Global

Special Report

Highlights The End Of APP?: Economic growth in the euro area has lost momentum, but it is not clear that an extended period of below-trend growth is unfolding. With most measures of spare capacity showing a lack of it, the ECB must still move forward with its plans to begin removing policy accommodation. Policy Choices: If the ECB downgrades its growth and inflation forecasts next month, delaying the end of the APP into 2019 is unlikely, as is altering the country weightings within the APP portfolio. More plausible options include pushing out forward guidance on future rate hikes, extending the maturity of the existing bond holdings, or introducing a new TLTRO to support lending. Impact On European Bonds & The Euro: The ECB is most likely to take a less hawkish slant in December, but will not signal any rapid move to begin hiking rates. This outcome will be bearish for the euro, but only neutral at best for overvalued European government bonds. Feature For the European Central Bank (ECB), the countdown is on to the December policy meeting, when a final decision will have to be made on the end of the Asset Purchase Program (APP). The central bank has been signaling throughout 2018 that net new APP bond purchases will stop at the end of the year, with a potential interest rate increase coming in September 2019 at the earliest. That decision on APP, however, will be conditional on the ECB remaining confident in its forecast that inflation will sustainably return to the target of "just below" 2%. Slumping European economic growth in 2018 means that the ECB's forecasts may prove to be too optimistic. This is especially true given the risks to growth and financial stability stemming from Italy's fiscal policy debate with the European Union, softening Chinese demand for European exports, and the uncertainties related to U.S. trade protectionism and the final U.K.-E.U. Brexit deal. Some pundits are even suggesting that the ECB may be forced to extend the APP program beyond December - or look for other ways to prevent a tightening of monetary conditions - even with headline inflation and wage growth having picked up across most countries. Against this increasingly muddled backdrop, what can the ECB credibly announce in December? In this Special Report, jointly published by BCA's Global Fixed Income Strategy and Foreign Exchange Strategy services, we discuss the state of the euro area economy and then consider the ECB's next potential policy moves, with ramifications for European bond yields and the euro. Our conclusion is that there are a few policy tools available to the ECB in case of a prolonged slump in growth, without having to bring on the operational difficulties from extending the APP beyond December. Such a "dovish" shift would be bearish for the euro but neutral, at best, for European government bonds which remain deeply overvalued. ECB Policy Dilemma: Slowing Growth Vs. Accelerating Inflation At last month's monetary policy meeting, ECB President Mario Draghi noted that the slowing economy was merely returning to trend (or potential) growth from an unsustainably fast pace in 2017 that was fueled by strong export demand. Looking at the broad swath of euro area economic data, Draghi's relatively optimistic assessment is not far off the mark. The euro zone has seen a clear loss of economic growth momentum since the start of the year (Chart 1). The initial read on real GDP for the third quarter, released last week, showed a deceleration to a below-potential quarterly growth pace of 1.7%. The manufacturing purchasing managers index (PMI) has fallen from a peak of 61 in December 2017 to 52 in October, mirroring a -1% decline in the OECD's leading economic indicator for the region. Chart 1A European Growth Slump, Not Yet A Downtrend Yet not all the economic news has been that weak. Both consumer and business confidence remain at elevated levels according to the European Commission (EC) surveys, consistent with above-trend real GDP growth (bottom two panels). Even though exports have weakened substantially from the booming pace in 2017 - largely due to China's slowing growth - the EC survey on firms' export order books remains at robust levels and overall export growth has rebounded of late (Chart 2). The current conditions component of the euro area ZEW index has also ticked higher (top panel), as has the bank credit impulse (bottom panel). Chart 2Not All The Economic News Is Bad The bigger issue for the ECB is that the recent cooling of growth comes at a time when, by almost all measures, there is little economic slack in the euro area. Capacity utilization is running at an 11-year high of 84%, while the output gap is effectively closed according to estimates from the IMF (Chart 3). Chart 3No Spare Capacity In Europe With that gap projected to turn positive in 2019, core inflation in the euro zone should be expected to drift higher. Yet core inflation now remains stuck around 1%, well below the headline inflation figure of 2% that has been heavily influenced by past increases in energy prices (bottom panel). The labor market is sending signals that the current period of low euro area inflation may be turning around. The unemployment rate for the entire region fell to a 10-year low of 8.1% in September, well below both the ECB's latest 2018 forecast and the OECD's estimate of the full employment NAIRU (Chart 4). This tightening labor market is a broad-based phenomenon across the euro area, with nearly 80% of countries in the region having an unemployment rate below NAIRU (middle panel).1 The last two times there was such a broad-based decline in unemployment in the region, in 2001-02 and 2006-07, a significant tightening of monetary policy was required as measured by a simple Taylor Rule. Chart 4Broad-Based Labor Market Strength Already, the tightening labor market is starting to put upward pressure on labor costs. The annual growth in wages & salaries accelerated to just over 2% in the second quarter of 2018. Similar to the fall in unemployment rates, the faster wage growth has also been widely seen throughout the region, with nearly three-quarters of euro area countries showing faster wage growth from one year ago (bottom panel). The mix of slowing growth momentum with some inflationary pressures can be seen in our ECB Monitor, which measures the cyclical pressures to tighten or ease monetary policy in the euro area. The Monitor had been signaling a need for tighter policy for most of the past two years, but has now fallen back to levels consistent with no change in policy (Chart 5). When breaking down the Monitor into its inflation and growth components, the latter has fallen the most. The inflation components remain in the "tight money required" zone above the zero line. Chart 5Our ECB Monitor Says 'Do Nothing' Looking across the balance of the euro area data, President Draghi's assessment that the recent economic weakness is not the beginning of a sustained move to below-trend growth is justified. Given the broad evidence pointing to a lack of excess capacity across the euro area economy, it will take a much bigger growth slump before the ECB can shift to a more dovish policy bias. The critical series to monitor will be business confidence, capital spending and export orders. All are at risk of downshifting due to slowing global trade activity and sluggish Chinese demand. BCA's China experts continue to have doubts that the Chinese government will undertake any typical initiatives to stimulate demand, like interest rate cuts or fiscal spending, given worries about high domestic debt levels. Without the impetus from strong Chinese import demand boosting euro area exports, the current tightness of euro area labor markets, and uptrend in wage growth, may be at risk of a reversal, as we discussed in a recent Special Report.2 Bottom Line: Economic growth in the euro area has lost momentum, but it is not clear that an extended period of below-trend growth is unfolding. With most measures of spare capacity showing a lack of it, the ECB must still move forward with its plans to begin removing policy accommodation. What Tools Are Available For The ECB? Net-net, when looking at the broad balance of growth and inflation data at the moment, there is not yet enough evidence to suggest that the ECB needs to back away from its current plans to end net new APP purchases in December. That does not mean that the ECB would not consider changes to its total mix of monetary policy measures. The ECB has treated the APP, which began in 2015, as a "deflation fighting tool" during a period when there was excess capacity and very low inflation in the euro area. That is no longer the case, so it will be difficult for the ECB Governing Council to argue in December that new APP purchases are still necessary. It would take a substantial downward adjustment to the ECB growth and inflation forecasts, with a subsequent upward revision to the expectations for the unemployment rate, for the ECB to reconsider the plans to stop new bond purchases at year-end. Yet the ECB has also made it clear that interest rate hikes will not happen soon after the APP purchases end. Going back over the entire 20-year history of the ECB, there have only been three tightening episodes through rate hikes: 1999-2000, 2003-07 and 2011. In all three cases, what prompted the rate hikes was a period of broad-based increases in euro zone inflation that followed a period of equally broad-based euro zone economic growth. This can be seen in Chart 6, which shows "diffusion indices", or breadth across countries, for euro area real GDP and inflation. A higher number means that a greater percentage of individual nations is experiencing faster growth or inflation, and vice versa. During those three previous tightening cycles, the diffusion indices all reached elevated levels for growth and, more importantly, inflation. With more countries enjoying the upturn, the ECB could be more confident in seeing the need for interest rate increases to cool off demand to prevent an inflation overshoot. Chart 6No Need For ECB Rate Hikes Anytime Soon At the moment, the diffusion indices are quite low, suggesting that few countries are witnessing accelerating growth or inflation. This means that there is no pressure for the ECB to move up its current dovish guidance to the markets about the timing of the first rate hike in late 2019. That also means that there is a risk that the ECB is forced to consider options for providing additional monetary accommodation if there was a large enough downgrade to its growth and inflation forecasts. If the ECB were to indeed lower its growth forecasts in December and consider additional easing options, there are only four plausible options at their disposal: 1) Extending the APP purchases