reacceleration of the economy. Going forward, data for the month of August will be very important to monitor, as many of the easing measures were not announced until late July. For all the reasons outlined in this two-part Special Report, we would view a sharp increase in total credit as a game-changer that would point toward a "stimulus overshoot" (Appendix). Such an overshoot is less likely if the government relies more heavily on fiscal spending this time around, which is what we expect. Meanwhile, turmoil in emerging markets - which we fully anticipated based on China's policy headwinds this year and our dollar bullish view - will only be exacerbated by China's unwillingness to stimulate massively.48 Matt Gertken, Associate Vice President Geopolitical Strategy mattg@bcaresearch.com Qingyun Xu, Senior Analyst qingyun@bcaresearch.com Yushu Ma, Contributing Editor yushum@bcaresearch.com Appendix Appendix Appendix 1 Please see BCA Geopolitical Strategy Weekly Report, "Political Risks Are Understated In 2018," dated April 12, 2017, available at gps.bcaresearch.com. 2 Please see BCA Geopolitical Strategy Special Report, "China: Looking Beyond The Party Congress," dated July 19, 2017, available at gps.bcaresearch.com. 3 Please see The Bank Credit Analyst Special Report, "A Long View Of China," dated December 28, 2017, available at bca.bcaresearch.com. 4 The fact that he began tightening financial policy in late 2016 and early 2017 was especially significant because only a very self-assured leader would attempt something so risky ahead of a midterm party congress. 5 Please see BCA Geopolitical Strategy Weekly Reports, "Trump, Year Two: Let The Trade War Begin," dated March 14, 2018, and "Trump's Demands On China," dated April 4, 2018, available at gps.bcaresearch.com. 6 The statement declared in its first paragraph that China would "maintain the stability of employment," with employment being the first item in a list. A similar emphasis on employment has not been seen in Politburo statements since the troubled year of 2015, and it has not been mentioned substantively in 11 key meetings since the nineteenth National Party Congress last October. 7 Please see footnote 2 above. 8 After the State Council meetings on July 23 and 26, Vice-Minister of Finance Liu Wei elaborated on the government's thinking: "These [measures] further add weight to the overall broad logic at the start of the year ... It isn't at all that the macro-economy has undergone any major volatility, and we are not undertaking any irrigation-style, shock-style measures." Please see "Beijing Sheds Light On Plans For More Active Fiscal Policy," China Banking News, July 27, 2018, available at www.chinabankingnews.com. 9 Our colleagues in BCA's Emerging Markets Strategy service have dubbed this policy "triple tightening." Please see BCA Emerging Markets Strategy Weekly Report, "EM And China: A Deleveraging Update," dated November 8, 2017, available at gps.bcaresearch.com. 10 This spike in net new issuance in the single month of June is equivalent to 19.8% of the total net new issuance in 2017. It is also much higher than the average monthly issuance in 2014-17 or in 2017 alone. However, since June and July have typically seen the largest spikes in new issuance, it will be critical to see if new issuance in 2018 remains elevated after July. Notably, local government bond issuance is currently divided between brand new bonds, debt swap bonds, and refinancing bonds, but the debt swap program will expire in August, and the refinancing bonds are separate, meaning that a larger share of the allowed new issuance will involve new spending. 11 The IMF expects the change in local government explicit debt this year to be 1.9 trillion RMB. That is, a rise from 16.5 trillion existing to 18.4 trillion estimated. 12 This number is derived by assuming that total debt reaches 92.2% of the debt limit in 2018, which is the share it reached in 2015 (since 2015 the share has fallen to 87.5% in 2017). However, 2015 was a year of fiscal easing, so it is not unreasonable to apply this ratio to 2018 as an upper estimate, now that the government's easing signal is clear. One reason that local governments have been increasing debt more slowly than allowed was that the central government was tightening investment restrictions, for instance on urban rail investment. Many new subway projects of second-tier cities have been suspended, and after raising the qualifications for subway and light rail, the majority of third- and fourth-tier cities were not qualified to build urban rail at all. As a result, local governments' investment intentions were dropping. Now this may change. 13 This estimate comes from the Ministry of Finance. The previous estimate was from the National Accounting Office and stood at 7 trillion RMB as of June 2013. 14 Maturities will spike in the coming years, so this policy signal suggests that further support for refinancing will be forthcoming. There are even unconfirmed rumors of a second phase of the local government debt swap program, which would cover "hidden debt." 15 We say "minimum" because we do not include projections of the impact of tax deductions, lacking details. We only estimate the headline savings to household incomes - loss to government revenues - based on the increase of the individual income tax eligibility threshold and the reduction in tax rates for different income brackets. 16 Additional fiscal measures include corporate tax cuts, R&D expense credits, VAT rebates, and reductions in various fees. 17 Please see BCA Geopolitical Strategy Monthly Report, "What Geopolitical Risks Keep Our Clients Awake?" dated March 9, 2016, available at gps.bcaresearch.com. 18 In fact it is more like 1.9 trillion due to strings attached, but a fourth or even fifth RRR cut could push it 3.5 trillion for the year, assuming the average 800 billion cut. 19 Ultimately this trend will result in tightening liquidity conditions in China, but for now forex reserves are not draining massively, while the RRR cuts are easing domestic liquidity. 20 Please see "China Said To Ease Bank Capital Rule To Free Up More Lending," Bloomberg, July 25, and "China's Central Bank Steps Up Effort To Boost Lending," August 1, 2018, available at www.bloomberg.com. 21 Please see BCA Emerging Markets Strategy Special Report, "Ms. Mea Challenges The EMS View," dated October 19, 2017, available at ems.bcaresearch.com. 22 Please see BCA Research Special Report, "China: Party Congress Ends ... So What?" dated November 2, 2017, available at bca.bcaresearch.com. 23 Please see BCA Foreign Exchange Strategy Weekly Report, "The Dollar And Risk Assets Are Beholden To China's Stimulus," dated August 3, 2018, available at fes.bcaresearch.com. 24 Please see BCA Global Investment Strategy Weekly Report, "Three Macro Paradoxes Are About To Come True," dated August 3, 2018, available at gis.bcaresearch.com. 25 Please see BCA China Investment Strategy Special Report, "The Data Lab: Testing The Predictability Of China's Business Cycle," dated November 30, 2017, available at cis.bcaresearch.com. 26 Please see BCA China Investment Strategy Weekly Report, "China Is Easing Up On The Brake, Not Pressing The Accelerator," dated July 26, 2018, available at cis.bcaresearch.com. 27 Please see BCA Geopolitical Strategy Weekly Report, "Political Risks Are Understated In 2018," dated April 12, 2017, available at gps.bcaresearch.com. 28 Please see BCA Geopolitical Strategy Special Report, "China: Looking Beyond The Party Congress," dated July 19, 2017, available at gps.bcaresearch.com. 29 Please see BCA Geopolitical Strategy Special Report, "Geopolitics - From Overstated To Understated Risks," dated November 22, 2017, available at gps.bcaresearch.com. 30 Please see BCA Geopolitical Strategy Special Report, "Politics Are Stimulative, Everywhere But China," dated February 28, 2018, available at gps.bcaresearch.com. 31 Please see BCA Geopolitical Strategy Special Report, "The Wrath Of Cohn," dated July 26, 2017, available at gps.bcaresearch.com. 32 Please see footnote 31 above. 