Economy

Highlights Duration & Fed: Our late-1990s & 2015/16 roadmap for the economy still holds, but risks are mounting. Despite the risks, we expect that trade tensions will calm enough for the economic data to improve during the next few months. The result will be one more Fed rate cut this month, followed by an extended on-hold period. Investors should keep portfolio duration low in that environment. Junk Quality Spreads: This year’s divergence between the Caa/Ba quality spread and the high-yield index spread is highly unusual, but has more to do with movements in Treasury yields and changing index duration than with broader concerns about corporate credit quality. Investment Grade Risk & Reward: We present a novel approach for assessing the risk/reward trade-off among investment grade corporate bond sectors. We note that Saudi Arabian and Mexican Sovereign bonds, Foreign Agency bonds and Conventional 30-year Agency MBS look particularly attractive in risk-adjusted terms. Feature Contagion? This publication has repeatedly pointed to the late-1990s and the 2015/16 periods as appropriate comparables for today’s global growth slowdown. That is, we expect that the current spate of weakness will stay confined within the manufacturing sector and will not spread into the broader economy, leading the U.S. into recession. This call is important from an investment perspective because it implies that the Fed is not currently engaged in an easing cycle that will bring the funds rate back to zero. Rather, we anticipate only three rate cuts this year (we’ve already seen two), followed by the eventual resumption of hikes. Bond yields will not make new lows in that environment. Chart 1Manufacturing Weakness Spreading? Chart 2"Hard" Data Still Firm But some data received this month challenge our economic narrative. Specifically, September’s drop in the ISM Non-Manufacturing PMI from 56.4 to 52.6 and the year-over-year decline in the Conference Board’s survey of consumer confidence (Chart 1). Both are sending tentative signals that economic weakness might be spreading from the manufacturing sector into the broader U.S. economy. The Fed is worried about the same thing, as evidenced by this passage from the September FOMC minutes: One risk that the economy faced was that the softness recorded of late in firms’ capital formation, manufacturing, and exporting activities might spread to their hiring decisions, with adverse implications for household income and spending. Participants observed that such an eventuality was not embedded in their baseline outlook; however, a couple of them indicated that this was partly because they assumed that an appropriate adjustment to the policy rate path would help forestall that eventuality. This passage makes two important points. First, it stresses the risk of contagion from manufacturing into services and consumer spending as a precondition for recession. This risk has clearly increased, but we are not yet ready to abandon our base case outlook. For one thing, Chart 1 shows that the ISM Non-Manufacturing survey printed at 51.8 for one month in 2016, before rebounding sharply. Second, the “hard” economic data paint a much rosier picture that the “soft” survey data (Chart 2). Industrial production has already bounced off its lows and, unlike the ISM Manufacturing PMI, has not yet approached 2015/16 levels. Similarly, new orders for capital goods are much stronger than during the 2015/16 period. As for consumer spending, it continues to grow at a rapid pace despite the drop in confidence. Chart 3Expect One Rate Cut In October The most logical explanation for the divergence between “hard” and “soft” data is that business and consumer sentiment are being pulled down by concerns about the ongoing trade war. Our sense is that some positive news on that front is now required to bring the survey data back into line with the “hard” numbers. On that note, we anticipate that the looming 2020 election will provide enough incentive for President Trump to reach some sort of détente with China. In fact, as we go to press, optimism about a potential trade deal has pushed the 10-year Treasury yield up above 1.70%. If this optimism is not vindicated, then weak survey data will eventually drag the “hard” data lower. The economy is at a critical and highly uncertain juncture. Amidst so much uncertainty, and with so much hinging on near-term political decisions, how should we expect the Fed to respond? The above passage from the September FOMC minutes gives us a strong clue. It illustrates that the Fed believes that sufficiently accommodative monetary policy will help mitigate the risk of contagion from manufacturing into services and consumer spending. In other words, the Fed must help weather the current storm by ensuring that financial conditions remain supportive. This means refraining from delivering hawkish surprises to market expectations.1 The Fed believes that sufficiently accommodative monetary policy will help mitigate the risk of contagion from manufacturing into services and consumer spending. With that in mind, we note that the market has mostly priced-in an October rate cut (Chart 3), and we expect the Fed to deliver on that expectation. Assuming an October cut, the market is only pricing-in a 28% chance of another cut in December. Overall, the market is priced for 59 basis points of rate cuts during the next 12 months. We anticipate a 25 bps cut this month, followed by an improvement in the economic data that will make further cuts unnecessary. Bottom Line: Our late-1990s & 2015/16 roadmap for the economy still holds, but risks are mounting. Despite the risks, we expect that trade tensions will calm enough for the economic data to improve during the next few months. The result will be one more Fed rate cut this month, followed by an extended on-hold period. Investors should keep portfolio duration low in that environment. High-Yield Quality Spreads: Less Than Meets The Eye Corporate bonds have generally performed quite well this year, but oddly, the lowest tier of junk has not kept pace (Chart 4). Investment grade excess returns have followed a typical risk-on pattern. That is, the lowest rated / riskiest credit tiers have performed best in a bull market. However, in the high-yield space, Caa-rated debt has bucked the trend and actually underperformed the duration-matched Treasury index by 33 bps. Chart 4Caa-Rated Junk Is Not Keeping Pace Is this a potentially worrying sign for corporate spreads more generally? To consider the question, we looked at the historical relationships between quality spreads – the spread differential between low-rated and high-rated credit tiers – and the overall index spreads for both investment grade and high-yield. We found a strong positive correlation in both cases, but no leading or lagging properties. That is, quality spreads tend to follow the same trend as the overall index spread, but do not flag signs of trouble before the overall index. Nonetheless, the current divergence between the Caa/Ba quality spread and the high-yield index spread is highly unusual (Chart 5). Our sense, however, is that the divergence has less to do with concerns about credit quality and more to do with this year’s large moves in Treasury yields and changes to bond index duration. Chart 5De-Coupling In Quality Spreads... Chart 6...Is Due To Duration   Specifically, we note that this year’s large decline in Treasury yields has caused junk index duration to plunge, but the drop has been greater for the Ba credit tier than the Caa credit tier (Chart 6). Ba index duration has fallen by 0.8 this year (from 4.4 to 3.5), while Caa index duration has fallen by 0.6 (3.4 to 2.8). The result is that if we control for changes in duration by looking at a 12-month breakeven spread instead of the average index option-adjusted spread (OAS), we see that the quality spread widening is roughly consistent with the overall index (Chart 6, panel 3).2 In other words, the steep drop in Treasury yields has not led to the same reduction in risk in the Caa credit tier as it has in the other junk credit tiers. Caa spreads have widened on a relative basis, as a result. This year’s large decline in Treasury yields has caused junk index duration to plunge. It’s also interesting to note that the opposite dynamic is afoot within the investment grade corporate space. The Baa/Aa quality spread is more or less consistent with the overall index spread in OAS terms (Chart 5, top panel), but the quality spread widening is exacerbated when the impact of changing duration is considered (Chart 6, panels 1 & 2). That is, index duration has lengthened by more for the upper credit tiers than it has for the Baa credit tier. This makes Baa corporates look particularly attractive in risk-adjusted terms, as we have noted in prior research.3 From a big picture perspective, it is unusual for Treasury yields to fall so much without a concurrent widening in credit risk premiums. Eventually, this anomaly will be resolved by either: Higher Treasury yields in the event that recession is avoided, or Wider credit spreads in the event of a contraction in U.S. economic activity But in the meantime, negatively convex sectors such as high-yield corporates and Agency MBS look particularly attractive on a risk-adjusted basis. These sectors have benefited from the drop in Treasury yields by seeing their durations fall. They should perform well as long as the current environment of low Treasury yields and stable credit spreads persists. We take a more detailed look at the prospects for risk-adjusted performance within the different investment grade bond sectors in the next section. Risk And Reward In Investment Grade Bond Sectors As mentioned above, in this week’s report we present a novel approach for considering the risk/reward trade-off between different investment grade sectors of the U.S. bond market. We consider 23 sectors in total: 4 corporate credit tiers Conventional 30-year Agency MBS and Agency CMBS Aaa-rated non-Agency CMBS, credit card ABS and auto loan ABS Domestic and Foreign Agency bonds Supranationals Local Authority bonds (mostly taxable munis and USD-denominated Canadian provincial debt) USD-denominated Sovereign bonds for 10 different emerging markets Reward First, we consider the reward side of the equation. We do not impose any macro view, but instead, use the average index OAS as the best estimate for each sector’s 12-month expected excess returns relative to a duration-matched position in Treasuries. Chart 7 shows the expected excess returns for each sector. Right away, the attractiveness of Mexican sovereign debt is apparent. Mexico carries an A rating, but offers a greater spread than the Baa corporate index. Chart 7Expected Returns Risk We decided to assess risk using a breakeven spread framework. We calculate a 12-month breakeven spread for each sector. This spread represents the basis point spread widening required for each sector to break even with a duration-matched position in Treasury securities on a 12-month horizon. We calculate the breakeven spread using the following equation: 0 = OAS – D(B) + 0.5*CVXs*(dYs)2 - 0.5*CVXT*(dYT)2 Where: OAS = the sector’s option-adjusted spread D = the sector’s duration B = the breakeven spread CVXs = the sector’s convexity CVXT = the convexity of a duration-matched Treasury security dYs = trailing 1-year volatility of the sector’s yield dYT = trailing 1-year volatility of the duration-matched Treasury yield Chart 8 shows each sector’s 12-month breakeven spread, and it illustrates that the breakeven spread