beyond December, either at the current pace of €15bn/month or a slower pace between €5-10bn/month Extending the APP into 2019 is the least likely choice because the ECB is already close to some of the self-imposed constraints on its government bond holdings. The ECB has set a limit of owning no more than 33% of an individual country's allowable government bonds, with maturities of between 1-31 years. Right now, the ECB owns about 31% of all eligible German government debt (Chart 7), and would breach that 33% level sometime in the first half of 2019 if the current pace of buying was maintained without any increase in German bond issuance (i.e. smaller budget surpluses).3 A similar outcome would also occur for smaller bond markets, like the Netherlands and Finland (bottom panel). Chart 7ECB Will Hit Country Issuer Limits If Current APP Is Maintained Of course, this is a self-imposed rule by the ECB that can easily be changed. That already occurred back in 2016 when the ECB allowed the purchase of bonds below the deposit rate as part of its APP operations. This meant that the ECB would buy bonds with negative yields, essentially guaranteeing a loss assuming that the bonds were held to maturity. Yet given how much emphasis the ECB has placed on abiding by the issuer limits, we think the ECB would consider other policy choices before raising them. 2) Changing the composition of the APP portfolio Changing the mix of bonds within the APP portfolio is a more likely option, but even this has its limits. The ECB could choose to buy more corporate bonds or covered bonds, but those are less liquid markets where there is arguably more evidence that ECB buying has impacted market functionality. The ECB may be reluctant to take on more credit risk in its bond portfolio, as well. At the country level, the ECB could choose to move away from using its Capital Key weightings to determine the allocation of its bond purchases by country. In the current heated political atmosphere in Europe, however, with the populist Italian government in a very public battle with the E.U. over its 2019 budget, the ECB will not want to be seen as favoring any country more than another by buying more government bonds in places like Italy or Spain over Germany and France. That can already be seen in how bond purchases have been allocated in 2018, with purchases sticking closer to the Capital Key weightings in Italy and France from the larger weightings seen in 2017 (Charts 8 & 9). Chart 8The ECB Capital Key ... Chart 9... Is Not Always Adhered To A more likely reallocation of bond holdings could occur within each country by adjusting the maturities held within the ECB's portfolio. Following the template of the Fed's 2012 "Operation Twist", the ECB could aim to sell shorter-dated bonds in exchange for longer-maturity debt, thereby exacting a flattening influence on government yield curves. There is scope for that in Germany, where the weighted-average-maturity (WAM) of the ECB's bond holdings has decline by 18 months since peaking in late 2015 (Chart 10). Large declines in WAW have also occurred for Spanish, Italian and Portuguese bonds owned by the ECB, if policymakers were willing to take on more duration risk in the Periphery. Chart 10The ECB Has Room To Extend Its APP Maturities 3) Extend forward guidance on the first rate hike The easiest option for the ECB in the event of a downgrade of its growth/inflation projections is to simply extend the forward guidance on the timing of the first interest rate hike. Right now, our Months-to-Hike indicators, which measure the time until a full rate hike is discounted in the European Overnight Index Swap (OIS) curve, are discounting a hike of 10bps by November 2019 and a hike of 25bps by May 2020 (Chart 11). The ECB could easily signal that any rate hike, of any size, would not occur before the latter half of 2020 if an additional easing move was required. This would mostly likely result in lower bond yields and a weaker euro, all else equal, helping easy monetary conditions in the euro area. Chart 11Extending Forward Guidance Is An Option 4) Introduce a new Targeted Long-Term Lending Operation (TLTRO) One final intriguing option for an ECB policy ease would be the introduction of another TLTRO. The last such targeted lending program occurred in 2016, but the first wave of the much larger program that began in 2014 has already started to run off the ECB's balance sheet. This is the most effective way to get European banks to extend credit to borrowers at lower interest rates, since the banks would be able to fund that borrowing via the TLTRO at a rate lower than market rates. President Draghi did note last month that some members of the Governing Council brought up the idea of a new TLTRO at the ECB's policy meeting, and some well-known investment banks have recently discussed the implications of a new operation. In our view, a new TLTRO is the most effective way for the ECB to provide stimulus via lower private borrowing rates. It would also help offset any negative ramifications of the reduction of the ECB's balance sheet from the expiration of prior TLTROs. This would likely only happen, though, if there was evidence that the credit channel was impaired in the euro area. The previous TLTROs occurred after a period when banks were tightening credit standards, corporate borrowing rates and credit spreads were widening, European bank stocks were falling and European bank lending standards were becoming more restrictive (Chart 12). Chart 12A New TLTRO? Watch Lending Standards Today, bank stocks are falling and corporate bond yields/spreads are low but slowly rising, while European banks are actually easing lending standards according to the ECB's Q3 Bank Lending Survey. If the latter were to flip into the "tightening standards" zone, without any rebound in European bank shares or decline in corporate borrowing rates, the ECB could be tempted to go down the TLTRO route once again. Bottom Line: If the ECB downgrades its growth and inflation forecasts next month, delaying the end of the APP into 2019 is unlikely, as is altering the country weightings within the APP portfolio. More plausible options include pushing out forward guidance on future rate hikes, extending the maturity of the existing bond holdings, or introducing a new TLTRO to support lending. Likely ECB Options & Investment Implications In our view, the most realistic outcomes for the December ECB meeting can be boiled down to two decisions, conditional on how the ECB's economic forecasts are presented: 1) Unchanged growth & inflation forecasts: The ECB will signal the end of new APP bond purchases at the end of December, while maintaining the current forward guidance on rate hikes that no move will occur until at least September 2019. 2) Downgraded growth & inflation forecasts: The ECB will signal the end of new APP bond purchases at the end of December, but will also push out forward guidance on the first rate hike to at least sometime in mid-2020. In the latter scenario, the ECB could also consider two other options: extending maturities within its German bond holdings, or announcing a new TLTRO. We think that the ECB will wait to see how financial markets absorb the end of new APP buying before considering any move on maturity extension. At the same time, the ECB would signal that a TLTRO is a possibility if lending standards deteriorate and borrowing rates climb higher. While the ECB has talked a lot about how they will continue to reinvest the proceeds of maturing bonds in its portfolio, similar to what the Federal Reserve did after it ended its QE buying, the bigger impact on bond yields will come from a worsening of the supply/demand balance for European bonds. The ECB has been buying amounts greater than the entire net bond issuance of all euro area governments since the APP started in 2015, which has created a scarcity of risk-free sovereign debt for private investors. The result: extremely low bond yields, with a negative term premium (Chart 13). Reduced ECB buying will result in more bonds that have to be purchased by private investors, and a less negative term premium, going forward. Chart 13Bund Term Premium Unwind? How high euro area bond yields eventually go will then be determined by more traditional factors, like inflation expectations and the expected path of ECB rate hikes. Going back to the ECB's previous tightening cycles over its existence, actual rate hikes did now occur before inflation expectations - as measured by 5-year CPI swaps, 5-years forward - rose above 2% (Chart 14). Those inflation expectations are now 32bps below that level, and the ECB will not begin to shift to less dovish forward guidance unless the markets begin to discount more stable inflation close to the ECB's "near 2%" target. Chart 14Not Enough Inflation (Yet) To Justify Rate Hikes Dovish guidance on future ECB rate hikes will continue to widen the U.S.-Europe interest rate differentials that have helped weaken the euro versus the U.S. dollar in 2018 (Chart 15). This will continue to put downward pressure on EUR/USD cross, particularly with neutral momentum and positioning indicators suggesting that the euro is not yet oversold (bottom panel). Chart 15Likely ECB Actions Are Euro-Bearish Bottom Line: The ECB is most likely to take a less hawkish slant in December, but will not signal any rapid move to begin hiking rates. This outcome will be bearish for the euro, but only neutral at best for overvalued European government bonds. Robert Robis, CFA, Senior Vice President Global Fixed Income Strategy rrobis@bcaresearch.com 1 Since not every country in the euro area is also part of the OECD, we could only use 14 of the 19 countries in the euro area in the indicator shown in the middle panel of Chart 5. 2 Please see BCA Foreign Exchange Strategy/Global Fixed Income Strategy Special Report, "Will Rising Wages Cause An Imminent Change In Policy Direction In Europe And Japan?, dated October 6th 2018, available at fes.bcaresearch.com and gfis.bcaresearch.com. 3 The ECB does allow the purchase of both federal government bonds, as well as the debt of government agencies and supranationals, as part of its APP. For our projections, we have assumed that of the €15bn in net new bonds that the ECB buys each month, 82% are debt issued by government-related entities (i.e. 18% goes to credit instruments like corporate bonds and covered bonds), with 10% of those government purchases going to supras. From that reduced number, we assume anywhere from 10-30% of purchases go to agencies, depending on the country. For the sake of simplicity, we also assume a pace of net government bond issuance in line with that seen over the past year, rather than make specific assumptions on changes in individual country budget deficits.