33 Please see BCA Geopolitical Strategy Weekly Report, "Italy, Spain, Trade Wars... Oh My!" dated May 30, 2018, available at gps.bcaresearch.com. 34 Please see Part I of this report. 35 Please see BCA China Investment Strategy Weekly Report, "Now What?" dated June 27, 2018, available at cis.bcaresearch.com. Note that according to the new asset management rules, financial institutions will be required to have a risk reserve worth 10% of their fee income, or corresponding risk capital provisions. When the risk reserve balance reaches 1% of the product balance, no further risk provision will be required. We estimate that setting aside these funds will be a form of financial tightening worth about 1.2% of GDP. 36 Please see Liansheng Zheng, "The Macro Prudential Assessment Framework of China: Background, Evaluation and Current and Future Policy," Center for International Governance Innovation, CIGI Papers No. 164 (March 2018), available at www.cigionline.com. 37 Recall that the second category of the MPA consists of bank assets and liabilities. This category also has a rule for broad credit growth, which is that it should not exceed broad money (M2) plus 20%-25%. Therefore passing this part of the exam already requires banks to meet a 28%-33% speed limit on new credit. Assuming that that the pro-cyclical parameter of the CAR category remains at its current minimum of 0.4, then the structural parameter cannot be effectively pushed any lower than 0.6-0.8. The bottom line is that pushing the CAR structural parameter lower is not going to yield a significant increase in the allowable rate of credit growth. 38 To reach this estimate, we began with the fact that the outstanding level of broad credit growth was around 207 trillion RMB by the end of 2017 (that is, loans plus bonds plus equities plus wealth management products and other off-balance-sheet assets). The 2017 growth rate was about 10% and is assumed to be the same in 2018. Therefore broad credit should reach 227.7 trillion by the end of the year. Then, if we assume that all banks lend at the maximum weighted growth rate allowed by adjusting the structural parameter in the MPA CAR requirement (which is 30%), outstanding broad credit would reach 269.1 trillion by the end of the year. Hence an extra 41.4 trillion RMB in broad credit growth would be released. For comparison, please see CITIC Bond Investment, "Deep Analysis: Impact of Parameter Adjustments in the MPA Framework," July 30, 2018, available at www.sohu.com. 39 Based on actual CARs in 2017, the limit to broad credit growth was 17%-22% for large state-owned banks, 10%-20% for joint-equity banks, and 15%-20% for city or rural commercial banks. However, the actual broad credit growth for most banks was a lot lower than that. For example, for all five state-owned banks (nationally systemically important financial institutions), it was below 10%, well beneath the 17%-22% determined by their actual CARs and C*. 40 Under current regulations, the loan provision ratio is 2.5% while the NPL provision coverage ratio is 150%. The higher of the two is the regulatory standard for commercial banks. On February 28, 2018, the China Banking Regulatory Commission issued a notice declaring that the coverage requirement would change to a range of 120%-150%, while the loan provision requirement would change to a range of 1.5%-2.5%. Banks would qualify for the easier requirements according to how accurately they classified their loans, whether they disposed of their bad loans, and whether they maintained appropriate capital adequacy ratios. This could result in a release of about 800 billion RMB worth of provisions that can be kept as core tier-1 capital or support new lending. 41 Please see BCA China Investment Strategy Special Report, "Stress-Testing Chinese Banks," dated July 27, 2016, available at cis.bcaresearch.com. 42 Please see BCA Emerging Markets Strategy Weekly Report, "Mind The Breakdowns," dated July 5, 2018, and Special Report, "Long Indian / Short Chinese Banks," dated January 17, 2018, available at ems.bcaresearch.com. 43 Please see Jamie P. Horsley, "What's So Controversial About China's New Anti-Corruption Body?" The Diplomat, May 30, 2018, available at thediplomat.com. 44 The NSC is operationally very close to the Central Discipline Inspection Commission (CDIC), which is the Communist Party corruption watchdog formerly headed by heavyweight Wang Qishan. It received only a 10% increase in manpower over the CDIC in order to expand its target range by 200% (covering all state agencies and state-linked organizations). It has allegedly meted out 240,000 punishments in the first half of 2018, up from 210,000 during the same period last year and 163,000 in H1 2016. About 28 of these cases were provincial-level cases or higher. The controversy over the "rights of the detained" has been highlighted by the beating of a local government official's limousine driver in one of the organization's first publicly reported actions. The NSC has also arrested local government officials tied to "corruption kingpin" Zhou Yongkang and known for misappropriating budgetary funds, and has secured the repatriation of fugitives who fled abroad and recovered the assets that they stole or embezzled. 45 The provinces include Tianjin, Chongqing, Liaoning, Inner Mongolia, etc. Please see BCA Geopolitical Strategy "Trump, Year Two: Let The Trade War Begin," dated March 14, 2018, available at gps.bcaresearch.com. There is empirical evidence that anti-corruption probes are correlated with debt defaults. Please see Haoyu Gao, Hong Ru and Dragon Yongjun Tang, "Subnational Debt of China: The Politics-Finance Nexus," dated September 12, 2017, available at gcfp.mit.edu. 46 Please see BCA Emerging Markets Strategy Special Report, "China Real Estate: A Never-Bursting Bubble?" dated April 6, 2018, available at ems.bcaresearch.com, and Commodity & Energy Strategy Weekly Report, "Blue Skies Drive China's Steel Policy," dated August 9, 2018, available at ces.bcaresearch.com. 47 Please see "As economy cools, China sets deadline for local government special bond sales," Reuters, dated August 14, 2018, available at www.reuters.com. For more on local government bond issuance, see Part I of this series in footnote 1 above. Note also rumors in Chinese media suggesting that a new local government debt swap program could be launched with the responsibility of tackling off-balance-sheet debts that are guaranteed by local governments. The program has thus far only swapped debts that local governments were obligated to pay. It is not clear what would happen to a third class of local debt, that which is neither an obligation upon local governments nor guaranteed by them but that nevertheless is deemed to serve a public interest. 48 Please see BCA Geopolitical Strategy Weekly Report, "The EM Bloodbath Has Nothing To Do With Trump," dated August 14, 2018, available at gps.bcaresearch.com.