is a sub-optimal measure of risk. In theory, the highest breakeven spreads should be the least likely to see losses, but this is obviously not the case. Baa-rated South African Sovereign debt carries the largest breakeven spread, but it should be among the riskiest of the sectors. Chart 812-Month Breakeven Spreads The missing piece of the puzzle is spread volatility. South African sovereign spreads need to widen by 39 bps before losses are incurred, while Aaa-rated credit card ABS spreads only need to widen by 13 bps. However, if spread volatility is much higher for South African sovereigns than for credit card ABS, then the sovereign sector still might be more likely to see losses. To control for this difference we calculate the standard deviation of annual spread changes for each sector, starting from May 2014 when all sectors have available data. We then divide each sector’s breakeven spread by the result. This calculation gives us a volatility-adjusted 12-month breakeven spread. In other words, it is the number of standard deviations of spread widening required for each sector to see losses on a 12-month horizon (Chart 9). Chart 912-Month Volatility-Adjusted Breakeven Spreads Risk & Reward We bring risk and reward together in Charts 10-12. Chart 10 shows expected returns on the y-axis and the vol-adjusted 12-month breakeven spread on the x-axis. Sectors plotting near the top-right of the chart give the best returns and lowest risk of losses, while sectors plotting near the bottom-left provide low expected returns and high risk of losses. Immediately, Saudi Arabian sovereigns and Foreign Agency debt stand out as offering high expected returns for their risk levels. Note that South African sovereigns plot off the charts, toward the top-left of Charts 10-12, as indicated by the arrows. Chart 10Expected Returns Vs. Risk Of Negative Excess Returns Chart 11Expected Returns Vs. Risk Of Losing 100 BPs Chart 12Expected Returns Vs. Risk Of Losing 200 BPs In Charts 11 and 12 we make one further refinement to our risk measure. In these charts, instead of calculating 12-month breakeven spreads, we calculate the spread change necessary for each sector to underperform Treasuries by 100 bps and 200 bps, respectively. Saudi Arabian sovereigns and Foreign Agency debt stand out as offering high expected returns for their risk levels. This adjustment arguably gives a more useful perspective on risk. For example, because spreads are quite narrow in the Supranational and Domestic Agency sectors, the risk of negative returns versus Treasuries is quite elevated. However, these sectors also carry high credit ratings and low spread volatility, making it exceedingly unlikely that they would deliver losses of 100 bps or more. Considering Charts 11 and 12, we look for sectors that clearly dominate other ones, i.e. plotting both higher and further to the right. Once again, Foreign Agencies and Saudi Arabian sovereigns both look very appealing. Mexican sovereign debt also offers very high expected return, and less risk that the Baa corporate sector. We would also like to point out the attractiveness of Agency MBS. As we noted in a recent report, Agency MBS offer considerably less risk than high-rated corporate debt, and similar expected returns. Note that this analysis doesn’t impose any macroeconomic view, and our sense is that the macro back-drop is more favorable for MBS spreads than for corporates.4 All in all, we reiterate our recommendation to favor Agency MBS over Aaa-, Aa- and A-rated corporate bonds. We will continue to refine this approach to measuring the risk/reward trade-off in the coming weeks, including incorporating high-yield debt into our analysis. Stay tuned. Ryan Swift, U.S. Bond Strategist rswift@bcaresearch.com Footnotes 1 For further discussion on this topic please see U.S. Bond Strategy Weekly Report, “Act As Appropriate”, dated August 27, 2019, available at usbs.bcaresearch.com 2 The 12-month breakeven spread is the spread widening required on a 12-month horizon to break even with a duration-matched position in Treasury securities. It can be approximated by dividing the option-adjusted spread by duration, as is done in Chart 6. 3 Please see U.S. Bond Strategy Weekly Report, “Two Themes And Two Trades”, dated October 1, 2019, available at usbs.bcaresearch.com 4 Please see U.S. Bond Strategy Weekly Report, “Two Themes And Two Trades”, dated October 1, 2019, available at usbs.bcaresearch.com Fixed Income Sector Performance Recommended Portfolio Specification

Highlights Portfolio Strategy The trade-weighted U.S. dollar’s appreciation along with the still souring manufacturing data are weighing on SPX profit growth, at a time when heightened geopolitical uncertainty and a looming reversal in financial conditions has the potential to wreak havoc on stock prices. Stay cautious on the prospects of the broad equity market on a cyclical 9-12 month time horizon. Firming operating metrics, the resilient U.S. dollar, compelling valuations and depressed technicals, all signal that there is an exploitable tactical trading opportunity in a long S&P industrials/short S&P tech pair trade, irrespective of the trade war outcome. A tentative tick up in EM and China data along with improving relative operating metrics signal that the time is ripe to initiate a long machinery/short semis pair trade. Recent Changes Initiate a long S&P Industrials/short S&P Tech pair trade on a tactical three-to-six month time horizon, today. Initiate a long S&P Machinery/short S&P Semiconductors pair trade on a tactical three-to-six month time horizon, today. Feature The S&P 500 oscillated violently again last week, as the barrage of declining economic data, heightened trade war-related volatility and political upheaval dominated the news flow. While the Fed remains the backstop of last resort, we doubt additional interest rate cuts, which are already aggressively priced in the bond market, will boost lending and entice CEOs to invest in capital expenditure projects. Investors have to stay patient and disciplined, let this economic slowdown play out and allow for the natural healing of the economy. As a reminder, the ISM manufacturing index has been decelerating for twelve months and only been below the boom bust line for two. If history is an accurate guide, an additional three-to-six months of manufacturing pain are in store before a definitive bottom is in place (bottom panel, Chart 1). Such a macro backdrop, still warrants caution on the prospects of the broad equity market. Chart 1Allow Time For Economic Healing Beginning in August, a number of BCA publications became a tad more cautious on risk assets. Following our October editorial view meeting last week, this cautiousness was cemented with a tactical downgrade of global equities to neutral from previously overweight in the BCA House View matrix. While this marks a clear shift toward this publication’s less sanguine view of the U.S. equity market adopted during the summer, BCA's cyclical 12-month House View remains overweight global equities. Worryingly, the majority of the indicators we track continue to emit distress signals and warn that the SPX has further downside (Chart 2), especially absent profit growth. Importantly, we first correctly posited last May that the back half of the year global growth reacceleration was in jeopardy and would go on hiatus courtesy of rising policy uncertainty.1 Such a backdrop would boost the U.S. dollar and simultaneously take a bite out of SPX EPS.2 Chart 2Soft Data Red Flag Last week we highlighted that the U.S. dollar is the most important indicator to monitor given its global deflationary/reflationary properties. Were the greenback to maintain its year-to-date gains, it will continue to dent SPX profitability via P&L translation loss effects and likely sustain the profit recession into early 2020 (trade-weighted U.S. dollar shown inverted, bottom panel, Chart 3). Chart 3Greenback Weighing On Profits U.S. Equity Strategy’s S&P 500 four-factor macro EPS growth model remains downbeat (middle panel, Chart 4). Were we to isolate the U.S. dollar as a single variable and re-run the regression it is clear that additional greenback appreciation will further weigh on SPX profit growth (bottom panel, Chart 4). Meanwhile, the easing in financial conditions and drubbing of the 10-year Treasury yield since the Christmas Eve lows is already reflected in the 23% jump in the forward PE multiple, which explains over 90% of the SPX’s rise since the Dec 24, 2018 trough (top & middle panels, Chart 5). In other words, for multiples to expand anew, financial conditions would have to further ease, which in our view is a tall order (bottom panel, Chart 5). Chart 4EPS Model Warrants Caution Chart 5Financial Conditions Are The Forward P/E This week we are initiating two related pair trades to exploit the mispricing of the trade war within the deep cyclical sector universe.  Thus, we would lean against the narrative that easy financial conditions are not fully reflected into stocks. In contrast, our worry is that junk spreads are on the verge of a breakout and such a backdrop would tighten financial conditions and aggravate an SPX drawdown (junk OAS shown inverted, Chart 6). Adding it all up, the trade-weighted U.S. dollar’s appreciation along with the still souring manufacturing data are weighing on SPX profit growth, at a time when heightened geopolitical uncertainty and a looming reversal in financial conditions has the potential to wreak havoc on stock prices. Stay cautious on the prospects of the broad equity market on a cyclical 9-12 month time horizon. This week we are initiating two related pair trades to exploit the mispricing of the trade war within the deep cyclical sector universe. Chart 6Watch Junk Spreads Initiate A Long Industrials/Short Tech Pair Trade… Ever since the Sino-American trade war started in March 2018, the market has punished industrials, but tech has escaped unscathed. While the global growth soft patch preceded the U.S./China trade spat, courtesy of the Fed’s tightening cycle and Chinese policymakers’ slamming on the brakes, the trade war has served as a catalyst to aggressively shed deep cyclical equities except for tech stocks (Chart 7). We think this misalignment presents a playable opportunity to generate alpha by going long industrials/short tech, irrespective of the trade war’s outcome. In other words, this market neutral trade will be in the black either because the trade spat gets resolved or because there will effectively be no “real” deal including intellectual property and the tech sector. If the two sides manage to iron out their differences and strike a deal, industrials stocks should benefit from a greater catch-up phase because they have been depressed over the past two years, while tech stocks are near relative all-time highs. In contrast, a “no deal” scenario, should also re-concentrate investors’ minds and lead to a relative selling in tech stocks versus their already beaten-down deep cyclical peers: industrials. Chart 7Bifurcated Deep Cyclicals Market Chart 8Lots Of Bad Trade War News Reflected In Prices Chart 8 shows the drubbing in relative share prices as three key macro drivers have felt the trade war’s wrath. In more detail, were a deal to get struck, growth expectations will reverse course and a bond market sell-off will almost immediately reflect such an improvement in the global macro backdrop. Rising interest rates on the back of a reflationary/inflationary impulse are a boon for industrials and a bane for high growth tech stocks (top panel, Chart 