Highlights Investors are worrying too much about the things that caused the global financial crisis, and not enough about those that could cause the next downturn. Despite the recent patch of soft data, the U.S. housing market is in good shape. Go long homebuilders relative to the S&P 500. Imbalances in the corporate debt market have increased, but are not severe enough to generate systemic economic distress. U.S. rates will need to rise quite a bit more than the market anticipates before the economy slows by enough to force the Fed to back off. The combination of a stronger dollar and inadequate Chinese stimulus will continue to pressure emerging markets. Even Brazil's pro-capitalist new president may not be able to reverse the country's bleak fiscal dynamics. Our MacroQuant model, which predicted the correction, points to further near-term downside risk for global equities. The cyclical (12-to-18 month) outlook looks much better, however. Feature The Market's Maginot Line One of the most reliable ways to make money as an investor is to figure out the market's collective biases and trade against them. Behavioral economists have long noted that people tend to assign too much weight to recent experience in taking decisions. As a result, in finance, as in military strategy, there is a constant temptation to fight the last war. The last war policymakers waged was against the scourge of deflation that followed the housing bust and financial crisis. For much of the past decade, investors have held a magnifying glass over anything that could possibly resemble the conditions that led up to the Global Financial Crisis. While such behavior is understandable, it is misplaced. History suggests that both lenders and borrowers tend to act prudently for years, if not decades, following major financial crises. Mistakes are still made, but they are different mistakes. People overcompensate. They obsess about the past rather than focusing on the future. U.S. Housing Is Okay There is no denying that the U.S. housing market has softened this year (Chart 1). Housing starts, building permits, and home sales have all fallen. Residential investment has subtracted from GDP growth over three consecutive quarters. Chart 1Housing Has Been A Drag On The U.S. Economy This Year There is little mystery as to why the housing market has been on the back foot. The Trump tax bill capped the deduction on state and local property taxes, while reducing the amount of mortgage debt on which homeowners can deduct interest payments from $1 million to $750,000. This had a negative effect on housing activity, especially in high-tax Democrat-leaning states with elevated real estate prices. More importantly, mortgage rates have risen by over 100 basis points since last August. Chart 2 shows that home sales and construction almost always decline after mortgage rates rise. In this respect, the weakness in housing activity is reminiscent of the period following the taper tantrum, when housing activity also slowed sharply. Chart 2No Mystery Why U.S. Housing Has Been Weak... We do not expect mortgage rates to fall from current levels. But they are not going to rise at the same pace as they have over the past year. Thus, while the headwinds from higher financing costs will not disappear, they will abate to some extent. Fundamentally, the housing market is on solid ground (Chart 3). Mortgage rates are still well below their historic average. Home prices have risen considerably, but do not appear excessively stretched compared to rents or incomes. Unlike in 2006, the home vacancy rate is near its historic lows. Residential investment stands at only 3.9% of GDP, compared with a peak of 6.7% of GDP in the second half of 2005. The average age of the residential capital stock has risen by nearly five years since 2006, the largest increase since the Great Depression. With household formation rebounding briskly from its post-recession lows, homebuilders are still arguably not churning out enough new homes. Chart 3A...But Fundamentals Are Still In Good Shape (I) Chart 3B...But Fundamentals Are Still In Good Shape (II) Mortgage lenders have learned from past mistakes (Chart 4). While lending standards have eased modestly over the past 4 years, underwriting standards have remained high. The average FICO score for new borrowers is more than 40 points above pre-recession levels. The Urban Institute Housing Credit Availability index, which measures the percentage of home purchase loans that are likely to default over the next 90 days, is at reassuringly low levels. This is particularly the case for private-label mortgages, whose default risk has hovered at just over 2% during the past few years, down from a peak of 22% in 2006. Moreover, banks today hold much more high-quality capital than in the past, which gives them additional space to absorb losses (Chart 5). Chart 4Lending Standards Have Been Tight, But Are Starting To Loosen Chart 5U.S. Banks Are Well Capitalized With all this in mind, we are initiating a new strategic trade to go long U.S. homebuilders relative to the S&P 500.1 Corporate Debt: How Big Are The Risks? Unlike household debt, U.S. corporate debt has risen over the past decade and now stands at a record high level as a share of GDP. The quality of the lending has also been less than pristine, as evidenced by the proliferation of "covenant lite" loans. The interest coverage ratio for the economy as a whole - defined as the volume of profits corporations generate for every dollar of interest paid - is still above its historic average (Chart 6). However, this number is skewed by a few mega-cap tech companies that hold a lot of cash and have little debt. Chart 6Interest Coverage Looks Relatively High My colleague Mark McClellan, who writes our monthly Bank Credit Analyst publication, has shown that the interest coverage ratio for companies comprising the Bloomberg Barclays index would drop close to the lows of the Great Recession if interest rates were to rise by a mere 100 basis points across the corporate curve. The damage would be far worse if profits also fell by 25% in this scenario.2 While the corporate debt market has become increasingly frothy, it does not pose an imminent danger to the economy. There are several reasons for this. First, while U.S. corporate debt is high in relation to the past, it is still quite low in comparison with many other economies (Chart 7). The ratio of corporate debt-to-GDP, for example, is 30 percentage points higher in the euro area. This suggests that U.S. businesses still have the "carrying capacity" to take on additional debt. Chart 7U.S. Corporate Debt Is Not That High By Global Standards Second, the average maturity of U.S. corporate debt has risen over the past decade, with an increasing share of companies opting for fixed over floating-rate borrowings. This implies that it will take a while for the effect of higher rates to make their way through the system. Third, and perhaps most importantly, corporate bonds are generally held by non-leveraged investors such as pension funds, insurance companies, and ETFs. Bank loans account for only 18% of nonfinancial corporate-sector debt, down from 40% in 1980 (Chart 8). The share of leveraged loans held by banks has declined from about 25% a decade ago to less than 10% today. Chart 8Banks Have Reduced Their Exposure To The Corporate Sector Tellingly, we already had a dress rehearsal for what a corporate debt scare might look like. Credit spreads spiked in 2015. Default rates rose, but the knock-on effects to the financial system were minimal (Chart 9). This suggests that corporate America could withstand quite a bit of monetary tightening without buckling under the pressure. Chart 9The 2015 Debt Scare Did Not Topple The Economy Government Debt: No Worries... Yet If the risks posed by both the housing market and corporate debt market are contained, what about the risks posed by soaring government debt? The long-term fiscal outlook is certainly bleak, but the near-term risks are low.3 President Trump's tweets aside, the U.S. has an independent central bank which has been able to keep inflation expectations well anchored. The U.S. private sector is also running a financial surplus at the moment, meaning that it earns more than it spends (Chart 10). Not only does this make the economy more resilient, it also provides the government with additional savings with which to finance its fiscal deficit. Chart 10The U.S. Private Sector Is A Net Saver The private sector's financial balance will deteriorate over the next two years as household savings decline and corporate investment rises. This will put upward pressure on Treasury yields. However, if rising yields are reflective of stronger aggregate demand, this is unlikely to derail the economy. When Things Break Recessions are usually caused when the Fed raises rates by enough to undermine spending on interest rate-sensitive purchases such as housing, or when higher rates prick an asset bubble just waiting to burst. Given the lack of clear imbalances either in the real economy or financial markets, the Fed may have to raise rates significantly more than the market is currently anticipating. In fact, far from having to press the pause button midway through next year, our baseline expectation is that the Fed will expedite the pace of rate hikes in late 2019 as inflation finally starts to accelerate. Aggressive Fed rate hikes combined with an incrementally less expansionary fiscal policy will sow the seeds of a recession in late 2020 or 2021. Before the next U.S. downturn arrives, the dollar will have strengthened further. A resurgent greenback will cast a long shadow over emerging markets and commodity producers. As we discussed last week, China is unlikely to save the day by launching a massive stimulus program of the sort that it orchestrated in both 2009 and 2015.4 True, not all emerging markets are equal. Emerging Asia is more resilient now than it was two decades ago. Thailand, for example, was patient zero for the Asian crisis in 1997. Today, it sports a current account surplus of over 10% of GDP and low levels of external debt. This resilience will not prevent Asian economies from experiencing slower growth on the back of weaker Chinese demand, but it will prevent a full-blown balance of payments crisis from spiraling out of control. In contrast to Emerging Asia, Latin America looks more vulnerable (Table 1). BCA's chief emerging market strategist, Arthur Budaghyan, wisely upgraded Brazilian assets on a tactical basis on October 9th ahead of the presidential