Informe especial

Highlights Globalization, technological progress, weak trade unions, high debt levels, and population aging are often cited as reasons for why inflation will remain dormant. None of these reasons are inherently deflationary, and in some contexts, they may actually turn out to be quite inflationary. The combination of a stronger dollar and rising EM stress means that U.S. Treasury yields are more likely to fall than rise during the coming months. Over the long haul, however, bond yields are going higher - potentially much higher - as inflation surprises on the upside. Long-term bond investors should maintain below-benchmark exposure to duration risk in their portfolios. Gold offers some protection against rising inflation. That said, the yellow metal is still quite expensive in real terms, which limits its appeal. Investors would be better off simply buying inflation-protected securities such as TIPS. Historically, stocks have not performed well in inflationary environments. A neutral allocation to global equities is appropriate at this juncture. Feature Will Structural Forces Limit Inflation? In Part 1 of this report, we argued that inflation could surprise materially on the upside over the coming years due to the growing conviction among policymakers that: The neutral real rate of interest is extremely low; The natural rate of unemployment has fallen significantly over time; There is an exploitable trade-off between higher inflation and lower unemployment; The presence of the zero lower-bound on nominal short-term interest rates implies that it is better to be too late than too early in tightening monetary policy. A common refrain in response to these arguments is that the structural features of today's economy are so deflationary that policymakers simply would be not able to lift inflation even if they wanted to. Four features are often cited: 1) globalization; 2) modern technologies such as automation and e-commerce; 3) the declining influence of trade unions; and 4) population aging, high debt levels, and other contributors to "secular stagnation." In this week's report, we discuss all four features in turn. In every case, we conclude that the purported deflationary forces are not nearly as strong as most observers believe. Inflation And Globalization Imagine two closed economies, identical in every way other than the fact the one economy is larger than the other. Would one expect inflation to be structurally higher in the smaller economy? Most people would probably say no. After all, if one economy has more workers and capital than another economy, it will be able to generate more output. But all those additional workers will also want to spend more, so it is not immediately obvious why inflation should differ in the two regions. Now let us change the terminology a bit. Suppose the larger economy refers to the world as a whole. What would happen to the balance between aggregate demand and supply if we were to shift from a setting where countries do not trade with one another to a globalized world where they do? As the initial example suggests, to a first approximation, the answer is nothing. Since one country's exports are another's imports, globally, net exports will always be zero. Thus, it stands to reason that simply moving from autarky to free trade will not, in itself, boost global aggregate demand. Could a move towards free trade increase aggregate supply? Yes. Global production will rise if countries can specialize in the production of goods in which they have a comparative advantage. Productivity will also benefit from the fact that a large global market will allow companies to better exploit economies of scale by spreading their fixed costs over a greater quantity of output. But here's the catch: More production also means more income, and more income means more spending. Thus, if globalization increases aggregate supply, it will also increase aggregate demand. And if both aggregate demand and aggregate supply increase by the same amount, there is no reason to think that inflation will change. Granted, it is possible that desired demand will rise more slowly than supply in response to increasing globalization, putting downward pressure on inflation and interest rates in the process. This could be the case, for example, if globalization increases the share of income going towards rich people. As Chart 1 shows, rich people tend to save more than poor people. Chart 1Savings Heavily Skewed Towards Top Earners If globalization has increased income inequality, it is possible that this has had a deflationary effect. However, for this effect to persist, the world has to become even more globalized. This does not seem to be happening. Global trade has been flat as a share of GDP for over a decade (Chart 2). The share of U.S. national income flowing to workers has also been rising in recent years as the labor market has tightened (Chart 3). Chart 2Global Trade Has Peaked Chart 3Rising Labor Share Of Income Occurring ##br##Alongside Labor Market Tightening Globalization As An Inflationary Safety Valve The discussion above suggests that the often-heard argument that globalization is deflationary because it leads to an overabundance of production is not as straightforward as it seems. What about the argument that globalization is deflationary because it limits the ability of companies to raise prices? While this is a seemingly compelling argument, it runs square into the problem that profit margins are near record-high levels in many economies. Far from making companies more price-conscious, globalization has often created oligopolistic market structures. Granted, free trade can still provide a safety valve for countries suffering from excess demand. To see this, return to our earlier example of the large country versus the small country. Suppose that because of its well-diversified economy, the large country often encounters situations where one region is booming, while another is down in the dumps. When this happens, workers and capital will tend to flow to the thriving region, alleviating any capacity pressures there. The same adjustments often occur among countries. If desired spending exceeds a country's productive capacity, it can run a trade deficit with the rest of the world. Rather than the prices of goods and services needing to rise, excess demand can be satiated with more imports. However, for that realignment in demand to occur, exchange rates must adjust. In today's context, this means that the dollar may need to strengthen further. Notice that this dynamic only works if there is slack abroad. This is presently the case, but there is no assurance that this will always be so. The implication is that inflation could rise meaningfully as global spare capacity is absorbed. Technology And Inflation If the price of electronic goods is any guide, it would seem undeniable that technological innovation is a deflationary force. However, this belief involves a fallacy of composition. Above-average productivity gains in one sector of the economy will cause prices in that sector to decline relative to other prices. But falling prices will also boost real incomes, leading to more spending. It is possible that prices elsewhere in the economy will rise by enough to offset the decline in prices in the sector experiencing above-average productivity gains, so that the overall price level remains unchanged. Ultimately, whether inflation rises or falls in response to faster productivity growth depends on what policymakers do. Over the long haul, productivity growth will lead to higher real wages. However, real wages can go up either because the price level declines or because nominal wages rise. The extent to which one or the other happens depends on the stance of monetary policy. In any case, just as in our discussion of globalization, the whole narrative about how faster productivity growth is deflationary seems rather antiquated considering that productivity growth has been quite weak in most of the world for over a decade (Chart 4). Consistent with this, the price deflator for electronic goods has been falling a lot less rapidly in recent years than it has in the past (Chart 5). Chart 4Globally, Productivity Growth Has Been ##br##Falling For Over A Decade Chart 5Steadier Prices For Computer Hardware ##br##And Software In Recent Years Admittedly, it is possible to imagine a scenario where the pace of productivity growth slows but the nature of that growth changes in a more deflationary direction. However, evidence that this has happened is fairly thin. Take the so-called Amazon effect, which purports to show sizable deflationary consequences from the spread of e-commerce. As my colleague Mark McClellan has shown, outside of department stores, profit margins in the retail sector are well above their historic average (Chart 6).1 This calls into doubt claims that online shopping has undermined corporate pricing power. Recent productivity growth in the U.S. distribution sector has actually been slower than in the 1990s, a decade which produced large productivity gains stemming from the displacement of "mom