8). Similarly, the middle panel of Chart 8 highlights that the ISM manufacturing survey should climb above the boom/bust line and outshine the San Francisco Fed’s Tech Pulse Index (that comprises “coincident indicators of activity in the U.S. information technology sector”3) on news of a successful deal. Finally, relative capital expenditure outlays should also veer in favor of industrials as previously mothballed infrastructure projects will come out of hibernation (bottom panel, Chart 8). In contrast, tech capex has been resilient of late with analytics, security and cloud computing being the most defensive capex corner, leaving little room for additional relative capex gains. Taking the opposite side i.e. a “no deal”, we doubt the metrics we depict in Chart 8 would sink that much further. If anything we believe that there is an element of exhaustion and relative share prices would jump on news of a breakdown in trade talks as tech sector fire sales would trump the sell-off in already depressed industrials. Meanwhile, the U.S. dollar and relative share prices have been steeply diverging recently and this gap will likely narrow via a catch-up phase in the latter (top & middle panels, Chart 9). According to Factset’s latest data the S&P industrials sector garners 37% of its sales from abroad, whereas the S&P information technology sector’s foreign exposure stands at 57% of total revenues.4 Therefore, given this 20% delta, a rising greenback should be beneficial to the more domestically geared industrials stocks (bottom panel, Chart 9). On the operating front, industrials also have the upper hand. The relative wage bill is sinking like a stone (shown inverted, middle panel, Chart 10) at a time when relative selling price inflation is holding its own (top panel, Chart 10). The upshot is that a relative profit margin jump is in store in the coming months which should boost the relative share price ratio (bottom panel, Chart 10). Chart 9Unsustainable Divergence Chart 10Industrials Have The Upper Hand U.S. Equity Strategy’s proprietary relative Cyclical Macro Indicators and relative profit growth models capture all these drivers and both signal that an industrials versus tech earnings-led outperformance phase looms into year end (Chart 11). Chart 12 shows that the relative earnings breadth and relative net earnings revisions are both deep in negative territory. In terms of technicals, the relative percentage of groups trading with a positive 52-week rate of change has hit the lowest level in the past two decades (second panel, Chart 12) and our composite relative technical indicator is roughly one standard deviation below the historical mean (bottom panel, Chart 11). Chart 11Profit Models And...  Chart 12...Washed Out Breadth Say Buy Industrials At The Expense Of Tech Finally, relative valuations are also bombed out. Our relative valuation indicator has been in a six-year uninterrupted drop, falling from two standard deviations above the mean to one standard deviation below the mean (fourth panel, Chart 11). Such entrenched bearishness in relative value is unwarranted. Bottom Line:  Firming operating metrics, the resilient U.S. dollar, compelling valuations and depressed technicals, all signal that there is an exploitable tactical trading opportunity in a long S&P industrials/short S&P tech pair trade, irrespective of the trade war outcome. …And A Long Machinery/Short Semis Pair Trade A more speculative and higher octane vehicle to explore this trade war-related mispricing is via a long S&P machinery/short S&P semiconductors pair trade. Most of the drivers mentioned above also hold true in this subsector market-neutral trade. However, in this section we will drill deeper in the China/EM drivers. The Emerging Asia leading economic indicator (EALEI) has plummeted to levels last hit around the 1998 LTCM bailout (top panel, Chart 13). While more pain is likely in the coming months as global trade has ground to a halt, we doubt the carnage in the EALEI can continue indefinitely. In fact, a tentative trough in the Emerging Markets (EM) manufacturing PMI heralds a brighter outlook for relative share prices (bottom panel, Chart 13). Chart 13Same Trade War Theme, Different Vehicles To Play It Chart 14China...  Encouragingly, China’s fiscal and credit impulse also signals that a bottom in relative share prices is likely already in place. If this leading indicator proves accurate in the coming months, then relative share prices can spike 20% near the late-2018 highs (Chart 14).   Chinese money supply growth is showing some signs of life and capital committed to infrastructure spending is coming out of hibernation. Goldman Sachs’ China current activity indicator is on a similar upward trajectory, underscoring that the path of least resistance is higher for relative share prices (Chart 15). Chart 15...Holds The Key Chart 16Firming Final Demand... On the operating front, relative new orders and relative shipment growth have both ticked higher (top & middle panels, Chart 16). Importantly, our relative demand proxy suggests that the relative end-demand backdrop is also firming. Using Caterpillar’s global sales to dealers data compared with global chip sales reveals that a wide gap has formed between relative share prices and our relative demand gauge (bottom panel, Chart 16). If our thesis pans out in the upcoming three-to-six months then machinery will trounce semis. Finally, relative pricing power corroborates that machinery demand has the upper hand versus semiconductor final demand. The Commodity Research Bureau’s raw industrials index is climbing relative to Asian DRAM prices. The upshot is that the compellingly valued relative share price ratio will gain steam in the months ahead (Chart 17). In sum, a tentative up-tick in EM and China data along with improving relative operating metrics signal that the time is ripe to initiate a long machinery/short semis pair trade. Bottom Line: Initiate a long S&P machinery/short S&P semiconductors pair trade today. The ticker symbols for the stocks in the S&P machinery and S&P semis indexes are: BLBG – S5MACH – CAT, DE, ITW, IR, CMI, PCAR, PH, SWK, FTV, DOV, XYL, IEX, WAB, SNA, PNR, FLS, and BLBG – S5SECO – INTC, TXN, NVDA, AVGO, QCOM, MU, ADI, AMD, XLNX, QRVO, MCHP, MXIM, SWKS, respectively. Chart 17...Is A Boon To Relative Pricing Power Key Risk To Monitor One important risk to both of our newly recommended market-neutral trades is China. We recently touched base with our ex-Chief Geopolitical Strategist and currently Chief Strategist at the Clocktower Group, Marko Papic. He warned us that all bets would be off because: “I think we will look back at the recession of 2020 and it will be known as the “China recession”. Basically, China just decided to stop playing, pick up its toys, and go home”. If Marko’s wise words were to ring true, then such a Chinese policy shift will truly be a game changer with negative global economic growth implications. With regard to our pair trades, they would both be offside.   Anastasios Avgeriou, U.S. Equity Strategist anastasios@bcaresearch.com   Footnotes 1      Please see BCA U.S. Equity Strategy Weekly Report, “Consolidation” dated May 21, 2019, available at uses.bcaresearch.com. 2      Please see BCA U.S. Equity Strategy Weekly Report, “On Edge” dated May 13, 2019, available at uses.bcaresearch.com. 3      https://www.frbsf.org/economic-research/indicators-data/tech-pulse/ 4      https://www.factset.com/hubfs/Resources%20Section/Research%20Desk/Earnings%20Insight/EarningsInsight_100419A.pdf Current Recommendations Current Trades Size And Style Views Stay neutral cyclicals over defensives   (downgrade alert) Favor value over growth Favor large over small caps (Stop 10%)

Highlights The manufacturing slowdown, on its own, is unlikely to tip the economy into a recession. The sector accounts for a small share of U.S. output and employment, and will gain a tailwind from a pick-up in global growth. A larger and more stable service sector mitigates manufacturing’s impact on the employment and consumption outlook. The bar is too high for manufacturing job losses to lift the overall unemployment rate towards recession-inducing levels. The recent divergence between alternative measures of U.S. manufacturing activity confirms the resilience of the domestic manufacturing sector relative to the rest of the world. Feature Manufacturing activity has been the most prominent casualty of the trade war between the U.S. and China, and global manufacturing PMIs have languished as tensions have intensified with no clear end in sight. Throughout the spring and early summer, manufacturing activity in the comparatively closed U.S. economy held up better than it did overseas. In August, however, the ISM Manufacturing PMI finally crossed the 50 expansion/contraction line and subsequently dipped well below it in September. Evidence of weakness was broad-based throughout September’s report and the fact that forward-looking components like new orders, new export orders and backlogs of orders all contracted further has caught our attention. Although, like most developed markets, the U.S. is a service economy, and consumption accounts for the lion’s share of its GDP, it is certainly not immune to manufacturing cycles. We are not turning a blind eye to the global manufacturing slowdown, nor downplaying its magnitude, but for now we are not overly worried about it. Regular readers know that we continue to believe that the fundamentals of the U.S. economy remain strong, supported most of all by an especially robust labor market. The manufacturing slowdown is near the top of investors’ concerns, however, so we measure how severe a manufacturing slowdown would have to be to cause serious harm to the U.S. economy. We find that the bar is high and the slowdown has low odds of getting that bad if, as we expect, global growth eventually recovers. Until a pick-up truly materializes, we remain comforted by our expectation that buoyant consumption and government spending will keep the U.S. economy out of too much trouble. David And Goliath Chart 1Services May Be Larger, But Goods Punch Harder Technology and globalization have revolutionized the manufacturing process and disrupted the global economic landscape. As the outsourcing of manufacturing activities to lower-cost countries has become more and more prevalent, developed markets have steadily transitioned to service economies. Since the 1950s, goods-producing sectors’ share of U.S. GDP has decreased from half to 29%. Nevertheless, a third of the economy is not negligible, especially when it swings much more wildly than the services sector, which is more than twice its size (Chart 1). In a previous report1 where we looked at the components of the U.S. GDP equation, we showed that smaller, more volatile fixed investment was considerably more likely to negate trend growth in the rest of the economy than giant, but stable, consumption. This narrative echoes the dynamics at play with the manufacturing portion of the U.S. economy. Given their greater variability, goods-producing sectors are just as likely to wipe out 2% trend growth in services as services are to wipe out 2% manufacturing growth (Table 1). Table 1Another Road To Recession This would be bad news if we thought the manufacturing slowdown had a lot more downside. We continue to believe in a global growth recovery narrative, however, powered by impending Chinese stimulus and revived trade negotiations. U.S. industrial production and capacity utilization both surprised to the upside in August and global