elections. Nevertheless, Arthur still worries that Brazil's daunting fiscal challenges - the budget deficit currently stands at 7.8% of GDP and the IMF expects government debt to rise to nearly 100% of GDP over the next five years (Chart 11) - are so grave that even South America's answer to Donald Trump may not be able to save the Brazilian economy. Table 1Vulnerability Heat Map For Key EM Markets Chart 11Brazil Is Fiscally Challenged A Correction, Not A Bear Market The current market environment bears some similarities to the late 1990s. The Fed is tightening monetary policy in order to keep the domestic economy from overheating. The U.S. economy is responding to higher rates to some extent, but the main effects are being felt overseas. The Asian Crisis did not end the bull market in U.S. stocks, but it did generate a few nasty selloffs, the most notable being the 22% peak-to-trough decline in the S&P 500 between July 20 and October 8, 1998. We witnessed such a selloff this October. The bad news is that our MacroQuant model is pointing to additional equity weakness over the coming weeks (Chart 12). The model tends to downgrade stocks whenever growth is slipping, financial conditions are tightening, and sentiment is deteriorating from bullish levels. All three of these things are currently occurring. Chart 12MacroQuant* Model Suggests Caution Is Warranted The good news is that none of our recession indicators are flashing red. Since recessions and bear markets typically overlap (Chart 13), the odds are high that the current stock market correction will be just that, a correction. Chart 13Recessions And Bear Markets Usually Overlap Peter Berezin, Chief Global Strategist Global Investment Strategy peterb@bcaresearch.com 1 The corresponding ETFs are long ITB/short SPY. 2 Please see The Bank Credit Analyst Special Report, "The Long Shadow Of The Financial Crisis," dated October 25, 2018. 3 It is actually not even clear that a loss of confidence in America's fiscal management would cause a recession. The Fed largely determines borrowing costs at the short-to-medium end of the yield curve, which is where the government finances most of its debt. If people lose confidence in the dollar, they will either need to run down their cash balances by purchasing more goods and services or try to move their wealth abroad. The former will directly increase aggregate demand, while the latter will indirectly increase it through a weaker currency. To be clear, we are not suggesting that such an outcome would be beneficial to the economy; it would, among other things, greatly slow potential GDP growth by discouraging investment. But the near-term effect would likely be economic overheating and rising inflation rather than a recession. 4 Please see Global Investment Strategy Weekly Report, "Chinese Stimulus: Not So Stimulating," dated October 26, 2018. Strategy & Market Trends Tactical Trades Strategic Recommendations Closed Trades

Special Report

Highlights Investors looking for equity upside, along with fixed-income-like downside protection, coupled with a hedge against rising rates, should consider convertible bonds. As we near the end of the business cycle, the attractions of convertibles are becoming clearer: investors will benefit from more upside capture in case of a last run-up in stocks, but at the same time suffer less downside in a recession. Moreover, in periods of rising rates, convertible bonds perform well compared to other traditional fixed-income securities. However, multi-asset portfolio managers should note that the risk-return profile of convertible bonds is more like equities than bonds, and so convertibles have no place in a conservative fixed-income portfolio. Investors have a number of options to choose from when customizing equity-versus-fixed-income exposure in their convertible allocations. Feature Introduction An ideal financial instrument would have large equity exposure in an equity bull market, and increased fixed-income exposure in a bear market. Financial engineering can create synthetic positions using derivatives to replicate just this sort of hybrid exposure - or an investor can just buy convertible bonds. In this current, late, phase of the business cycle - with increased volatility, rising interest rates, and a pickup in inflation - where can investors find shelter, but without sacrificing returns in the event of a last blow-out run-up in stocks? In this report, we discuss how convertible bonds - despite their somewhat complex structure1 - could be the answer. Issuers prefer convertibles to traditional corporate bonds because of: 1) a lower coupon rate and fewer covenants, 2) the opportunity to sell equity at a premium to the current price, 3) a faster process for raising capital, compared to a secondary equity issue, and 4) easier access to capital markets for non-investment grade firms. On the demand side, the composition of convertible investors has evolved over time. Prior to the 2007-9 Global Financial Crisis (GFC), proprietary trading desks and leveraged hedge funds were the most important players, since convertible arbitrage2 was very profitable. But the liquidity freeze in 2008 and 2009 forced these short-term investors out of the market and brought back long-term buy-and-hold investors. Currently 65% of U.S. convertible bonds are held by long-only investors. This change in market structure has had important implications for arbitrage opportunities (Chart 1). Chart 1Fewer Short-Term Investors In the first half of 2018, issuance of global convertible securities topped $57 billion, the largest amount for a six-month period since 2008. The U.S. led the way, with issuance of $34 billion (Chart 2), followed by Asia ex-Japan at $12 billion, and Europe, $10 billion. The U.S. total includes $13.4 billion in convertible bond issuance by tech companies, the highest amount in the post-GFC period (Chart 2, panel 2). Bank of America Merrill Lynch estimates that full-year global issuance could be the highest in 12 years. The macro-backdrop for convertibles remains favorable: Chart 2Issuance Similar To Pre-Crisis Levels The hybrid equity/fixed-income exposure offers protection against rising rates because of its shorter duration; The new U.S. tax code limits interest deductibility, which strengthens the relative appeal of issuing a convertible security instead of a traditional bond; The return of volatility means investors benefit from holding a security with an embedded option; The flexibility of the asset class gives investors room to customize their exposure in terms of coupon rate, premium, and maturity. In this report, we start with the market structure and mechanics of convertible bonds. Next, we look at the four types of convertible bonds, which provide different risk-return profiles. In the following section, we analyze historical returns and performance in different market environments. Finally, we discuss the key asset allocation decisions involved in investing in convertible bonds. Our main findings are: Investors can customize their risk-return profile by choosing between high-volatility equity exposure (equity-sensitive convertibles), or more stable fixed-income exposure (credit-sensitive convertibles); Convertible bonds historically have generated an annualized return of 9.5% compared to 9.8% from equities, but with 2% lower volatility; Convertible bonds have a risk-return profile more like that of equities and junk bonds than that of investment-grade credit; In periods of rising rates and inflation, convertible bonds have outperformed their traditional fixed-income counterparts; In comparison to equities, convertibles capture more upside in bull markets than downside in bear markets; Investing in convertible bonds requires active management because of their varying degree of equity- and fixed-income sensitivity that changes over time. The Convertibles Market Convertible securities can be broken into three key groups: 1) convertible bonds (cash-pay3 and zero-coupon), 2) convertible preferred shares, and 3) mandatory convertibles. Cash-pay convertible bonds make up almost 80% of the outstanding market (Chart 3), while zero-coupon convertible bonds are almost non-existent. Mandatories and convertible preferred equities make up 15% and 7% respectively. Chart 3Convertibles Bonds Are 80% Of Convertibles Market... Before we delve deeper into the convertible bond markets, here are few key characteristics (Chart 4) of the other two groups: Chart 4...And Have The Best Risk-Adjusted Returns Convertible Preferred Equities are issued with a specific dividend rate that is generally higher than the dividend on common shares. They include an embedded option to convert to a specified number of common shares. Additionally, preferred dividends usually accumulate in arrears should the firm be unable to make a payment. The conversion rate increases with any increase in the common-share dividend. After the call protection expires, the company has the option of redeeming the issue at the stated par value. Mandatory Convertibles. These bonds automatically convert to common shares at a specified time. However, they do not offer downside protection since conversion can be into shares worth less than the original issue price. Rating agencies view these securities more as equities than bonds, giving firms an incentive to issue them from a balance-sheet perspective. Table 1 shows us that cash-pay (coupon paying convertible bonds) generated the highest return with the lowest volatility, thereby providing investors with the best risk-adjusted returns. Mandatory convertibles have a large excess kurtosis - driven by the forced conversion into equities at inopportune times. In bull and bear markets, it is clear convertible bonds did not enjoy the full upside provided by preferred shares and mandatories, but had 50% less downside in bear markets. Also, in periods of rising rates convertible bonds produced positive returns, but lagged both preferred shares and mandatory convertibles. Table 1Convertible Bonds' Risk-Return Profile A niche market exists for contingent convertibles (CoCos) - or, as they are sometimes called, anti-convertibles. Banks in the euro area issue CoCos to meet capital requirements and provide a cushion should they find themselves in a serious predicament. These typically pay a higher coupon than the bank's straight bonds to compensate for the possibility of a complete wipeout. In short, if all goes well you receive your fixed coupons and principal back at maturity. But, if things turn sour, the bonds convert to equity and the investor