and pop" stores with "big box" retailers such as Walmart and Costco. The Waning Power Of Unions The declining influence of trade unions is also often cited as a reason for why inflation will remain subdued. There are a number of empirical and conceptual problems with this argument. Empirically, unionization rates in the U.S. peaked in the mid-1950s, more than a decade before inflation began to accelerate. While the unionization rate continued to decline in the U.S. during the 1980s and 1990s, it remained elevated in Canada. Yet, this did not prevent Canadian inflation from falling as rapidly as it did in the United States (Chart 7). The widespread use of inflation-linked wage contracts in the 1970s appears mainly to have been a consequence of rising inflation rather than the cause of it (Chart 8). Chart 6Retail Sector Profit Margins Are Strong Chart 7Inflation Fell In Canada, Despite A ##br##High Unionization Rate Chart 8Higher Inflation Led To More Inflation-Indexed ##br##Wage Contracts, Not The Other Way Around Conceptually, the argument that strong unions tend to instigate price-wage spirals is highly suspect. Yes, firms may be forced to raise wages in response to union pressures, which could prompt them to increase prices, leading to demands for even higher wages, etc. However, the price level cannot increase on a sustained basis independent of other things such as the level of the money supply. Central banks must still play a decisive role. One can imagine a scenario where the presence of powerful trade unions creates a dual labor market, one with well-paid unionized workers and another with poorly-paid non-unionized workers. Governments may be tempted to run the economy hot to prop up the wages of non-unionized workers. On the flipside, one could also imagine a scenario where the absence of strong unions exacerbates income inequality, causing governments to pursue more demand-boosting macroeconomic policies. In either case, however, the ultimate cause of rising inflation would still be macroeconomic policy. Inflation And The Neutral Rate As the discussion so far illustrates, inflation is unlikely to rise unless policymakers let it happen. But what if the neutral rate of interest is so low that policymakers lose traction over monetary policy? In that case, central banks may not be able to bring inflation up even if they wanted to. This is not just an academic question. Japan has had near-zero interest rates for over two decades and this has not been enough to spur inflation. Chart 9Long-Term Inflation Expectations In The Euro Area ##br##Are Still Much Higher Than In Japan We do not disagree with the notion that the neutral rate of interest is lower today than it was in the past. However, magnitudes are important here. In thinking about the secular stagnation thesis, which underpins the rationale for why the neutral rate has fallen, one should distinguish between the "weak" form and the "strong" form versions of the thesis. The weak form says that the neutral nominal rate of interest is low but positive, whereas the strong form says that the neutral nominal rate is negative.2 While this may seem like a minor distinction, it has important policy and market implications. Under the strong form version of the thesis, central banks really do lose control of their most effective policy tool: the ability to change interest rates to keep the economy on an even keel. By definition, if the neutral nominal rate is deeply negative, then even a policy rate of zero would mean that monetary policy is too tight. Under such circumstances, an economy could easily succumb to a vicious circle where insufficient demand causes inflation to fall, leading to higher real rates and even less spending. Such a vicious circle is less probable when the weak form version of the secular stagnation thesis dominates. As long as the neutral nominal rate is positive, central banks can always choose a policy rate that is low enough to allow the economy to grow at an above-trend pace. If they keep the policy rate below neutral for an extended period of time, the economy will eventually overheat, generating higher inflation. The fact that the U.S. unemployment rate has managed to fall during the past few years, even as the Fed has been raising rates, strongly suggests that the weak form of the secular stagnation thesis is applicable to the United States. The euro area is a much tougher call, given the region's poor demographics and high debt levels. Nevertheless, at least so far, the euro area has one thing on its side: Long-term inflation expectations are still much higher than they are in Japan (Chart 9). Whereas a neutral real rate of zero implies a nominal rate of 1.8% in the euro area, it implies a much lower nominal rate of 0.5% in Japan. The Neutral Rate Will Likely Move Higher As we argued a few weeks ago, cyclically, the neutral real rate of interest has risen in the U.S., and to a lesser extent, the rest of the world.3 This has happened because deleveraging headwinds have abated, fiscal policy has turned more stimulative, asset values have risen, and faster wage growth has put more money into workers' pockets. Structurally, the neutral rate may also begin to creep higher as some of the very same long-term forces that have depressed the neutral rate in the past begin to push it up in the future. Demographics is a good example. For several decades, slower population growth has reduced the incentive for firms to expand capacity. Diminished investment spending has suppressed aggregate demand, leading to lower inflation. Population aging also pushed more people into their prime saving years - ages 30 to 50. By definition, more savings mean less spending. However, now that baby boomers are starting to retire en masse, they are moving from being savers to dissavers. Chart 10 shows that the "world support ratio" - effectively, the ratio of workers-to-consumers - has begun to fall for the first time in 40 years. As more people stop working, aggregate global savings will decline. The shortage of savings will put upward pressure on the neutral rate. Japan has been on the leading edge of this demographic transformation. The unemployment rate has fallen to a mere 2.4%, while the ratio of job openings-to-applicants has reached a 45-year high (Chart 11). The shackles that have kept Japan immersed in deflation for over two decades may be starting to break. Chart 10The Ratio Of Workers-To-Consumers Is Now Falling Chart 11Japan: Labor Market Tightening May Spur Inflation Debt Deflation Or Debt Inflation? The distinction between the weak form of secular stagnation and the strong form is critical for thinking about debt issues. Rising debt tends to boost spending, but when debt reaches very high levels, spending normally suffers as borrowers concentrate on paying back loans. As such, high indebtedness generally implies a lower neutral real rate of interest. There is an important caveat, however. The presence of a lot of debt in the financial system also creates an incentive for policymakers to boost inflation in order to erode the real value of that debt. This is particularly the case when governments are the main borrowers. When the strong form version of secular stagnation prevails, generating inflation is difficult, if not impossible. In such a setting, debt deflation becomes the main concern. In contrast, when the weak form version of secular stagnation prevails, higher inflation is achievable. Debt inflation becomes an increasingly likely outcome. If we are in a period where countries such as Japan are transitioning from a strong form of secular stagnation to a weak form, inflation could begin to move rapidly higher. We are positioned for this by being short 20-year versus 5-years JGBs. Inflation As A Political Choice There is a school of thought that argues that high inflation in the 1970s and early 80s was an aberration; that the natural state of capitalism is deflation rather than inflation. We reject this view. The natural state of capitalism is ever-increasing output. Whether prices happen to rise or fall along the way depends on the choice of monetary regime. This is a political decision, not an economic one. Regimes based on the gold standard tend to have a deflationary bias, whereas regimes based on fiat money tend to have an inflationary one. The introduction of universal suffrage in the first few decades of the twentieth century made inflation politically more palatable than deflation (Chart 12). There is little mystery as to why that was the case. In every society, wealth is unevenly distributed. Creditors tend to be rich while debtors tend to be poor. Unexpected inflation hurts the former, but benefits the latter. Chart 12Universal Suffrage Made Inflation Politically ##br##More Palatable Than Deflation Once universal suffrage was introduced, a poor farmer did not need to worry quite as much about losing his land to the bank, since he could now vote for someone who would ensure that crop prices increased rather than decreased. In William Jennings Bryan's colorful words, the rich and powerful "shall no longer crucify mankind on a cross of