growth is showing budding signs of a recovery (Chart 2). Moreover, our colleagues at Global Investment Strategy have found that industrial cycles last an average of 36 months, divided into an 18-month uptrend and an 18-month downtrend.2 Absent any major trade deterioration, the tenure of the current down leg suggests that an upturn in manufacturing activity is on its way (Chart 3). Chart 2Towards A Global Growth Pick-Up Chart 3The Global Manufacturing Cycle Has Likely Reached A Bottom   Another channel through which the manufacturing slowdown could hurt the U.S. economy is via manufacturing job losses and the detrimental effect they would have on overall U.S. consumption. Goods-producing sectors employ 21.1 million, including 12.9 million in manufacturing roles - a puny 14% and 8% of total nonfarm payrolls, respectively. The productivity gains that technological improvements and automated processes have unlocked over the years have allowed a modest share of U.S. workers who make tangible things to produce double their proportionate share of U.S. output. Bottom Line: Goods-producing sectors represent less than a third of U.S. GDP and less than a sixth of U.S. jobs. That’s enough for the global manufacturing slowdown to cause some domestic slowing, but not enough to end the expansion on its own. A High Pain Threshold Akin to the goods-producing sectors’ contribution to overall U.S. GDP, aggregate manufacturing payrolls tend to exhibit more volatility than aggregate services payrolls, particularly on the downside (Chart 4). Before the 1980s, because manufacturing activity accounted for a larger share of the U.S. economy and created a larger portion of jobs, a mere deceleration in the pace of payroll expansion was sufficient to tip the economy into a recession. The paradigm has shifted and it now takes a more severe manufacturing downturn to inflict real harm on the U.S. economy. Since the 1980s, no recession has occurred independent of a full-on contraction in manufacturing employment. We are not there yet, as manufacturing payrolls are still growing at a 1.1% pace. Aggregate manufacturing payrolls tend to exhibit higher volatility than aggregate services payrolls, particularly on the downside. Chart 4A Paradigm Shift Our Global Investment Strategy colleagues have previously shown that throughout the post-war era, whenever the 3-month moving average of the unemployment rate has risen by at least a third of a percentage point from its cyclical lows, a recession has ensued (Chart 5). The U.S. unemployment rate just made a fifty-year low and we do not expect a quick material reversal in the short run. A resilient service sector, ambitious hiring plans and elevated levels of job openings, coupled with a revival in global growth, should hold the U.S. unemployment rate in check for the time being (Chart 6). Chart 5The Recession-Inducing Level Of Unemployment... Chart 6...Is Not Imminent Given Strong Hiring Plans Investors are right to be concerned about the manufacturing slowdown nonetheless. To address those concerns more closely, and to challenge our own view, we calculated the number of manufacturing job losses that would be required to push the unemployment rate up to recession-inducing levels. The U.S. unemployment rate fell to a fifty-year low of 3.5% in September, tugging the 3-month moving average down to 3.6%. There are several paths the unemployment rate can take from current levels for its 3-month moving average to grow by a third of a percentage point. It may gain a linear 10 basis points a month and reach a 3.9% average in the fifth month. Myriad non-linear paths could get the moving average to 3.9% in more or less than five months. For the sake of this exercise, we do not choose a particular path, but simply assume that the 3-month moving average of the unemployment rate reaches 3.9% over three, six and twelve months. We build on the work of the economists at the Atlanta Fed and calculate the number of manufacturing job losses required to achieve a 3.9% target unemployment rate over those three timeframes. We used the Atlanta Fed Jobs Calculator’s3 default inputs, and the details and results of our subsequent calculations are summarized in Table 2. Table 2The Payroll Road To Recession Chart 7The Bar Is High For Manufacturing To Trigger A Recession Under the default assumptions of a constant participation rate and a population growth rate unchanged from the past twelve months’, it would take 313,000 job losses over three months for the overall U.S. unemployment rate to reach 3.9%. We assume that the private service sector, which shows no sign of distress, will continue to add jobs. It has done so at a historical average monthly growth rate of 0.19% but given that the overall economy has clearly slowed, we assume instead that the service sector will continue to add jobs at the slower 0.13% pace of the past twelve months. Under this more conservative assumption, the economy would gain 560,000 nonmanufacturing jobs over the next three months. Consequently, it would take 873,000 manufacturing job losses alone to offset these gains and lift the unemployment rate to 3.9% within three months. Over a six- and twelve-month horizon, the number of manufacturing job losses required to offset payroll expansion in services reaches 1.1 and 1.6 million, respectively.4 These levels of manufacturing job losses – equivalent to a 7% to 12% contraction in manufacturing payrolls - seem like a stretch in the current macroeconomic backdrop. The only time in the past seventy years when the U.S. economy experienced manufacturing job losses of this magnitude on a 3- month time period was in the first quarter of 1975, when the U.S. economy confronted a tripling of oil prices from the oil embargo. Manufacturing job losses in excess of 1.1 and 1.6 million jobs over a 6- and 12-month horizon have historically been more attainable (Chart 7). That said, manufacturing payrolls are still expanding on a 6- and 12-month horizon, albeit at a decelerating pace. Not only are manufacturing payrolls gains far from recession-inducing levels, manufacturing employment will gain a tailwind from the pick-up in global growth and turn in global industrial production cycles that we expect. These levels of manufacturing job losses – equivalent to a 7% to 12% contraction in manufacturing payrolls – seem like a stretch in the current macroeconomic backdrop. Bottom Line: The bar seems a little too high for the manufacturing slowdown alone to destroy enough jobs to tip the U.S. economy into a full-fledged recession. What Oil Shock? One can argue that the September oil shock caused by attacks on Saudi energy infrastructure will exert further pressure on global manufacturing activities. While it is true that large jumps in oil prices have often preceded recessions, we think the probability is slight that September’s event will jeopardize the prospects of a global growth recovery (Chart 8). Chart 8Oil Spikes And Recessions Chart 9U.S. Output Is Less Dependent On Oil First, not only was the September surge in oil prices tame relative to the spikes that have preceded past recessions, but the quicker-than-expected return of Saudi oil production has calmed markets. For now, the oil scare ended as quickly as it appeared. Second, higher oil prices are less of a drag on the U.S. economy than they were in the 1970s, as the country has become one of the largest oil-producing countries in the world and approaches true energy independence. The gradual shift from a manufacturing to services economy has also reduced the oil intensity of the U.S. economy to a little more than a third of what it was at the time of the 1970s oil embargo (Chart 9). Moreover, higher gasoline prices are less likely to hurt U.S. consumers now that filling the tank takes up a smaller portion of their wallets (Chart 10). As Fed Chair Jay Powell put it in a speech last week, “we now judge that a price spike would likely have nearly offsetting effects on U.S. GDP.” Chart 10Filling The Tank Takes Up A Smaller Portion Of Consumers' Wallets Conflicting Messages? Chart 11The ISM Manufacturing PMI's Sensitivity To Global Growth The ISM Manufacturing Composite PMI is our favored measure of U.S. manufacturing activity as its long track record allows for comparison across multiple business cycles. Although it only offers insights back to 2011, the alternative IHS Markit Manufacturing PMI is nevertheless widely watched by investors and we take note of its moves. While the recent ISM readings have been dismal, the Markit Manufacturing PMI for the U.S. accelerated to 51.1 in September. At first glance, it might seem that both readings are contradicting each other. In fact, the current divergence is not unprecedented and stems from differences in sub-component weighting methodology and in sample size and composition. The ISM reading focuses on larger multinational companies, whereas the U.S. Markit PMI polls a wider array of companies by size. Multinationals’ earnings are more directly affected by global growth developments than smaller and domestically-focused firms. Therefore, in periods of accelerating global and ex-U.S. growth, the ISM PMI tends to score higher than the Markit PMI, and vice versa (Chart 11). A still-expanding Markit Manufacturing PMI combined with a contracting ISM Manufacturing PMI simply reinforces the argument that the domestic manufacturing sector is more resilient than ex-U.S. manufacturing activity, and highlights the potential for an improvement in business confidence if the U.S. and China can reach some sort of detente. Investment Implications In spite of evidence that global manufacturing weakness is spreading, our overall assessment of the U.S. economy remains intact. Assuming an exogenous event does not snuff out the expansion, we do not expect the next recession to occur until after monetary policy turns restrictive. Since the Fed has pivoted to accommodation, along with the world’s other major central banks, we have pushed out our recession timetable back to at least the middle of 2021. We therefore think it is too early to de-risk investment portfolios. We have previously shown that bull markets tend to sprint to the finish line and we remain bullish on a 12-month cyclical horizon. Though we are not concerned that the end of the cycle is at hand, tariff tensions are squeezing trade flows and business confidence. Volatility is likely to remain elevated in the near term until trade tensions die down and the global economy demonstrates that an upturn is at hand. We are therefore neutral on equities over the tactical 0-to-3-month timeframe and recommend investors overweight cash to keep some dry powder at hand. We still recommend that investors underweight bonds in balanced portfolios.   Jennifer Lacombe, Senior Analyst jenniferl@bcaresearch.com Footnotes 1 Please see U.S. Investment Strategy Weekly Report, “If We Were Wrong”, dated April 8, 2019, available at usis.bcaresearch.com. 2 The reason underpinning this cyclicality is that most purchased goods retain some value for a certain amount of time before they need to be replaced. 3 The Atlanta Fed Jobs Calculator tool is available at https://www.frbatlanta.org/chcs/calculator?panel=1 4 Had we assumed that the nonmanufacturing payrolls continue to grow at the historical average monthly rate of 0.19% instead, the levels of manufacturing job losses required to offset the nonmanufacturing gains and lift the unemployment rate to 3.9% would be 1.1 million, 1.6 million and 2.6 million manufacturing job losses over a 3-, 6- and 12-month time horizon, respectively.