potentially loses everything. Mechanics Of Convertible Bonds Convertible bonds are a hybrid security issued as a senior unsecured bond with a fixed maturity (normally five years) with optionality to convert to a fixed number of shares. In exchange for the equity kicker, these bonds typically yield less and carry a lower coupon rate (Chart 5) than the issuer's comparable non-convertible debt. We describe the basics of convertible bonds in the Appendix. Chart 5The Cost Of An Embedded Option An investor considering an allocation to convertibles has four groups to choose from depending on his or her risk-return tolerance. The trade-off is between high volatility equity exposure versus more stable credit exposure. If the underlying stock does well, the convertible increases in value even without the investor exercising the option to convert into shares. If the stock does not appreciate, the investor retains the bond and collects regular coupons and par value at maturity. The interaction of market price with investment value and conversion price creates convertible bonds with different risk-return profiles: Credit Sensitive: A large decrease in the stock price has pushed the convertibles to trade close to their investment value (bond floor). These are out-of-the money convertibles, with a delta ranging from 10% to 40%, and also with large premium over investment value. The main factors affecting the pricing of such instruments are the level of interest rates and credit spreads. An investor has a small probability of generating large unexpected gains from underlying stock appreciation. Balanced: The stock price is close to the conversion price, making these at-the-money convertibles. They have a moderate premium to conversion value, and deltas in the range of 40-80%. Rising stock prices make the embedded call option more valuable, pushing the convertible price closer to the stock price. Long-term buy-and-hold investors looking to maintain a core allocation to convertibles should invest in balanced convertibles. Equity Sensitive: Convertibles that are deep in-the-money, trading near parity, with high deltas of over 80%, and generating returns that closely track equities. They still retain some downside protection due to seniority and par value at maturity even if they have most of the common share's upside potential. Distressed: As a company threatens to default or goes bankrupt, the value of the straight bond component declines to trade significantly below par. These bonds tend to have high degree of price volatility and low probability of return of capital. Risk & Return Convertible bond returns are driven by: 1) the bond component that is a function of rates, credit spreads, and curve effects; 2) the equity component, supported by the delta to the underlying stock price; and 3) the option component, that is a function of the underlying stock price and time to maturity. Convertibles combine characteristics of stocks and bonds (Chart 6), so they represent either lower-volatility equity exposure or enhanced fixed-income exposure. Over the past 24 years (Table 2), U.S. convertible bonds generated returns similar to U.S. equities, but with a lower volatility. However, relative to traditional corporate bonds, convertibles outperformed massively, but with much higher volatility. Looking at risk-adjusted returns, we see that convertible bonds have more similarity to equities and high-yield credit than to investment-grade credit (Chart 7). However, defaults in the convertible bond space have been close to 1%, which is significantly lower than the 4% in the high-yield credit market (Chart 8). This is because convertible bonds include a smaller proportion of issuers with high operating leverage, such as energy producers, and have a high representation of mature healthcare and technology companies. Chart 6Convertibles Vs. Traditional Table 2Better Than Equities, But More Volatile Than Traditional Bonds Chart 7Close To Equities & Junk Chart 8Lower Defaults Than Junk Bonds Short-term performance of the convertible bond market is driven by the composition of issuers, but long-term performance is driven by the performance of the different variables described above. In 1Q 2018, convertible bonds outperformed equities, largely due to technology and consumer staples convertibles. Technology convertibles saw a 11% gain, while the S&P technology sector was up only 3.5%. This was because technology convertible issuers were concentrated in the mid-cap growth segment, whereas the large-cap equity names are more heavily weighted in semiconductors. BCA has for two or three years been warning about the return of inflation and rising interest rates. Convertible bonds outperform traditional fixed income in periods of rising interest rates because: 1) rising rates are often coupled with periods of positive equity momentum, which benefits convertibles; 2) convertibles have lower duration than straight bonds. Since 1994, there have been 10 instances when the 10-year U.S. Treasury yield rose by more than 100 bps: convertible bonds outperformed in every instance. Additionally, convertible bonds enjoy a yield advantage: the average income return (coupon rate) on a convertible is greater than the dividend yield on the underlying stock. When investors allocate to convertible bonds from either their equity or fixed-income portfolio, the key consideration is upside versus downside exposure. When the underlying stock price rises, convertibles will capture a portion of the capital appreciation but, on the downside, convertibles continue to provide a consistent income flow and principal repayment at maturity. History tells us that convertibles capture more upside in bull markets than downside in bear markets. If the share price falls sharply below the conversion price, the convertible will react less and less to fluctuations in the underlying stock price. In short, convertible bonds provide more downside protection than stocks as market value will not drop below the investment value (bond floor). Convertibles also have a mechanism to offset rising equity volatility and rising rates. The embedded equity option in a convertible bond rises in value when volatility rises, providing a meaningful offset in contrast to equities that may suffer a drawdown. Over the long-run, convexity enables this asset to make the most of favorable stock market conditions, whilst suffering less in difficult conditions. As mentioned earlier, the risk-return profile of convertible bonds tends to have a closer relation with equities than with fixed income. Within fixed income, high-yield credit, which tends to have a return profile closely aligned with equities, has a strong correlation with convertible bonds. The greatest diversification potential is when convertible bonds are added to a portfolio of government bonds. However, investors should realize the risk-return profiles for convertibles and government bonds are very different, and an allocation to the former is only a possibility for an investor with a higher risk tolerance. What To Choose From? Equity Sensitive Versus Credit Sensitive Investors need to choose the right type of convertible bond depending on their risk tolerance. Equity-sensitive convertibles made up over 60% of the market prior to the GFC, but this proportion fell to around 20% during the recession (Chart 9). As stock prices tumble, the market price of convertibles get closer to the investment value (bond floor), and convertibles start behaving more like pure credit-sensitive bonds. Looking at total returns (Chart 10 & Table 3), it is clear that aggressive investors with a higher risk tolerance should invest exclusively in equity-sensitive convertibles. But investors looking to maintain a core long-term allocation to convertibles should focus on the balanced group. Despite being a small piece of the market, distressed convertibles are attractive return enhancers immediately after a recession. Investors looking for income return should prefer credit-sensitive or distressed convertibles over equity-sensitive ones. Equity-sensitive convertibles have the highest delta, making them the most vulnerable to underperformance in a downturn. Balanced convertibles have the highest vega, which means they are most impacted by increasing volatility - driven by both equity and rate volatility. In times of rising interest rates, equity-sensitive convertibles provide their best protection given their short duration. Credit- and rate-sensitive convertibles have almost double the duration, making them more vulnerable to rising rates. Chart 9Equity Vs. Fixed Income Exposure Chart 10Massive Outperformance By Equity Sensitive Table 3Equity Sensitive For The Aggressive, Credit Sensitive For The Conservative, Balanced For Everyone Small Cap Versus Large Cap Issues Investors can choose between convertible issues from companies of different size. Since the middle of the financial crisis, large-cap issues have grown to over 50% of the market (Chart 11), up from below 30%. The increase in market share was taken from small-cap issues, with mid-cap issues stable at 20% of the market. In terms of total returns (Chart 12 & Table 4), small cap outperformed both mid and, particularly, large caps. Part of this outperformance was due to the higher yield offered by small-cap issuers compared to their larger counterparts. In terms of equity sensitivity, small-cap issues currently have significantly lower delta than large caps. However, in times of rising volatility, small-cap issues lose more, driven by their higher vega. In terms of interest-rate sensitivity, all three sizes are roughly equally exposed given similar durations. Chart 11Bigger Is Not Always Better Chart 12Small Cap Outperforms Table 4Small Cap Provides The Best Value Investment Grade Versus The Rest A credit investor has one particularly important call: investment-grade versus high-yield. The situation is trickier for convertibles as over 60% of the bonds are unrated (Chart 13), thereby giving managers amply opportunity for alpha generation. Historical performance (Chart 14 & Table 5) shows that non-rated convertible bonds have a close relationship with non-investment-grade issues. Moreover, the relative performance of non-investment-grade and non-rated issues with investment grade issues follows a similar path. From an income-return perspective, both non-rated and non-investment-grade issues have lost their yield advantage since 2016. Investors are not receiving adequate yield for the additional risk they are taking with riskier issues. The return of volatility