gold." Today, populism is on the rise. Trumpist Republicans have clobbered mainstream Republicans in one primary election after another. The democrats are also shifting to the left, as the ousting of ten-term incumbent Joe Crowley by the firebrand socialist candidate Alexandria Ocasio-Cortez in June illustrates. And the U.S. is not alone. Italy now has an avowedly populist government. Other European nations may not be far behind. Meanwhile, a growing chorus of prominent economists have argued in favor of raising inflation targets on the grounds that a higher level of inflation would allow central banks to push real interest rates deeper into negative territory in the event of a severe economic downturn. We doubt that any central bank would proactively raise its inflation target in the current environment. However, one could imagine a situation where inflation begins to gallop higher because central banks find themselves behind the curve in normalizing monetary policy. Confronted with the choice between engineering a painful recession and letting inflation stay elevated, it would not be too surprising in the current political context if some central banks chose the latter option. Investment Conclusions As we discussed last week, the combination of a stronger dollar and rising EM stress means that U.S. Treasury yields are more likely to fall than rise during the coming months.4 Over the long haul, however, bond yields are going higher - potentially much higher - as inflation surprises on the upside. Long-term bond investors should maintain below-benchmark exposure to duration risk in their portfolios. Gold offers some protection against inflation risk. However, the yellow metal is still quite expensive in real terms, which limits its appeal (Chart 13). Investors would be better off simply buying inflation-protected securities such as TIPS. Chart 13Gold Is Not Cheap Historically, equities have not performed well in inflationary environments. U.S. stocks are quite expensive these days (Chart 14). Analyst expectations are also far too rosy (Chart 15). Non-U.S. stocks are more attractively priced, but face a slew of near-term headwinds. A neutral allocation to global equities is appropriate at this juncture. Chart 14U.S. Stocks Are Expensive Chart 15Analysts Are Far Too Optimistic Peter Berezin, Chief Global Strategist Global Investment Strategy peterb@bcaresearch.com 1 Please see Global Investment Strategy Special Report, "Did Amazon Kill The Phillips Curve?" dated September 1, 2017. 2 To keep things simple, we are assuming that nominal interest rates cannot be negative. In practice, as we have seen over the past few years, the zero lower-bound constraint is rather fuzzy. Nevertheless, it is doubtful that interest rates can fall too far into negative territory before people begin to shift negative-yielding bank deposits into physical currency. 3 Please see Global Investment Strategy Weekly Report, "U.S. Housing Will Drive The Global Business Cycle... Again," dated July 6, 2018. 4 Please see Global Investment Strategy Weekly Report, "Hot Dollar, Cold Turkey," dated August 17, 2018. Strategy & Market Trends Tactical Trades Strategic Recommendations Closed Trades

Highlights Lesson 1: Inflation is a non-linear phenomenon. Lesson 2: Beware government interference in monetary policy. Lesson 3: An emerging markets shock is deflationary for developed markets. Lesson 4: The 'Rule of 4' for equities and bonds. Feature We took a much needed holiday last week, hoping that financial markets would enter a midsummer slumber. Our hopes were dashed. The timing of the Turkish lira crisis reminded us of the old adage: time, tide - and financial markets - wait for no man. But on reflection, our summer holiday gave us the time for some, well... reflection: a precious quality in a world that is rapidly neglecting the value of reasoned analysis. The addiction to minute-by-minute commentary and knee-jerk reaction - epitomised by the Twitterati - means that we are 'thinking fast', when we should be 'thinking slow'. So here, after some reflection, are four long-term lessons from the Turkish lira crisis. Lesson 1: Inflation Is A Non-Linear Phenomenon. Turkey's recent experience clearly demonstrates that inflation is non-linear - meaning that inflation doesn't move in a gradual or controlled fashion. Non-linear phenomena experience sudden and explosive phase-shifts (Chart I-2). In Turkey's case, a major cause of its currency crisis was that inflation recently phase-shifted out of a well-established channel to its current 16 percent rate (Chart of the Week). Chart of the WeekTurkish Inflation Experienced A Non-Linearity Chart I-2Inflation Can Experience A Phase-Shift People struggle with the concept of non-linearity because the vast majority of our day to day experiences are linear, meaning the output is proportionate to the input. The speed of our car depends linearly on the pressure on the accelerator pedal; the temperature in our home depends linearly on the thermostat setting; the volume of music in our headphones depends linearly on the volume setting; and so on. Likewise, the vast majority of economic models - including the infamous DSGE inflation models used by central banks - assume linear relationships.1 But some phenomena are non-linear. An example you might relate to is trying to get a small amount of tomato ketchup out of crusted-over squeezy bottle. It is impossible. You squeeze and no ketchup comes out; you squeeze harder and still nothing comes out; and then suddenly you get the explosive phase-shift: the entire bottle empties on your plate! Inflation also experiences violent phase-shifts. The main reason is that people cannot perceive small changes in inflation, making inflation expectations very sticky, which is to say non-linear. The Turkish people might not perceive inflation rising from 8 percent to 10 percent, but they would certainly perceive it rising to 16 percent. Hence, as policymakers squeeze the ketchup bottle, nothing happens at first. But at a tipping point, the self-reinforcement of inflation expectations becomes explosive. Whereupon, the whole bottle comes out. The broad money supply, M, gaps up because it becomes rational for banks to lend as much as possible. And its velocity, V, also gaps up because it becomes rational to spend the money - both newly created and pre-existing balances - as quickly as possible (Chart I-3-Chart I-6). So the product MV, which equals nominal GDP, experiences an even sharper non-linearity. Chart I-3The Velocity Of Money... Chart I-4...Is A Non-Linear Phenomenon Chart I-5The Money Multiplier... Chart I-6...Is A Non-Linear Phenomenon This begs the question: when should we worry about a sudden phase-shift in developed market inflation rates? The answer comes from Lesson 2. Lesson 2: Beware Government Interference In Monetary Policy. An economy's broad money supply, M, is dominated by loans. So to expand the broad money supply, somebody has to borrow money. This means that the danger of an inflation phase-shift rises sharply if the government can borrow and spend money at will, with the central bank creating it.2 Over the past few centuries, the British government - by periodically leaving the gold standard - did exactly this to pay for the Napoleonic Wars, the Crimean War and the First World War (Chart I-7). Chart I-7The British Government Created Inflation To Pay For Wars Which answers the question of when to worry. The government has to get into cahoots with the central bank. In other words, the central bank loses its independence and fiscal policy has the scope to become ultra-loose. This describes the situation in Turkey, where President Erdogan has forced the central bank to suppress interest rates, while putting his son-in-law in charge of the Turkish treasury. Could something similar happen in developed economies? President Trump's fiscal stimulus combined with his recent attempt to influence Federal Reserve policy (to more dovish) is a small step in this direction. Nevertheless, the major developed market central banks are on a hawkish path. They are squeezing less on the ketchup bottle. Therefore, the real risk of a phase-shift in developed market inflation will arise not before the next global downturn, but after it - when desperate policymakers might resort to desperate measures. In the near term, we expect developed market inflation to remain contained, and one supporting reason comes from Lesson 3. Lesson 3: An Emerging Markets Shock Is Deflationary For Developed Markets. The slowdown and recent shock in emerging markets has caused the dollar and yen to surge. Even the euro - on a broad trade-weighted basis - has held up very well through the Turkish lira crisis and is up 2 percent in 2018 (Chart I-8). Chart I-8An EM Shock Boosts DM Currencies... Meanwhile, since May, industrial metal prices have plunged 20 percent (Chart I-9) and even the crude oil price is down by 10 percent. Chart I-9...And Depresses Industrial Commodity Prices An emerging market shock also threatens the developed market banking system by impairing its foreign loans. Thereby, it risks stifling domestic credit creation. The