Highlights Geopolitical risks are starting to abate as a result of material constraints influencing policymakers. China needs to ensure its economy bottoms and a debt-deflationary tendency does not take hold. President Trump needs to avoid further economic deterioration arising from the trade war. The U.K. is looking to prevent a recession induced by leaving the EU without an agreement. Iran and the risk of an oil price shock is the outstanding geopolitical tail risk. Feature Readers of BCA’s Geopolitical Strategy know that what defines our research is our analytical framework – specifically the theory of constraints. Chart 1The Electoral College – An Overlooked Constraint The theory holds that policymakers are trapped by the pressures of their office, their nation’s global position, and the stream of events. These pressures emerge from the material world that we inhabit and as such are measurable. If a leader lacks popular approval, cannot command a majority in the legislature, rides atop a sinking economy, or suffers under stronger or smarter foreign enemies, then his policy preferences will be compromised. He will have to change his preferences to accommodate the constraints, rather than the other way around. Case in point is the U.S. electoral college: it proved an insurmountable political constraint on the Democratic Party in 2016. The college is intended to restrain direct democracy or popular passions; it also restrains the concentration of regional power. In 2012, Barack Obama won a larger share of the electoral college than the popular vote, while in 2016 Hillary Clinton won a smaller share (Chart 1). Clinton’s lack of appeal in the industrial Midwest turned the college and deprived her of the prize. The rest is history. In this report we highlight five key constraints that will shape the direction of the major geopolitical risks in the fourth quarter. We recommend investors remain tactically cautious on risk assets, although we have not yet extended this recommendation to the cyclical, 12-month time frame. China’s Policy: The Debt-Deflation Constraint We have a solid record of pessimism regarding Chinese President Xi Jinping’s willingness and ability to stimulate the economy – but even we were surprised by his tenacity this year. His administration’s effort to contain leverage, while still stimulating the economy, has prevented a quick rebound in the global manufacturing cycle. The constraint limiting this approach is the need to avoid a debt-deflation spiral. This is a condition in which households and firms become pessimistic about the future and cut back their spending and borrowing. The general price level falls and drives up real debt burdens, which motivates further cutbacks. A classic example is Japan, which saw a property bubble burst, destroying corporate balance sheets and forcing the country into a long phase of paying down debt amid falling prices. China has not seen its property bubble burst yet. Prices have continued to rise despite the recent pause in the non-financial debt build-up (Chart 2). Looser monetary and fiscal policy have sustained this precarious balance. But the result is a tug-of-war between the government and the private sector. If the government miscalculates, and the asset bubble bursts, then it will be extremely difficult for the government to change the mindset of households and companies bent on paying down debt. It will be too late to avoid the vicious spiral that Japan experienced – with the critical proviso that Chinese people are less wealthy than the Japanese in 1990 and the country’s political system is less flexible. A Japan-sized economic problem would lead to a China-sized political problem. This is why the recent drop in Chinese producer prices below zero is a worrisome sign (Chart 3). Policymakers have loosened monetary and fiscal policy incrementally since July 2018 and they are signaling that they will continue to do so. This is particularly likely in an environment in which trade tensions are reduced but remain fundamentally unresolved – which is our base case. Chart 2China's Property Bubble Intact Chart 3China's Constraint Is Debt-Deflation Are policymakers aware of this constraint? Absolutely. If the trade talks collapse, or the global economy slumps regardless, then China will have to stimulate more aggressively. Xi Jinping is not truly a Chairman Mao, willing to impose extreme austerity. He oversaw the 2015-16 stimulus and would do it again if he came face to face with the debt-deflation constraint. Is China still capable of stimulating? High debt levels, the reassertion of centralized state power, and the trade war have all rendered traditional stimulus levers less effective by dampening animal spirits. Yet policymakers are visibly “riding the brake,” so they can remove restraints and increase reflation if necessary. Most obviously, authorities can inject larger fiscal stimulus. They have insisted that they will prevent easy monetary and credit policies from feeding into property prices – and this could change. They could also pick up the pace when it comes to reducing average bank lending rates for small and medium-sized businesses.1 In short, stimulus is less effective, but the government is also preferring to save dry powder. This preference will be thrown by the wayside if it hits the critical constraint. The implication is that Chinese stimulus will continue to pick up over a cyclical, 12-month horizon. There is impetus to reduce trade tensions with the U.S., discussed below, but a lack of final resolution will ensure that policy tightening is not called for. Bottom Line: China’s chief economic constraint is a debt-deflation trap. This would engender long-term economic difficulties that would eventually translate into political difficulties for Communist Party rule. If a trade deal is reached, it is unlikely alone to require a shift to tighter policy. If the trade talks collapse, stimulus will overshoot to the upside. Trade War: The Electoral Constraint The U.S. and China are holding the thirteenth round of trade negotiations this week after a summer replete with punitive measures, threats, and failed restarts. Tensions spiked just ahead of the talks, as expected. Immediately thereafter President Trump declared he will meet with Chinese negotiators to give a boost to the process and reassure the markets.2 Trump’s major constraint in waging the trade war is economic, not political. Americans are generally sympathetic to his pressure campaign against China. Public opinion polls show that a strong majority believes it is necessary to confront China even though the bulk of the economic pain will be borne by consumers themselves (Chart 4). Yet Americans could lose faith in Trump’s approach once the economic pain fully materializes. Critically, the decline in wage growth that is occurring as a result of the global and manufacturing slowdown is concentrated in the states that are most likely to swing the 2020 election, e.g. the “purple” or battleground states (Chart 5). Chart 4Americans To Confront China Despite The Costs? Chart 5Trump Faces Pressure To Stage A Tactical Trade Retreat Furthermore, a rise in unemployment, which is implied by the recent decline in the University of Michigan’s survey of consumer confidence regarding the purchase of large household goods, would devastate voters’ willingness to give Trump’s tariff strategy the benefit of the doubt (Chart 6). Wisconsin and Pennsylvania, two critical states, have seen a net loss of manufacturing jobs on the year. The fear of an uptick in U.S. unemployment will prevent Trump from escalating the trade war. An uptick in unemployment would be a major constraint on Trump’s trade war – he cannot escalate further until the economy has stabilized. And that may very well require tariff rollback while trade talks “make progress.” We expect that Trump is willing to do this in the interest of staying in power. As highlighted above, the Xi administration is not without its own constraints. Our proxies for China’s marginal propensity to consume show that Chinese animal spirits are still vulnerable, particularly on the household side, which has not responded to stimulus thus far (Chart 7). Since this constraint is less immediate than Trump’s election date, Xi cannot be expected to capitulate to Trump’s biggest demands. Hence a ceasefire or détente is more likely than a full bilateral trade agreement. Chart 6Waning Consumer Confidence On Big Ticket Items Foreshadows Rise In Unemployment Trump’s electoral constraint also suggests that he needs to remove trade risks such as car tariffs on Europe and Japan (which we expect he will do). We have been optimistic on the passage of the USMCA trade deal but impeachment puts this forecast in jeopardy. Chart 7China's Trade War Constraint? Animal Spirits   Bottom Line: Trump will stage a tactical retreat on trade in order to soften the negative impact on the economy and reduce the chances of a recession prior to the November 3, 2020 election. China’s economic constraints are less immediate and it is unlikely to make major structural concessions. Hence we expect a ceasefire that temporarily reduces tensions and boosts sentiment rather than a bilateral trade agreement that initiates a fundamental deepening of U.S.