will have a smaller impact on investment-grade issues compared to the rest of the market because the former have a lower effective duration. Additionally, implied volatility is lower for investment-grade issues. Chart 13Over 60% Has No Credit Rating Chart 14Similar Return, But Different Risk Table 5No Rating = Source Of Alpha The Asset Allocation Decision The key question here is: are investors looking at convertible bonds (Chart 15) as part of an equity or a fixed-income allocation? Investors considering convertibles as part of their equity allocation are looking for a more defensive exposure and yield pick-up, and so should focus on balanced convertibles and not equity-sensitive ones. On the other hand, considering convertibles as part of fixed-income allocation will deliver equity exposure, and so investors should focus on credit-sensitive or balanced convertibles. Chart 15Somewhere Between Equities & Junk Another major factor is the investment horizon of the convertible allocation. A core strategic allocation to convertibles will require a hybrid exposure, providing lower-volatility equity exposure over multiple full market cycles. Such investors are looking for long-term equity upside, but are concerned about shorter-term downside equity volatility and should consider balanced convertibles. On the other hand, investors using convertibles as part of a tactical allocation, to make a short-term bet in order to diversify away from traditional fixed-income or equity exposure, should consider either equity-sensitive or credit-sensitive convertibles. The bottom-line is that convertible investing requires active management because these securities have varying degrees of equity and fixed-income sensitivity that change over time. In periods of rising equity markets, an investor with passive exposure to convertibles would automatically have a large holding in equity-sensitive convertibles with a high delta, thereby increasing his or her exposure to equity downside risk. For example, in February 2009, when markets troughed after the GFC, more than two-thirds of convertibles were trading as credit-sensitive instruments. An investor following a passive index in this situation would have had minimal exposure to equity-sensitive convertibles, and would thereby have had limited participation in the equity upside. Finally, the convertible universe is constantly evolving. The typical convertible bond is issued with a five-year life by a company in the early to mid stage of its corporate life cycle, seeking capital to grow. As time passes, the issuer matures to a point where it no longer needs convertibles in its capital structure. Nearly two-thirds of the current issuers of convertible were not in the market 10 years ago, while two-thirds of the S&P 500 members remain unchanged over this time. Aditya Kurian, Senior Analyst Global Asset Allocation adityak@bcaresearch.com 1 Despite the complexities, the first convertible bond was issued as long ago as 1874 by Rome, Watertown and Ogdensburg Railroad to finance a project. The bond was never converted since the underlying shares failed to rise enough and the company refinanced the bond in 1904. 2 For an explanation of convertible arbitrage, please see A Note On Convertible Arbitrage at the end of this report. 3 Convertible bonds that make regular coupon payments. A Note On Convertible Arbitrage A market-neutral hedge fund strategy where the manager goes long the convertible bond and short the underlying stock. The short position in the underlying stock creates a delta-neutral position, but maintaining this position requires dynamic hedging which is expensive. There is a possibility of large mispricing because of the over-the-counter nature of the market and uncertainty regarding call or redemption features of convertibles. Often, the embedded equity option is a source of cheap volatility compared to the underlying stock's listed options. A quick measure for convertible valuations is comparing the volatility of options in the market to the volatility priced in the embedded option in the convertible. If market volatility rises, but the price of convertible stays the same, the security could be cheap and attractive. Looking at historical performance (Table 6), convertible arbitrage generated almost 3% less than equities, but with less than half the volatility. However, all of the outperformance was during recessions or equity bear markets. Additionally, convertible arbitrage funds have large negative skew and kurtosis relative to both equities and the hedge-fund composite. We recommend investors allocate to convertible arbitrage hedge funds in preparation for a downturn. Table 6Convertible Arbitrage Versus Traditionals Appendix: The Basics Of Convertible Bonds Investment Value (Bond Floor): The fixed-income component of the convertible bond, or in other words, the value of the bond without the conversion feature (equity kicker). This remains stable over a wide range of stock prices but, when creditworthiness deteriorates, consequent stock price movements will have an impact on the investment value (IV). Holding creditworthiness constant, the IV provides the bond floor, below which the convertible should not trade. The IV fluctuates in tandem with the price of a straight corporate bond of similar quality. A convertible that is trading close to its IV will be more affected by changes in rates than one that is well above it. Investment Premium: The market price minus IV expressed as a percentage of IV. Premium over IV indicates the level of downside risk. A higher premium means the bond price is more sensitive to the price of underlying stock, which means less downside protection because the bond market price would have to decline significantly before reaching the IV. Higher premium is a result of rising underlying stock value, whereas a smaller premium is when the convertible is more interest-rate sensitive and behaves like a pure bond. Conversion Value (CV): The equity portion of the convertible bond. Conversion ratio is set at the time of issuance and it is the number of shares a bondholder will receive upon conversion. Conversion price is the price at which the number of converted shares is equal to the par value of the bond. At issuance, the underlying stock price is usually below conversion price. Conversion Premium: The market price minus CV expressed as a percentage of CV. As market price rises above CV, fixed-income attributes are lost and equity features take over, consequently decreasing conversion premium. Declining stock prices mean convertible market price approaches fixed-income value (bond floor) and conversion premium increases. Appendix Chart 1Preferred Shares & Mandatory Convertibles Have Higher Income Returns Appendix Chart 2Convertible Bonds' Delta & Vega Reduces In A Recession Appendix Chart 3Conversion Premium Far From Recessionary Levels Appendix Chart 4Average Duration Less Than 2.5 Appendix Chart 5U.S. Is 60% Of Global Appendix Chart 6U.S. Is Clearly The Best Performer Appendix Chart 7U.S. Also Provides The Best Income Return Appendix Chart 8But, U.S. Is The Most Equity Sensitive Appendix Chart 9U.S. Has A Higher Implied Volatility Appendix Chart 10Distressed Is The Best Solution Immediately After A Recession Appendix Chart 11Balanced Has The Lowest Coupon Appendix Chart 12Balanced Has Moderate Delta, But Highest Vega Appendix Chart 13Equity Sensitive Are The Best Rate Hedge Appendix Chart 14Premiums Are Stable Appendix Chart 15Mid-Cap Provides Low Income Return Appendix Chart 16Massive Delta & Vega Divergence Appendix Chart 17Large Cap Premium Has Risen The Most Appendix Chart 18Implied Volatility Is Similar Across The Board Appendix Chart 19ALl Coupon Rates Have Fallen Appendix Chart 20Investment Grade Has The Highest Delta Appendix Chart 21Underweight Duration = Investment Grade Convertibles Appendix Chart 22Premiums Stable

Dear Client, You will see in this Monthly Portfolio Update that we have expanded our table of Recommendations to include a wider range of the views that Global Asset Allocation (GAA) regularly discusses in its publications. Please see our most recent Quarterly Portfolio Outlook1 for a detailed explanation of those recommendations that we do not specifically touch on in this Monthly. A note on our publication schedule. We will not publish a Monthly for December, or a Q1 2019 Quarterly in mid-December. Instead, we will send you in late November the BCA 2019 Outlook (BCA's annual discussion with Mr. and Ms. X). This will be accompanied by a short GAA note, updating our recommendation tables with a brief commentary. Best Regards, Garry Evans A Correction, Not A Bear Market Investors have a tendency to forget that corrections are common in bull markets. The current equity run-up, which began in March 2009, has seen five corrections (defined as a 10-20% decline in the S&P500). We may now be experiencing the sixth, with the index already down 9.9% from its peak on September 20. Recommendations But we think the evidence is fairly strong that this is just a correction and not the beginning of a new bear market (using the common definition of a 20% or greater fall). It is highly unusual for bear markets to occur - and for bonds to outperform equities - except in the run-up to, and during, recessions (Chart 1). We see little to suggest that a recession in on the horizon over the next 12 months. Chart 1Corrections Are Not At All Rare What caused the correction? The immediate trigger was a seemingly concerted series of statements in early October from FOMC officials, including even doves such as Lael Brainard, that economic circumstances are "remarkably positive" and that rates remain "a long way from neutral" (to quote Fed Chair Jay Powell). In particular, New York Fed President John Williams argued that the neutral rate of interest (the r*) is very uncertain - even though he was joint creator of the main model that estimates it. The implication is that the Fed will keep on raising rates until the economy clearly slows. This pushed the 10-year Treasury yield above 3.2%. Markets are starting to worry that the Fed will make a policy mistake and that certain segments of the economy (housing, emerging markets?) may be too weak to withstand tighter monetary policy. Moreover, this is in a context in which global growth has been weakening (Chart 2), China appears to be slowing quite sharply (Chart 3), the trade war is escalating (with the U.S. now threatening to impose tariffs on all Chinese imports), and valuations for most assets are stretched. Chart 2Outside The U.S., Growth Is