combination of stronger currencies, lower commodity prices, and potentially weaker bank credit creation is a disinflationary headwind for developed markets in the near term. Lesson 4: The 'Rule of 4' For Equities And Bonds. If developed market inflation remains contained in the near term, it should also keep a lid on bond yields. This is significant because our non-consensus call is that the main threat to developed market risk-assets comes not from trade wars and/or a global economic slowdown; it comes from rich valuations which will become dangerously unstable if bond yields march much higher. The bond yield that matters is the global long bond yield. Effectively, this is the weighted average of its three main components: the 10-year yields on the U.S. T-bond, the German bund and the Japanese government bond (JGB). But for a useful rule of thumb, just sum the three yields. A sum above 4 - which broadly equates to the global 10-year yield rising above 2 percent - means it is time to go underweight equities. A sum between 3.5 and 4 means a neutral stance to equities. A sum well below 3.5 means an overweight stance to equities - because it would justify even richer valuations. Investment Recommendations Asset allocation: Our 'rule of 4' sum now stands at 3.3, indicating a close to neutral stance to equities. For bonds, we have since May recommended an overweight position in a portfolio of high-quality government 30-year bonds. The recommendation is performing well, and it is appropriate to stick with it for the time being. Sector allocation: Stay overweight the classical defensives versus the classical cyclicals: materials, industrials and banks. This recommendation has fared spectacularly well. Healthcare has outperformed banks by 20 percent since February, so the pressing question is: when to take profits? We anticipate at some point in the fourth quarter. Within the cyclical sectors, prefer banks over oil and gas. Regional and country equity allocation: the geographical allocation of equities follows directly from the sector allocation. Our preferred ranking of sectors necessarily means that our preferred ranking of major equity markets is: S&P500 first, Eurostoxx50 and Nikkei225 second (tied), FTSE100 third. Again, this recommendation has performed extremely well. Currency allocation: Since February, our main currency recommendations have been short EUR/JPY, long EUR/USD, and long EUR/CNY. In effect the recommendations reduce to: long JPY/USD and long EUR/CNY, and this combination has proved to be an excellent 'all-weather' position (Chart I-10). Stick with it for the time being. Chart I-10Long JPY/USD And EUR/CNY Has Been##br## A Good 'All-Weather Combination' Finally, our long-standing short Turkish lira versus South African rand position has returned a mouth-watering 73 percent in four years.3 It is time to close the short Turkish lira position and bank the profits. Dhaval Joshi, Senior Vice President Chief European Investment Strategist dhaval@bcaresearch.com 1 Dynamic Stochastic General Equilibrium models. 2 For example, by giving all public sector workers a 50% pay rise! 3 After the cost of carry, based on interest rate differentials. Fractal Trading Model* Market reaction to the Turkish lira crisis caused our two most recent trades to hit their stop-losses, but it has also created new opportunities. The aggressive sell-off in industrial commodities appears technically extended. So this week's recommended trade is an intra-cyclical equity sector pair-trade: long global basic resources, short global chemicals. The profit target is 3.5% with a symmetric stop-loss. For any investment, excessive trend following and groupthink can reach a natural point of instability, at which point the established trend is highly likely to break down with or without an external catalyst. An early warning sign is the investment's fractal dimension approaching its natural lower bound. Encouragingly, this trigger has consistently identified countertrend moves of various magnitudes across all asset classes. Chart I-11 The post-June 9, 2016 fractal trading model rules are: When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. Use the position size multiple to control risk. The position size will be smaller for more risky positions. * For more details please see the European Investment Strategy Special Report "Fractals, Liquidity & A Trading Model," dated December 11, 2014, available at eis.bcaresearch.com Fractal Trading Model Recommendations Equities Bond & Interest Rates Currency & Other Positions Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields Interest Rate Chart II-5Indicators To Watch - Interest Rate Expectations Chart II-6Indicators To Watch - Interest Rate Expectations Chart II-7Indicators To Watch - Interest Rate Expectations Chart II-8Indicators To Watch - Interest Rate Expectations

Highlights Just to be clear: The balance of price risks in oil markets remains to the upside - particularly if we see a supply shock resulting from the loss of as much as 2mm b/d of exports from Iran and Venezuela. Neither the supply side nor the demand side in base metals evidence outsized risks, which keeps us neutral ... for now. Still, downside risks for commodities - mostly via threats to trade - loom. In line with our House view, we believe markets are too complacent re the effects of a global trade war.1 However, focusing only on the trade war obscures growing risks to EM imports and exports arising from the Fed's rates-normalization policy, which is pushing the USD higher. A strong USD retards EM trade growth, which is particularly bearish for metals and oil (Chart of the Week). Chart of the WeekStronger USD, Slower EM Import Growth##BR##Bearish For Base Metals And Oil An oil-supply shock taking prices above $120/bbl, as one of our scenarios does, would generate a short-term inflationary impulse, and would depress aggregate demand, particularly in EM. Ultimately, it would become a deflationary impulse, as higher energy prices consume a larger share of discretionary incomes, and slow growth. A slowdown in EM trade on the back of a strong USD also would generate a deflationary impulse, as EM income growth slows and aggregate demand falls. Either way, the Fed's rates-normalization policy will be put on hold as current inflation risks morph to deflation risks, if the downside becomes dominant. Highlights Energy: Overweight. The U.S. Strategic Petroleum Reserve (SPR) will release 11mm of oil from its reserves in the October - November period, to allay concerns over the likely loss of 1mm b/d of Iranian exports to U.S. sanctions. We've been expecting this ahead of U.S. mid-term elections, but don't think it will fill the gap in lost exports. Base Metals: Neutral. Union and management leaders at BHP's Escondida mine in Chile averted a strike, after agreeing a contract at the end of last week. Precious Metals: Neutral. Gold rallied more than $35/oz off its lows of last week, as markets took notice of record speculative short positioning, which many view as a bullish contrary indicator. Gold was trading to $1195/oz as we went to press. Ags/Softs: Underweight. The USDA is expected to roll out a $12 billion relief package for farmers on Friday, which includes direct purchases of commodities that were not exported due to tariffs, according to agriculture.com's Successful Farming publication. Feature Overall, the balance of price risks in the industrial commodities are neutral (in base metals) and to the upside (in oil). In the base metals, we think fear of a Sino - U.S. trade war has market participants jittery, and may be getting to the point where it is starting to affect expectations for capex and investment on the production side, and growth on the demand side. Given our expectation EM trade will hold up this year (Chart 2), we continue to expect base metals demand to remain fairly stable, and perhaps pick up as China rolls out modest stimulus measures later this year.2 Chart 2USD Strength Slows EM Trade Growth We remain bullish oil demand - expecting growth of ~ 1.6mm b/d on average in 2018 - 19, and continue to expect a supply deficit next year, which will push Brent prices from $70/bbl on average in 2H18 to $80/bbl next year.3 However, if we see continued strength in the USD beginning to degrade actual EM demand, we will be forced to revise our assessment. Downside Risks To Metals And Oil Loom As mentioned above, we are aligned with our House view, and believe markets are all but ignoring the risk of an all-out trade war, spreading from the well-covered Sino - U.S. standoff to the broader global economy. The global economy already appears to be registering the first signs of a trade slowdown, according to the World Bank's July 2018 global outlook, where it observes "softening demand for imports in advanced economies - with the exception of the United States - and weaker exports from Asia."4 We also are picking it up in our modeling (Chart 2). The Bank also notes the slowdown in trade "is accompanied by rising barriers to trade, moderating growth in China, higher energy prices, and elevated policy uncertainty." A prolonged trade