-China economic engagement. U.S. Policy: The Economic Constraint The 2020 U.S. election is a critical political risk both because of the volatility it will engender and because of what we see as a 45% chance that it will lead to a change in the ruling party governing the world’s largest economy. Will Trump be the candidate? Yes. If Trump’s approval among Republicans breaks beneath the lows plumbed during the Charlottesville incident in 2017 (Chart 8A), then Trump has an impeachment problem, but otherwise he is safe from removal. Judging by the Republican-leaning pollster Rasmussen, which should reflect the party’s mood, Trump’s approval rating has not broken beneath its floor and may already be bouncing back from the initial hit of the impeachment inquiry (Chart 8B). The rise in support for impeachment and removal in opinion polls is notable, but it is also along party lines and will fade if the Democrats are seen as dragging on the process or trying to circumvent an election that is just around the corner. Chart 8ARepublican Opinion Precludes Trump’s Removal Chart 8BRepublican-Leaning Pollster Shows Support Holding Thus Far How will all of this bear on the 2020 election? Turnout will be high so everything depends on which side will be more passionate. A critical factor will be the Democratic nominee. Former Vice President Joe Biden, the establishment pick, has broken beneath his floor in the polling. His rambling debate performances have reinforced the narrative that he is too old, while the impeachment of Trump will fuel counteraccusations of corruption that will detract from Biden’s greatest asset: his electability. According to a Harvard-Harris poll from late September, 61% of voters believe it was inappropriate for Biden to withhold aid from Ukraine to encourage the firing of a Ukrainian prosecutor even when the polling question makes no mention of any connection with Biden’s son’s business interest there. Moreover, 77% believe it is inappropriate that Biden’s son Hunter traveled with his father to China while soliciting investments there. With Vermont Senator Bernie Sanders’s candidacy now defunct as a result of his heart attack and old age, Elizabeth Warren, the progressive senator from Massachusetts, will become the indisputable front runner (which she is not yet). In the fourth primary debate on October 15, she will face attacks from all sides reflecting this new status. Given her debate performances thus far, she will sustain the heightened scrutiny and come out stronger. This is not to say that Warren is already the Democratic candidate. Biden is still polling like a traditional Democratic primary front runner (Chart 9), while Warren has some clear weaknesses in electability, as reflected in her smaller lead over Trump in head-to-head polls in swing states. Nevertheless Warren is likely to become the front runner. Chart 9Biden Polling About Average Relative To Previous Democratic Primary Front Runners The recession call remains the U.S. election call. Two further considerations: Impeachment and removal of President Trump ensure a Democratic victory. There are hopes in some quarters that President Trump could be impeached and removed and yet his Vice President Mike Pence could go on to win the 2020 election, preserving the pro-business policy status quo. The problem with this logic is that Trump cannot be removed unless Republican opinion shifts. This will require an earthquake as a result of some wrongdoing by Trump. Such an earthquake will blacken Pence’s and the GOP’s name and render them toxic in the general election. Not to mention that Pence’s only act as president in the brief interim would likely be to pardon Trump and his accomplices. He would suffer Gerald Ford’s fate in 1976. Which means that a significant slide in Trump’s approval among Republicans will translate to higher odds of a Democratic win in 2020 and hence higher taxes and regulation, i.e. a hit to corporate earnings expectations. We expect this approval to hold up, but the market can sell off anyway because … The market is overrating the Senate as a check on Warren in the event she wins the White House. It is true that relative to Biden, Warren is less likely to carry the Senate. Democrats need to retain their Senate seat in Alabama, while capturing Maine, Colorado, and Arizona (or Georgia) in addition to the White House in order to control the Senate. Biden is more competitive in Arizona and Georgia than Warren. But this is a flimsy basis to feel reassured that a Warren presidency will be constrained. In fact, it is very difficult to unseat a sitting president. If the Democrats can muster enough votes to kick out an incumbent and elect an outspoken left-wing progressive from the northeast, they most likely will have mustered enough votes to take the Senate as well. For instance, unemployment could be rising or Trump’s risky foreign policy could have backfired. Chart 10Business Sentiment Threatens Trump Re-Election In our estimation the Democrats have about a 45% chance of winning the presidency, and Warren does not significantly reduce this chance. The resilient U.S. economy is Trump’s base case for success. But Trump’s trade policy and the global slowdown are rapidly eating away at the prospect that voters see improvement (Chart 10). This speaks to the constraint driving a ceasefire with China above, but it also speaks to the broader probability of policy continuity in the U.S. As Warren’s path to the White House widens, there is a clear basis for equities to sell off in the near term. Bottom Line: Trump’s approval among Republicans is a constraint on his removal via impeachment. But the status of the economy is the greater constraint. The recession call remains the election call. While we expect downside in the near term, we are still constructive on U.S. equities on a cyclical basis. War With Iran: The Oil Price Constraint The Senate will remain President Trump’s bulwark amid impeachment, notwithstanding the controversial news that Trump is moving forward with the withdrawal of troops from Syria, specifically from the so-called “safe zone” agreed with Turkey, giving Ankara license to stage a larger military offensive in Syria. This abandonment of the U.S.’s Kurdish allies at the behest of Turkey (which is a NATO ally but has been at odds with Washington) has provoked flak from Republican senators. However, it is well supported in U.S. public opinion (Chart 11). Trump is threatening to impose economic sanctions on Turkey if it engages in ethnic cleansing. The Turkish lira is the marginal loser, Trump’s approval rating is the marginal winner. The withdrawal sends a signal to the world that the U.S. is continuing to deleverage from the Middle East – a corollary with the return of focus on Asia Pacific. While the Iranians are key beneficiaries of this pivot, the Trump administration is maintaining maximum sanctions pressure on the Iranians. The firing of hawkish National Security Adviser John Bolton did not lead to a détente, as President Rouhani has too much to risk from negotiating with Trump. Instead the Iranians smelled U.S. weakness and went on the attack in Saudi Arabia, briefly shuttering 6 million barrels of oil per day. The response to the attack – from both Saudi Arabia and the U.S. – revealed an extreme aversion to military conflict and escalation. Instead the U.S. has tightened its sanctions regime – China is reportedly withdrawing from its interest in the South Pars natural gas project, a potentially serious blow to Iran, which had been hyping its strategic partnership with China. This reinforces the prospect for a U.S.-China ceasefire even as it redoubles the economic pressure on Iran. As long as the U.S. maintains the crippling sanctions on Iran, there is no guarantee that Tehran will not strike out again in an effort to weaken President Trump’s resolve. The fact that about 18% of global oil supply flows through the critical chokepoint of the Strait of Hormuz is Iran’s ace in the hole (Chart 12). It is the chief constraint on Trump’s foreign policy, as greater oil supply disruptions could shock the U.S. economy ahead of the election. Trump can benefit from minor or ephemeral disruptions but he is likely to get into trouble if a serious shock weakens the economy at this juncture. Chart 11U.S. Opinion Constrains Foreign Policy Chart 12Oil Price Constrains U.S. Policy Toward Iran An oil shock does not have to originate in Hormuz shipping or sneak attacks on regional oil infrastructure. Iran is uniquely capable of fomenting the anti-government protests that have erupted in southern Iraq. The restoration of stability in Iraq has resulted in around 2 million barrels of oil per day coming onto international markets (Chart 13). If this process is reversed through political instability or sabotage, it will rapidly push up against global spare oil capacity and exert an upward pressure on oil prices that would come at an awkward time for a global economy experiencing a manufacturing recession (Chart 14). Chart 13Iran's Leverage Over Iraq Chart 14Global Oil Spare Capacity Constrains Response To Crisis Bottom Line: Iran’s power over regional oil production is the biggest constraint on Trump’s foreign policy in the region, yet Trump is apparently tightening rather than easing the sanctions regime. The failure of the Abqaiq attack to generate a lasting impact on oil prices amid weak global demand suggests that Iran could feel emboldened. The U.S. preference to withdraw from Middle Eastern conflicts could also encourage Iran, while the tightening of the sanctions regime could make it desperate. An oil shock emanating from the conflict with Iran is still a significant risk to the global bull market. Brexit: The No-Deal Constraint The fifth and final constraint to discuss in this report pertains to the U.K. and Brexit. We do not consider the October 31 deadline a no-deal exit risk. Parliament will prevail over a prime minister who lacks a majority. Nevertheless the expected election can revive no-deal risk, especially if Boris Johnson is returned to power with a weak minority government. Chart 15U.K.: Public Opinion Constrains Parliament And No-Deal Brexit While parliament is the constraint on the prime minister, the public is the constraint on parliament. From this point of view, support for Brexit has weakened and the Conservative Party is less popular than in the lead up to the 2015 and 2017 general elections. The public is aware that no-deal exit is likely to cause significant economic pain and that is why a majority rejects no-deal, as opposed to a soft Brexit. Unless the Tory rally in opinion polling produces another coalition with the Northern Irish, albeit with Boris Johnson at the helm, these points make it likely that a no-deal Brexit will become untenable when all is said and done (Chart 15). If Johnson achieves a single party majority the EU will be more likely to grant concessions enabling him to get a withdrawal deal over the line. We remain long GBP-USD but will turn sellers at the $1.30 mark. Investment Implications The path of least resistance is for China’s stimulus efforts to increase – incrementally if trade tensions are contained, and sharply if not. This should help put a floor beneath growth, but the Q1 timing of this floor means that global risk assets face additional downside in the near term. We continue to recommend going long our “China Play” index. U.S.