Slowing Chart 3Sharp Slowdown Ahead For China? So how worried should investors be? Most of the usual indicators of generalized risk aversion have not flashed strong warning signals during the equity market sell-off (Chart 4). The move up in bond yields came mostly from a rise in real yields, not inflation expectations, and the yield curve steepened, suggesting that markets are pricing in stronger growth not excessive Fed action. Safe haven assets, such as gold and the Swiss franc, did not perform particularly strongly. Credit spreads rose a little, by around 70 basis points, but do not yet signal stress. Chart 4No Signals Of Strong Risk Aversion Moreover U.S. growth, in particular, remains robust. Though the r* may be tricky to estimate, monetary policy is still clearly accommodative and is likely to remain so until at least mid-2019, even if the Fed hikes by 25bp a quarter (Chart 5). Fiscal policy will be stimulative until the end of 2019, adding 1.1 percentage points to growth this year and 0.5 next, according to IMF estimates. Earnings growth will slow from its current lick - Q3 U.S. earnings look like coming in at 23% year-on-year, compared to a forecast of 19% before the results season - but our models suggest that 2019 bottom-up estimates are about right, with growth slowing to around 10% in the U.S. and to somewhat less in the euro area and Japan (Chart 6).2 Chart 5Fed Policy Still Accomodative Chart 6Earnings Growth To Continue, Albeit More Slowly If we have a concern, it is that a few interest-rate sensitive elements of the U.S. economy are showing signs of softness. Housing starts have been weak for a while, but higher mortgage rates may now be having an effect, with residential investment subtracting from GDP growth in all three quarters so far this year (Chart 7). However, mortgage rates are unlikely to continue to rise at the same pace and so the effect should weaken in further quarters. Capex intentions and durable orders have also slipped, perhaps suggesting that corporations have reined back investment plans due to global uncertainties (Chart 8). But these signs point to slower growth next year, not recession, with the U.S. likely to continue to grow above trend. Historically, higher long-term rates have proved a drag on the economy only when they have risen above trend nominal GDP growth, currently around 3.8% (Chart 9). We have some way to go before we reach that tipping-point. Chart 7Housing Is Hurting Chart 8...And Capex Is Getting Cautious Chart 9Rates Matter When They Exceed Nominal Growth We moved to neutral on risk assets, including equities, at the beginning of July. Many of the worries we flagged then have come about. This is late in the cycle, and so volatility will probably remain elevated. However, we do not expect the next recession to come until 2020 at the earliest. Moreover, none of the warning signals on our bear market checklist (which includes the shape of the yield curve, profit margins, a peak in cyclical spending as a percentage of GDP, Fed policy becoming restrictive etc.) are yet flashing, though several may do by mid next year. Equity market valuations are no longer expensive after the recent sell-off (Chart 10). If the current correction were to continue and the drop in the S&P 500 extend to 15% and in global equities to 20% from their most recent peaks, we might be inclined tactically to move back overweight on risk assets. Chart 10Stocks Are No Longer Expensive Currencies: We expect further U.S. dollar appreciation. Divergences in growth and monetary policy between the U.S. and other developed markets will continue. While we expect the Fed to continue to hike once a quarter until end-2019, we could imagine the ECB turning more dovish if euro zone growth continues to slow and Italian BTP 10-year bond yields rise above 4%. The Bank of Japan will stick to its Yield Curve Control policy, which will prevent the yen rising. Emerging market currencies look vulnerable as their economies slow as a result of central bank rate hikes earlier in the year. Asian currencies might undertake competitive devaluations if the renminbi falls below 7, as a result of a worsening trade war. Fixed Income: Long-term rates are unlikely to have peaked for this cycle. Core inflation will stay at around 2% for a few more months because of a favorable base effect, but underlying inflation pressures (the result of rising wages and increases in import tariffs) will push up U.S. inflation by mid next year (Chart 11). A combination of higher inflation, steady Fed hikes, and deteriorating supply/demand conditions (which will raise the term premium) will move 10-year rates above 3.5% by mid-2019 (Chart 12). We accordingly recommend being short duration and overweight TIPs. U.S. high-yield bonds look somewhat attractive, with a default-adjusted spread of 270 bps, after their recent modest sell-off (Chart 13). But this is dependent on our assumption (based on Moody's model) of credit defaults of only 1.04% over the next 12 months.3 Given where we are in the cycle, and considering the elevated corporate leverage in the U.S., we do not consider this a risk worth taking, and so maintain our moderate underweight in credit. Chart 11Underlying Inflation Pressures Are Strong Chart 12Indicators Point To Treasury Yields Above 3.5% Chart 13Are Junk Bonds Attractive Again? Equities: We prefer DM equities over EM, and favor the U.S. and, to a degree, Japan. Emerging markets continue their deleveraging process and will be hurt by rising U.S. rates, a stronger dollar, and slowdown in China. Valuations for EM equities, though one standard deviation cheap relative to global equities, are not yet sufficiently attractively valued to permit investors to buy EM stocks irrespective of their poor fundamentals. Moreover, analysts are still far too optimistic on the outlook for EM earnings, flattering the valuation metric (Chart 14). Stronger growth and an appreciating currency point to an overweight in U.S. equities which, moreover, would be likely to outperform in the event of a deeper correction, given their low beta. Chart 14EM Equities Aren't As Cheap As They Seem Commodities: The crude oil price has fallen back a little in recent weeks, as a result of increases in OPEC production, a modest slowing of demand, and releases of the U.S. Strategic Petroleum Reserve. Our energy strategists have slightly lowered their 2019 Brent forecast to $92 a barrel, from $95 (Chart 15). However, they warn that geopolitical risks, such as widespread application of sanctions on Iran and a collapse in Venezuela, and limits to capacity in Saudi Arabia and U.S. shale production could easily cause spikes above $100.4 A 100% year-on-year rise in oil prices has historically been a clear warning of recession. That would equal Brent at $120 in 1H 2019. Metal prices will continue to be driven by China. At the moment we see no sign of China implementing a major stimulus, which would boost infrastructure spending and therefore demand for commodities (Chart 16), and so we expect further falls in industrial commodities prices. Chart 15Oil Prices Can Rise Further Chart 16No Sings Of Big China Stimilus Garry Evans, Senior Vice President Global Asset Allocation garry@bcaresearch.com 1 Please see GAA Quarterly Portfolio Outlook - October 2018, available at gaa.bcaresearch.com 2 For details of these models and the assumptions behind them, please see The Bank Credit Analyst November 2018, available at bca.bcaresearch.com 3 For details please see BCA U.S. Bond Strategy Weekly Report, "What Kind Of Correction Is This?", dated October 30, 2018, available at usbs.bcaresearch.com 4 For details please see BCA Commodity & Energy Strategy & Bond Strategy Weekly Report, "Risk Premium In Oil Prices Rising; KSA Lifts West Coast Export Capacity", dated October 25, 2018, available at ces.bcaresearch.com GAA Asset Allocation

Highlights After tumbling more than 20% between June and August, copper prices have remained largely static. This reflects the tug-of-war between the near-term bullish physical market fundamentals, and the cloudier macro headwinds ahead, arising from a stronger U.S. dollar. Furthermore, Chinese policymakers are unlikely to abandon their reform agenda and stimulate massively, which will put downside pressure on copper prices further down the road. Despite our negative 12-month outlook, we do not rule out the potential for some upside going into year-end, on the back of falling inventories. Energy: Overweight. News leaked earlier this week indicates the Trump administration is divided over whether to grant waivers on Iranian crude oil imports to big importers like China, India and Turkey, following the re-imposition of sanctions on November 4.1 The U.S. State Department, in particular, appears worried the sanctions will produce a price spike that could derail growth in the U.S. and its allies. This suggests the Administration will be less determined to enforce its Iranian export sanctions, until it has been assured Saudi Arabia and Russia will be able to bring enough production on line in 1H19 to cover the lost Iranian exports, and possible deeper Venezuelan losses. Markets will remain focused on actual export losses from Iran - if they come in at the high end of expectations (i.e., greater than 1.5mm b/d), we expect higher prices; if it becomes apparent the U.S. will go soft on enforcing sanctions, prices would fall. Expect higher volatility. Base Metals: Neutral. Copper prices could rally over the short term, on the back of lower inventories. However, longer term, we see no catalysts to push prices toward recent highs of ~ $3.30/lb on the COMEX. Precious Metals: Neutral. Gold's break above $1,200/oz is holding, but it continues to grind in a $1,210 to $1,240/oz range. Ags/Softs: Underweight. The USDA will report on export sales of grains and oil seeds today. Soybean exports were down 21% y/y for the current crop year, based on the Department's October 18 report. Feature Tight Market Conditions Suggest A Brief Upswing ... After remaining in the $2.90-$3.30/lb trading range for the better part of 1H18, copper prices plunged ~20% since their June peak. The trigger? The escalation of the U.S.