war that spreads globally would be especially devastating to EM economies, as two-thirds of them are commodity exporters of one sort or another.5 Fed Policy Is An EM Growth Risk As important as a trade war is for global growth, focusing too heavily on it obscures growing risks to EM imports and exports arising from the Fed's rates-normalization policy, which is pushing the USD higher. Table 1USD Vs. Fed Policy Variables Per the Richmond Fed's Summary, the Fed is charged by Congress to "promote effectively the goals of maximum employment, stable prices, and moderate long term interest rates."6 One of the models we use to forecast the broad trade-weighted USD is a Fed policy-variables model, which uses lagged U.S. nonfarm payrolls, core PCEPI (the Fed's preferred measure), U.S. 10-year real rates, and U.S. short-term real-rate differentials vs. DM rates as proxies for these policy goals. We throw lagged copper futures prices in to pick up current industrial activity, as well (Table 1). This model highlights the long-term equilibrium between the USD TWIB and the Fed's policy variables going back to 2000.7 We average the output of the policy-variables model with four other models using close-to-real-time variables, and some other proxies for the Fed's policy variables to generate our forecast (Chart 3). Chart 3BCA USD TWIB Forecast The USD TWIB and EM trade volumes form a cointegrated system, as shown in Chart 2. Based on our modeling, we expect EM trade to hold up reasonably well over the next year, with y/y growth remaining positive most of the time. But, as close inspection of the chart reveals, the rate of p.a. growth is slowing as a result of the Fed's rates-normalization policy. This means the rate of growth in EM demand for base metals and oil will slow, although the level of demand will remain high following 20 years of solid growth.8 As a House, we expect the USD TWIB to rise another 5% over the next year, which, given the elasticities in our model, would translate into more than 10% declines in copper and Brent prices, all else equal. The Oil Wildcard As regular readers of this service know, we do not believe "all else equal" applies to commodity markets, particularly oil. We have been highlighting the risks of a confluence of negative supply shocks for months - i.e., the loss of up to 2mm b/d of oil exports from Iran and Venezuela - and the implications of this for prices (Chart 4). This is apparent in our ensemble forecasts, which reflect the physical deficit we expect to the end of 2019 (Chart 5). Chart 4U.S. SPR Release Doesn't Cover Lost Iranian Exports The U.S. government has taken notice of these risks. However, we believe this week's announcement by the Trump administration to release 11mm barrels of crude oil from the U.S. SPR over the October - November period might hold gasoline prices down ahead of the U.S. midterms, but will do next to nothing to make up for the lost export volumes we are expecting in 2019 (Chart 4). Chart 5BCA Continues To Expect Physical Deficits An oil-supply shock taking prices above $120/bbl - the projection from one of our scenarios in Chart 4 - would generate a short-term inflationary impulse in U.S. data the Fed follows. This would depress aggregate demand, particularly in EM, as oil is priced in USD. The Fed likely looks through this spike, but, should it misread the inflation impulse and tighten more aggressively, it would be delivering a double-whammy to EM economies: Higher oil prices and a stronger USD. Many EM governments have relaxed or removed subsidies on fuel prices following the 2015 collapse in oil prices engineered by OPEC. While some governments may re-introduce subsidies, not all will cover all of the price increase in such a shock.9 So, even if some subsidies are re-introduced, a price spike likely would hit EM consumers harder than previous high-price epochs. There is a non-trivial likelihood such an oil-price spike would trigger a recession in the U.S. - and likely in DM and EM economies - per Hamilton's (2011) analysis.10 This would force the Fed to change course and resume its accommodative policies. Ultimately, this would become a global deflationary impulse, as higher energy prices erode discretionary incomes, and slow growth. Bottom Line: An oil-supply shock and slower EM trade growth on the back of a strong USD ultimately produce deflationary impulses. Either way, Fed rates-normalization policy will be put on hold if these downside risks become the dominant theme in industrial commodity markets, and the current inflation risks morph to deflation risks. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Hugo Bélanger, Senior Analyst Commodity & Energy Strategy HugoB@bcaresearch.com 1 Please see BCA Research's Global Investment Strategy Weekly Report "How To Trade A Trade War," published July 13, 2018. It is available at gis.bcaresearch.com. 2 BCA Research's Geopolitical Strategy is expecting policymakers to deploy modest fiscal stimulus and reflationary policies to counter growing threats from the country's trade war with the U.S. This will be supportive, at the margin, for bulks and base metals. Please see "China: How Stimulating Is The Stimulus?" published by our Geopolitical Strategy August 8, 2018. It is available at gps.bcaresearch.com. 3 Please see BCA Research's Commodity & Energy Strategy Weekly Report "OPEC 2.0 Sailing Close To The Wind," which contains our most recent supply-demand balances and forecasts. It was published August 16, 2018, and is available at ces.bcaresearch.com. 4 Please see The World Bank's Global Monthly, July 2018, p. 2. 5 Please see remarks by World Bank Senior Director for Development Economics, Shantayanan Devarajan, who notes, "two-thirds of developing countries ... depend on commodity exports for revenues." His remarks are in "Global Economy to Expand by 3.1 percent in 2018, Slower Growth Seen Ahead," World Bank press release on June 5, 2018. 6 Please see Steelman, Aaron (2011), "The Federal Reserve's "Dual Mandate": The Evolution Of An Idea," published on the Federal Reserve Bank of Richmond's website. 7 We use a cointegration model to estimate these policy-driven regressions. The output is stout (R2 is greater than 0.95), and it has good out-of-sample results. We use a weighted-average of the five forecasts based on root-mean-square-errors to come up with our USD_TWIB forecast. 8 The World Bank estimates the seven largest EM economies - Brazil, China, India, Indonesia, Mexico, the Russian Federation, and Turkey - accounted for ~ 100% of the increase in metals consumption and close to 70% of the increase in energy demand over the past 20 years. Please see "The Role of Major Emerging Markets In Global Commodity Demand," in the Bank's June 2018 Global Economics Prospects, beginning on p. 61. 9 Please see BCA's Commodity & Energy Strategy Weekly Report, "OPEC 2.0 Scrambles To Reassure Markets," published June 28, 2018. It is available at ces.bcaresearch.com. 10 For an excellent discussion of the correlation between oil-price shocks and recessions, please see Hamilton, James D. (2011), "Historical Oil Shocks," Prepared for the Handbook of Major Events in Economic History. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed in 2018 Summary of Trades Closed in 2017

Highlights The persistent weakness of the RMB appears to be one important factor weighing on Chinese stocks, particularly the domestic market. CNYUSD may have some upside from current levels if the Trump administration applies only 10% rate to the second round of planned tariffs, but on balance is likely to come under further market pressure. This explains the PBOC's decision to try to support the currency. Interestingly, July brought some hopeful (albeit early) macro signals from China among the data that we track, some of which appear to have been overlooked by investors. Still, a neutral stance towards Chinese investable stocks versus the global benchmark continues to be warranted, at least until some clarity emerges about the magnitude and disposition of the export shock. Feature Economic and financial market conditions in China have not meaningfully improved since the publication of our last weekly report. Chart 1 highlights that China's economic surprise index remains in negative territory, and Chart 2 shows that Chinese investable and domestic stocks remain 22% and 29%, respectively, below their rolling 1-year high in local currency terms. In US$ terms, domestic Chinese stocks are 34% below their January peak, owing to the significant decline in CNYUSD. The BCA China Play Index and the relative performance of domestic infrastructure stocks versus global equities are two additional market indicators that we are watching closely as proxies for reflation, and neither is signaling a significant improvement (Chart 3). Chart 1Persistently Negative Economic Surprises... Chart 2...And Still In A Bear Market Chart 3Reflation Proxies Are Not Signaling A Major Economic Upturn The RMB Factor The persistent weakness of the RMB appears to be one important factor weighing on Chinese stocks, particularly the domestic market. While a weaker currency will actually help