-China trade tensions should decline as President Trump looks to prevent higher unemployment ahead of his election. China has reason to follow through on small concessions to encourage Trump’s tactical trade retreat, but it does not face pressure to make new structural concessions. We expect a ceasefire – with some tariff rollback likely – but not a big bang agreement that removes all tariffs or deepens the overall bilateral economic engagement. Stay long our “China Play” index. We remain short CNY-USD on a strategic basis but recognize that a ceasefire presents a short term (maximum 12-month) risk to this view, so clients with a shorter-term horizon should close that trade. We are long European equities relative to Chinese equities as a result of the view that China will stimulate but that a trade ceasefire will leave lingering uncertainties over Chinese corporates. U.S. politics are highly unpredictable but constraint-based analysis indicates that while the House may impeach, the Senate will not remove. This, combined with Warren’s likely ascent to the head of the pack in the Democratic primary race, means that Trump remains favored to win reelection, albeit with low conviction (55% chance) due to a weak general approval rating and economic risks. The risk to U.S. equities is immediate, but should dissipate. The U.S. is rotating its strategic focus from the Middle East to Asia Pacific, which entails a continued rotation of geopolitical risk. However, recent developments reinforce our argument in July that Iranian geopolitical risk is frontloaded relative to the China risk. This is true as long as Trump maintains crippling sanctions. Iran may be emboldened by its successes so far and has various mechanisms – including Iraqi instability – by which it can threaten oil supply to pressure Trump. This is a tail risk, but it does support our position of being long EM energy producers.   Matt Gertken, Vice President Geopolitical Strategist mattg@bcaresearch.com Footnotes 1 Please see BCA Research, China Investment Strategy Weekly Report, “Mild Deflation Means Timid Easing,” October 9, 2019, available at cis.bcaresearch.com. 2 China knows that Trump wants to seal a deal prior to November 2020 to aid his reelection campaign, while Trump needs to try to convince China that he does not care about election, the stock market, or anything other than structural concessions from China. Hence the U.S. blacklisted several artificial intelligence companies and sanctioned Chinese officials in advance of the talks. The U.S. opened a new front in the conflict by invoking China’s human rights abuses in Xinjiang, which is also an implicit warning not to create a humanitarian incident in Hong Kong where protests continue to rage. These are pressure tactics but have not yet derailed the attempt to seal a deal in Q4.

Analysis on Turkey is available below. Highlights A dovish Fed or robust U.S. growth does not constitute sufficient conditions for a bull market in EM. China’s business and credit cycles are much more important factors for EM than those of the U.S. A recovery in the Chinese economy and global manufacturing is not imminent. The common signal reverberating from various financial markets is that the risks to the global business cycle are still skewed to the downside. Feature Current investor perceptions of emerging markets are mixed. Some expect EM to benefit greatly from low U.S. interest rates. These investors view even a partial trade deal between the U.S. and China as sufficient for EM to embark on a bull market. BCA’s Emerging Markets Strategy team disagrees with this narrative. We deliberated the significance of the U.S.-China confrontation to EM in our September 19 report; therefore, we will not go over this subject here. Rather, in this report we discuss some of the more common misconceptions surrounding EM currently, and infer what these mean for investment strategies. Perception 1: The share of resource sectors (materials and energy) in the EM equity benchmark has declined substantially. This along with the expanded role of consumers and consumer stocks (Alibaba, Tencent and Baidu) in EM economies and equity markets has made their share prices less exposed to the global trade cycle and commodities prices. Reality: It is true that in many EM bourses, the weight of consumer stocks has been growing. Nevertheless, their financial markets in general, and equity markets in particular, remain very sensitive to the global trade cycle and commodities prices. Chart I-1 illustrates that the aggregate EM equity index has historically been and continues to be strongly correlated with the global basic materials stock index. The latter includes mining, steel and chemical companies. Global materials stocks also exhibit a very strong correlation with Chinese banks’ share prices. Moreover, global materials stocks also exhibit a very strong correlation with Chinese banks’ share prices (Chart I-2). The rationale for the high correlation is that both mainland banks’ profits and global demand for basic materials are driven by a common factor: China’s business cycle. Chart I-1EM And Global Materials Stocks Move Together Chart I-2Chinese Bank And Global Materials Share Prices Are Highly Correlated For example, construction in China is contracting (Chart I-3), which entails both higher NPLs for Chinese banks and lower demand for basic materials. China accounts for about 50% of global consumption of industrial metals, cement and many other basic materials. Finally, EM ex-China bank stocks also correlate strongly with global basic materials share prices. The basis is as follows: Many emerging economies export raw materials, and commodities price fluctuations impact their business cycle, exports and exchange rates. Chart I-3China: Construction Activity Is Contracting Chart I-4High-Yielding EM: Currencies And Local Bond Yields Historically, in high-yielding EM markets, currency depreciation has led to higher interest rates and lower bank share prices, and vice versa (Chart I-4). Lately, EM bond yields have not risen in response to EM currency depreciation. However, we believe this correlation will soon be re-established if EM currencies continue drifting lower.  In short, China’s money/credit cycles drive not only the mainland’s business cycle, banking profits and NPLs, but also global trade and commodities prices. The latter two - via their impact on exchange rates and in turn interest rates - have historically explained credit and domestic demand cycles in high-yielding EM. Perception 2:  EM stocks are a high-beta play on the S&P 500, i.e., EM equities outperform when the S&P 500 rallies, and vice versa. Reality: Since 2012, the beta for EM equity versus the S&P 500 has often been below one (Chart I-5). Furthermore, since 2012, EM share prices often failed to outpace their DM peers during global equity rallies. Indeed, EM relative equity performance versus DM, as well as the EM ex-China currency total return index, have been closely tracking the relative performance of global cyclicals versus global defensive stocks (Chart I-6). Chart I-5EM Equities Beta To The S&P 500 Chart I-6Global Cyclicals-To-Defensives Equity Ratio And EM   In short, EM equities and currencies have been, and will remain, sensitive to the global business cycle rather than the S&P 500. Since 2012, the latter has - on several occasions - decoupled from the global manufacturing and trade cycles. Perception 3:  EM stocks, currencies and fixed-income markets are very sensitive to U.S. interest rates. Hence, a dovish Fed will lead to EM currency appreciation.  Reality: Chart I-7 reveals that EM currencies, total returns on EM local currency bonds in U.S. dollar terms and EM sovereign credit spreads do not exhibit a strong relationship with U.S. Treasury yields. U.S. interest rate expectations have a much smaller impact on EM financial markets than commonly perceived by the investment community.  Overall, U.S. interest rate expectations have a much smaller impact on EM financial markets than commonly perceived by the investment community.  Chart I-7EM And U.S. Bond Yields: No Stable Correlation Chart I-8China Cycle And EM Stocks Led U.S. Bond Yields On the contrary, the declines in U.S. bond yields in both 2015/16 and in 2018/19 were due to the growth slowdown that emanated from China/EM. The top panel of Chart I-8 illustrates that Chinese import growth rolled over in December 2017, yet U.S. bond yields rolled over in October 2018. What is more, EM share prices have been leading U.S. bond yields in recent years, not the other way around (Chart I-8, bottom panel). Perception 4:  If the U.S. avoids a recession, EM risk assets will recover. Chart I-9EM Profits Are Driven By Chinese Not U.S. Business Cycle Reality: EM per-share earnings contracted in 2012-2014 and in 2019, despite reasonably robust growth in U.S. final demand (Chart I-9, top panel). This suggests that even if the U.S. economy avoids a recession, that will not be a sufficient condition to be bullish on EM. EM corporate profits are highly driven by China’s business cycle. The bottom panel of Chart I-9 illustrates that mainland domestic industrial orders have been the key driver of EM corporate profit cycles since 2008. Perception 5:  EM equities, fixed-income markets and currencies are cheap. Reality: EM stocks are not cheap. They are fairly valued. Equity sectors with very poor fundamentals have very low multiples. Hence, they are “cheap” for a reason. These include Chinese banks, state-owned enterprises in various countries and resource companies. Equity segments with robust fundamentals are overpriced. Given that Chinese banks, state-owned enterprises in various countries, resource companies, and cyclical businesses have very large market caps, EM market-cap based equity valuation ratios are low – i.e., they appear cheap.  