-China trade war. The increasingly acrimonious trade relationship acted as a reality check. Investors betting on strike-induced mine supply shortfalls earlier this year were forced to adjust expectations regarding the resilience of the global, and, more specifically, the Chinese business cycles.2 The negative impact of the trade war is clear: copper prices moved to the downside with each escalation in the dispute (Chart of the Week). While current market fundamentals do not necessarily warrant such drastic declines, we see these developments as a wake-up call to market participants. Copper sentiment - previously buoyed by expectations of mine strikes (which failed to materialize) - has come crashing down (Chart 2). Chart of the WeekCopper Down On Trade War Chart 2Sentiment Has Come Crashing Down However, the outlook in the very near term is not so bleak. The evidence below suggests tight physical conditions, indicating copper's next move could be to the upside: Chinese copper imports came in strong in September (Chart 3). While unwrought copper imports reached a 2.5-year high, ores and concentrates forged new record highs. Chart 3Chinese Imports Came In Strong The resilience of Chinese purchases comes on the back of restrictions on scrap imports, which account for a significant share global copper supply (Chart 4). As we have been highlighting, Chinese restrictions on the purity of scrap imports require other forms of the metal to fill the supply gap.3 At the same time, the 25% tariff imposed on Chinese imports of U.S. scrap since August also manifests itself in greater demand for other forms of the metal. This is evident in weak scrap copper imports (Chart 5). Chart 4A Dependable Secondary Market Is Essential For Global Supply Chart 5Scrap Import Restrictions Raise Need For Other Forms Of Copper Copper inventories at the three major global exchange warehouses have been declining steadily and together now stand at half their late April peak (Chart 6). This is their lowest level since late 2016. Chart 6Exchange Inventories At Two Year Low The above evidence of a tight market is in line with copper's futures curve, which is recently pricing a premium for physical delivery (Chart 7). Chart 7Markets Pricing A Premium For Physical Delivery Going into the winter, smelter disruptions may lend further upside support amid these tight conditions: The Vedanta copper smelter in the Indian state of Tamil Nadu was forced to shut down in May due to violent protests. The smelter has an annual production of over 400k MT. In Chile, Codelco gave notice to the market that two of its four smelters will undergo weeks-long outages, in order to comply with tightening of emissions rules - requiring smelters to capture 95% of emissions - due to take effect in December. This will halt production from smelters at the Chuquicamata and Salvador mines for 75 days and 45 days, respectively. Furthermore, in mid-October BHP Billiton reduced its 2018 copper production forecast by 3% to between 1.62mm MT and 1.7mm MT, due to shutdowns at its Olympic Dam facility in Australia and Spence in Chile. Bottom Line: Dynamics at the scrap level in China and disruptions at major smelters in India, Chile and Australia justify tight copper market conditions. This offers potential for a minor rebound in copper prices in the very near term. ... Ahead Of Macro Headwinds In the medium term, macro headwinds will dominate the physical market, capping gains in copper prices. Most notably, fall-out from the U.S.-China trade war in absence of aggressive traditional forms of stimulus, will weigh on demand there. Furthermore, U.S. dollar strength on the back of economic and monetary policy divergences, will make the red metal more expensive for global consumers. Ex-U.S. Growth Unpromising Given the stimulative fiscal policies in the U.S., our House View still does not expect a recession before late-2020. However in the meantime, the global economy will be characterized by divergence in favor of the U.S. (Chart 8). Chart 8Global Economic Divergence Favors U.S. Of utmost importance is, of course, China - where roughly half of global refined copper is consumed. The trade dispute with the U.S. has raised concerns over the resilience of the Chinese economy. Recent data releases have done little to ease fears of a manufacturing slowdown. The Li Keqiang Index and our China construction proxy - both of which are strongly correlated with copper prices - are on a slight downtrend (Chart 9). Chart 9Ominous Signs From China China's 3Q18 GDP data indicate the Chinese economy grew by the slowest pace in nearly a decade (Chart 10). At the same time, PMI's have fallen to or near the 50 level - the boom-bust line - reflecting pessimism in the manufacturing sector. The real estate market - where 45% of China's copper is consumed - also looks gloomy. Home sales rolled over, boding ill for future housing starts. Chart 10Weak Q3 GDP Mirrors Manufacturing And Property Sectors What's more, we are not betting on a flood of stimulus to rescue China's ailing economy. As our colleagues at BCA's Geopolitical Strategy service have been highlighting, the drive to combat vulnerabilities in financial markets raised the pain threshold of Chinese policymakers.4 As such, they are not likely to abandon their reform agenda at the first sign of weakness, as they traditionally have. Although some measures have already been implemented to ease policy, the current response is not yet as promising for commodity markets as has historically been the case. For one, credit growth is constrained by China's de-leveraging campaign. Although there is some evidence that the clampdown on shadow financing is easing, it is not yet at simulative levels (Chart 11). And while the money impulse is rebounding thanks to Reserve Requirement Ratio cuts, the credit impulse is still falling deeper into negative territory. Chart 11Shadow Banking Restrained By Reform Agenda Additionally, as Peter Berezin who heads BCA's Global Investment Strategy highlights, China's more recent forms of (consumption-based) stimulus such as income tax reforms do not boost commodity demand. The same goes for the other way in which Chinese authorities are trying to stabilize their economy: by depreciating the RMB. This is in clear contrast to traditional measures such as fixed asset investment, which stimulate demand for raw materials and capital goods.5 Overall, the current level of stimulus is not sufficient to boost the Chinese economy. Nor, by extension, is it enough to lift EMs, and commodity prices in the process. In fact, copper markets have been oblivious to various announcements by Chinese authorities that they are easing policy (Chart 12). Chart 12Copper Markets Oblivious To Chinese Stimulus Our Geopolitical Strategists warn that the U.S.-China trade war could get worse before it improves. Thus, while policymakers are not yet compelled to throw in the towel with their reform agenda, they are pragmatic and will likely intensify their response if conditions deteriorate further. If authorities were to deploy massively stimulative fiscal and monetary policy by propping up infrastructure and the real estate sector - as they traditionally have done - chances are that we would be able to escape further price weakness in copper markets. For now, the evidence points at a more modest policy approach. Green Dollar, Red Metal As a counter-cyclical currency, the U.S. dollar will shine in the current weaker ex-U.S. growth environment. What's more, limited spare capacity in the U.S. and a strong labor market foreshadow rising U.S. inflation readings. This will justify continued tightening by the Fed. Economic divergences favoring the U.S. economy will amplify the impact. Rising U.S. borrowing costs will be painful for debt-laden EM economies. Their Central Banks will struggle to keep the pace with the Fed. Similarly, the European Central Bank - conscious of turmoil in Italy - will be forced to maintain a more dovish stance. This will weigh down on the EUR/USD. A stronger dollar generally dents demand by making commodities - priced in U.S. dollars - more expensive for foreign consumers. While energy markets dominated by supply risks remain disconnected from their long-term negative correlation with the U.S. dollar, the relationship with metals has re-converged (Chart 13).6 This leaves copper more vulnerable to the downside amid dollar strengthening. The impact will be magnified for Chinese consumers as the RMB weakens further, forcing the top consumer to cut down on imports of the red metal. Chart 13USD-Copper Relationship Re-converged Bottom Line: Headwinds from weakness in China and a stronger dollar will be a drag on demand next year. Unless Chinese policymakers temporarily abandon their reform agenda and stimulate massively, medium term copper prices will face pressures to the downside. Model Updates Given the macro headwinds outlined above, we revised our copper demand forecast. Our balances now point to a slight surplus in 2019 (Chart 14). In the context of 24mm MT of consumption p.a., a 100k MT surplus can be characterized as a balanced market. This makes prices vulnerable to upside or downside surprises, which can easily tip the scale. Chart 14Broadly Balanced Market In line with our market assessment, we simulated forecasts for copper prices based on a 5% and 10% appreciation in the USD over the coming 12 months (Chart 15). Chart 15Macro Headwinds In 2019   Roukaya Ibrahim, Editor/Strategist Commodity & Energy Strategy RoukayaI@bcaresearch.com 1 Reuters published an interesting analysis containing the apparently leaked information re the internal disputes in the Trump administration entitled "Trump's sanctions on Iran tested by oil-thirsty China, India" on October 29, 2018. 2 In the Commodity & Energy Strategy Weekly Report published January 25, 2018, we highlighted the risk to mine supply in 2018 on the back of an unusually large number of labor contract renegotiations taking place this year - representing ~ 5 mm MT worth of mined copper. Most noteworthy was the risk of a strike at the Escondida copper mine in Chile. These have been largely resolved with minimal impact on supply. Please see "Stronger USD, Slower China Growth Threaten Copper," available at ces.bcaresearch.com. 3 Please see BCA Research Commodity & Energy Strategy Weekly Report titled "Copper: A Break Out, Or A Break Down?" dated May 17, 2018. Available at ces.bcaresearch.com. 4 Please see BCA Research Geopolitical Strategy Special Report titled "China Sticks To The Three Battles," dated October 24, 2018. Available at gps.bcaresearch.com. 5 Please see BCA Research Global Investment Strategy Weekly Report "Chinese Stimulus: Not So Stimulating" dated October 26, 2018, available at gis.bcaresearch.com. 6 Please see BCA Research Commodity & Energy Strategy Weekly Report titled "Correlations Vs. USD Weaken," dated June 14, 2018. Available at ces.bcaresearch.com. 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