offset some of the export shock, Chart 4 shows that domestic stocks have not responded positively to the decline: the rolling 3-month correlation between the two has soared even further into positive territory over the past month, which may explain recent actions from the PBOC to help stabilize the currency. In short, the RMB appears to be acting as the "panic barometer" for domestic equity investors. Chart 4The RMB Is Acting As A "Panic Barometer" For Domestic Stocks Chart 5Some Evidence Of PBOC-Driven Depreciation The PBOC continues to maintain that it is not actively manipulating the RMB, arguing that both last year's appreciation and Q2's depreciation have occurred due to market supply and demand. Chart 5 casts some doubt on this claim, suggesting that at least some of the recent decline has been purposeful. The chart shows the standardized 1-month percent change in official reserves, measured in SDRs to help remove the impact of currency fluctuations. It highlights that the change in currency-neutral reserves has been quite elevated over the past three months relative to recent history, which is what would be expected (absent major capital outflow) if the PBOC was buying foreign currency assets to push down the exchange rate. But we agree that the extent of the decline is now probably more than what policymakers are comfortable with, which raises the question of how much more market-based pressure the RMB is likely to come under. In attempting to answer this question, it is interesting to note that the magnitude of the decline in CNYUSD over the past two months seems to have been closely aligned with the share of proposed tariffs as a share of Chinese exports to the U.S., as would be implied in a simple open economy model with flexible exchange rates. Chart 6 illustrates the magnitude of the decline in CNYUSD that would be implied by this framework in a variety of tariff scenarios. The chart shows that the RMB has some upside from current levels if the rate on the second round of tariffs is limited to 10% (instead of the 25% that has been threatened), and no additional tariffs are levied. But it also shows that further market pressure on the exchange rate is likely if the Trump administration simply follows through with their stated plans, and especially if the U.S. moves to tariff all imports from China. Notably, in the scenarios showing a further RMB decline, all of them fall below the psychologically important level of 7 yuan to the dollar. Chart 6More Pressure On RMB To Come If Trump Merely Follows Through With His Threats Given this, it is easy to see why investors feel that they are in limbo regarding the outlook for Chinese stock prices. They can observe the reflationary outlook of Chinese policymakers, but they are also factoring in: A looming export shock of still uncertain magnitude A strong signal from authorities that the campaign to control leverage and crackdown on shadow banking will not be abandoned Persistent RMB volatility An ongoing "old economy" slowdown that was already underway prior to the imposition of tariffs Domestic Economy Crosscurrents Chart 7Closely Watched Data Releases Negatively Surprised In July Concerning the last of these factors, we have written about a slowdown in China's old economy for the better part of the past year, a view that is now sharply in the market's focus given the negative external outlook. Last week's disappointing release of the July retail sales, industrial production, and fixed asset investment data certainly did not help improve investor sentiment towards China's economy (Chart 7). Interestingly, however, July did bring some hopeful (albeit early) macro signals from China, some of which appear to have been overlooked by investors. Table 1 presents the dashboard of select macro series that we have showed in several reports over the past few months. It highlights the evolution of the key six components of our BCA Li Keqiang index Leading Indicator, four housing market series that we have found to have strongly leading properties, as well as the NBS and Caixin manufacturing PMIs. Credit growth and the PMIs are currently providing the most negative signals, in that they declined in July and are below their 12-month moving average. In the case of credit growth, this is a continuation of an almost 2-year downtrend, but the PMI weakness has been much more recent (in response to the worsening export outlook). But several indicators that we track ticked up in July, including 4 out of 6 components of our leading indicator for the Li Keqiang index (LKI). The fact that monetary conditions indexes have risen should not be surprising given the recent weakness in the currency, but growth in the money supply also ticked up non-trivially last month (possibly due to the PBOC's apparent manipulation of the RMB). In the case of M2, the tick up technically pushed the YoY growth rate (modestly) above its trend for the first time in 2 years. Table 1Some Hopeful Signs, But Credit Remains Weak There are two other points from Table 1 worth highlighting, the first of which is negative. While the LKI itself has looked reasonably strong over the past few months (in contrast to our slowing domestic demand view), it ticked down in July for the second time. In addition, the LKI has recently been propped up by two, presumably unsustainable, factors: a spurt of rail cargo volume growth that appears to be strongly linked to trade front-running in advance of the U.S. import tariffs, and a surge in electricity consumption from the services industry (which is not investment-intensive). Chart 8 controls for the second factor by presenting an alternative measure of the LKI that replaces overall electricity production with consumption in primary and secondary industries; the difference in the recent trend between the two measures is clear. Chart 8The LKI Is Being Held Up By Trade Front Running And Services The second important point from Table 1 is positive: both housing starts and sales accelerated very significantly in July, with sales being particularly notable. BCA's China Investment Strategy service has highlighted that the housing sector represented the best candidate for meaningful acceleration in Chinese economic activity, and the July data was particularly impressive. It remains unclear whether the authorities will continue to follow through with a crackdown on the property sector, despite recent statements suggesting they will: household leverage is not enormously elevated relative to GDP, but it has accelerated very significantly over the past couple of years. But if the recent strength in sales volume continues and policymakers do not respond aggressively with macroprudential measures, our conviction in a sustained residential construction boom in China would rise materially. This will be important for investors to monitor, as it could provide a critical source of investment-driven domestic demand over the coming 6-12 months. Investment Conclusions Despite the crosscurrents buffeting China's economic outlook, we can draw three conclusions that lead us to firm near-term investment strategy recommendations: Market proxies are not signaling that Chinese policymakers will end up overstimulating the economy For now, credit growth, and the domestic "old economy" more generally, continues to decelerate Further RMB weakness may be in the cards To us, these conclusions clearly argue for a neutral stance towards Chinese investable stocks versus the global benchmark, at least until some clarity emerges about the magnitude and disposition of the export shock. We also continue to recommend that investors favor low market beta sectors within the investable universe, such as classical defensives as well as industrials.1 In early-July, we opened a "shadow" trade of being long the MSCI China A Onshore index / short MSCI China index, which we said we would consider implementing in response to a 5% rally in relative dollar performance. Chart 9 highlights that this threshold has not yet been reached, and we continue to warn against trying to catch a falling knife. But Chart 10 underscores how stretched (to the downside) domestic stocks have become: versus the global benchmark, relative stock prices in US$ have fallen to an 11-year low. Panel 2 illustrates that this stretched performance is at least in part driven by the performance of U.S. equities, but domestic stocks prices are still at the very low end of their post-GFC range when compared with global ex-U.S. stocks. Chart 9Still Too Early To Buy A-Shares... Chart 10...But The Selloff Seems Extremely Late In short, the potential for a substantial bounce in relative domestic equity performance is considerable were the economic outlook to stabilize, and we will be watching closely for an opportunity to time a reversal. Stay tuned! Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com 1 Investable industrial stocks in China have become relatively low-beta, owing to the fact that they had already materially underperformed the investable benchmark prior to the emergence of trade frictions with the U.S. Cyclical Investment Stance Equity Sector Recommendations