To remove the impact of these large market cap segments, we constructed and have been publishing the following valuation ratios: median, 20% trimmed mean and equal-sub-sector weighted (Chart I-10). Each of these is calculated based on the average of trailing and forward P/E ratios, price-to-book value, price-to-cash earnings and price-to-dividend ratios. EM equities relative to DM are not cheap either. Chart I-11 demonstrates the same ratios – median, 20% trimmed-mean and equal-sub-sector weighted values for EM versus DM. Chart I-10EM Equities Are Not Cheap Chart I-11Relative To DM EM Stocks Are Not Cheap Further, when valuations are not at extremes as in the case of EM equities at the moment, the profit cycle holds the key to share price performance over a 6 to 12-month horizon. EM earnings are presently contracting in absolute terms, and underperforming DM EPS. Two currencies that offer value are the Mexican peso and Russian ruble. Chart I-12EM Local Yields Are Low In Absolute Terms And Relative To U.S. In the fixed-income space, EM local bond yields are very low in absolute terms and relative to U.S. Treasury yields (Chart I-12). EM sovereign and corporate spreads are not wide either. As to exchange rates, the cheapest currencies are those with the worst fundamentals, such as the Argentine peso, Turkish lira and South African rand. The majority of other EM currencies are not very cheap. Two currencies that offer value are the Mexican peso and Russian ruble. Yet foreign investors are very long these currencies, and a combination of lower oil prices and portfolio outflows from broader EM will weigh on these exchange rates as well. Takeaways And Investment Strategy Chart I-13EM Currencies And Industrial Metals Prices EM risk assets and currencies exhibit the strongest correlation with global trade and commodities prices. Chart I-13 indicates that the EM ex-China currency total return index closely tracks commodities prices. This corroborates the messages from Chart I-1 on page 1 and Chart I-6 on page 4.  China’s business and credit cycles are much more important for EM than those of the U.S. A dovish Fed or strong U.S. growth are not sufficient reasons to bet on an EM bull market. A recovery in the Chinese economy and global manufacturing is not imminent. Individual EM countries’ domestic fundamentals such as return on capital, inflation, banking system health, competitiveness and politics drive individual EM performance. On these accounts, the outlook varies among EM. Readers can find analyses on specific EM economies in our Countries In-Depth page. Asset allocators should continue underweighting EM stocks, credit and currencies versus their DM counterparts.  Absolute-return investors should outright avoid EM, or trade them on the short side. Within the EM equity space, our overweights are Mexico, Russia, Central Europe, Korea ex-tech, Thailand and the UAE. Our underweights are South Africa, Indonesia, Philippines, Hong Kong, Turkey and Colombia. The path of least resistance for the U.S. dollar is up. Continue shorting the following basket of EM currencies versus the dollar: ZAR, CLP, COP, IDR, MYR, PHP and KRW. We are also short the CNY versus the greenback. As always, the list of our country allocations for local currency bonds and sovereign credit markets is available at the end of our reports – please refer to page 16. Take Cues From These Markets We suggest investors take cues from the following financial market signals. They are unequivocally sending a downbeat message for global growth and risk assets: The ratio between Sweden and Swiss non-financial stocks in common currency terms is heading south (Chart I-14). Swedish non-financials include many companies leveraged to the global industrial cycle, while Swiss non-financials are dominated by defensive stocks. Hence, the persistent decline in this ratio presages a continued deterioration in the global industrial sector. Where is the next defense line for this ratio? To reach its 2002 and 2008 nadirs, it will need to drop by another 10%. In the interim, investors should maintain a defensive posture. Chart I-14A Message From Swedish And Swiss Equities Chart I-15A Breakdown In The Making? U.S. FAANG stocks appear to be cracking below their 200-day moving average. The relative performance of global cyclical versus global defensive stocks is relapsing below the three-year moving average that served as a support last December (Chart I-15). U.S. FAANG stocks appear to be cracking below their 200-day moving average (Chart I-16). If this support gives, the next one will be about 17% below current levels. Finally, U.S. high-beta share prices are on the verge of a breakdown (Chart I-17). The next technical support is 10% below current levels. Chart I-16FAANG Are On The Support Line Chart I-17U.S. High-Beta Stocks Are On The Edge Bottom Line: The common message reverberating from these financial markets corroborates our fundamental analysis that a global business cycle recovery is not imminent, and that global risk assets in general, and EM financial markets in particular, are at risk of selling off further. Arthur Budaghyan Chief Emerging Markets Strategist arthurb@bcaresearch.com   Turkey: Is The Mean-Reversion Rally Over? Turkish financial markets have rebounded to their respective falling trend lines (Chart II-1). Are they set to break out or is a setback looming? Chart II-1Back To Falling Trend Chart II-2TRY Is Cheap Pros The economy has undergone a considerable real adjustment and many excesses have been purged: The current account balance has turned positive as imports have collapsed. Going forward, lower oil prices are likely to help the nation’s current account dynamics. The lira has become cheap (Chart II-2).  According to the real effective exchange rate based on unit labor costs, the currency is one standard deviation below its fair value. Core and headline inflation have fallen, allowing the central bank to cut interest rates aggressively. However, the exchange rate still holds the key: if the currency depreciates anew, local bonds yields will rise and the ability of the central bank to reduce borrowing costs further will diminish. Finally, private credit and broad money growth have decelerated substantially and are contracting in inflation-adjusted terms (Chart II-3). Chart II-3Money & Credit Have Bottomed Chart II-4Banks Have Been Aggressively Buying Government Bonds The recent gap between broad money and private credit growth has been due to commercial banks buying government bonds (Chart II-4). When a commercial bank purchases a security from non-banks, a new deposit/new unit of money supply is created. Banks’ purchases of government bonds en masse have capped domestic bond yields. However, if pursued aggressively, such monetary expansion could weigh on the currency’s value.   Cons Presently, potential sources of macro vulnerability in Turkey are: Foreign debt obligations (FDOs) – which are calculated as the sum of short-term claims, interest payments and amortization over the next 12 months – are at $168 billion, which is sizable. The annual current account surplus has reached only $4 billion and is sufficient to cover only 2.5% of FDOs, assuming the capital and financial account balance will be zero. Clearly, Turkey needs to both roll over most of its foreign debt coming due and attract foreign capital to finance a potential expansion in its imports if its domestic demand is to recover. Critically, $20 billion of net FX reserves, excluding gold, swap lines with foreign central banks and net of domestic banking and non-banking corporations’ foreign exchange deposits, are not adequate either to cover foreign debt obligations. Even though headline and core inflation measures have fallen, wage inflation remains rampant (Chart II-5). If wage inflation does not drop substantially very soon, rapidly rising unit labor costs will feed into inflation leading to negative ramifications for the exchange rate. This is especially crucial in Turkey given President Erdogan has undermined the central bank’s credibility and is resorting to populist measures to revive his popularity. Finally, Turkish banks remain under-provisioned. Currently, the banking regulator is requiring banks to boost their non-performing loans (NPL) ratio to 6.3% of total loans.This a far cry from the 2001 episode when the NPL ratio shot up to 25% (Chart II-6).   Even though interest rates rose much more in 2001 than last year, the private credit penetration in the economy was very low in the early 2000s. A higher credit penetration usually implies weaker borrowers have borrowed money and heralds a higher NPL ratio. Typically, following a credit boom and bust, it is natural for the NPL ratio to exceed 10%. We do not think Turkish banks stocks, having rallied a lot from their lows, are pricing in such a scenario. Chart II-5Surging Wages Are A Risk Chart II-6NPL Ratio Is Unrealistic Investment Recommendation We recommend both absolute-return investors and asset allocators not to chase Turkish financial markets higher. Renewed market volatility lies ahead. Given we expect foreign capital outflows from EM, Turkish companies and banks will encounter difficulties in rolling over their external debt and attracting foreign capital into domestic markets. This will produce a new downleg in the exchange rate. In turn, currency depreciation will weigh on performance of local bonds as well as sovereign and corporate credit. Stay underweight.   Andrija Vesic, Research Analyst andrijav@bcaresearch.com Footnotes Equities Recommendations Currencies, Credit And Fixed-Income Recommendations

Business confidence peaked in March 2018 and has been in a freefall ever since, with the steepest drop taking place in recent months as the Sino-American trade war has re-escalated (CEO confidence shown inverted, top panel). Moreover, there is mounting evidence that the trade tensions are further infecting the economy beyond manufacturing including services and the consumer. Using data from the Conference Board’s Consumer Confidence survey and from the University of Michigan Sentiment survey the chart shows that consumer intentions to buy large household durable goods (shown inverted, second panel), cars (shown inverted, third panel) and homes (shown inverted, bottom panel), all have taken a massive hit of late. Historically, all three survey measures have been excellent leading indicators of the labor market and the current message is to expect a rise in the unemployment rate in coming months. Bottom Line: While we are on the sidelines on the defensive/cyclical portfolio bent we stand ready to move to a defensive over cyclical preference. Once our S&P software trailing stop gets triggered, which will move this heavyweight tech subgroup to neutral, then the broad tech sector will shift to underweight and our defensive/cyclical bent to overweight. Stay tuned.