Sectors

Special Report

This is the second of a two-part Special Report on the structural changes that have occurred as a result of the Great Recession and financial crisis. We look at three issues: asset correlation, the safety of the financial system, and the level of global debt. First, correlations among financial assets shifted dramatically during the financial crisis and the after-effects lingered for years. Some believe that the underlying level of correlation among risk assets has shifted permanently higher for two main reasons: (1) trading factors such as the increased use of exchange-traded funds and algorithms; and (2) the risk-on/risk-off environment in which trading has become more binary in nature, due to the sharp rise in policy uncertainty, risk aversion and risk premiums in the aftermath of the Great Recession. We have sympathy for the second explanation. The equity risk premium (ERP) was forced higher on a sustained basis by the financial crisis, driven by fears that the advanced economies had entered a ‘secular stagnation’. Elevated correlation among risk assets was a result of a higher-than-normal ERP. The ERP should decline as fears of secular stagnation fade, leading to a lower average level of risk asset correlation than has been the case over the last decade. Second, regulators have been working hard to ensure that the financial crisis never happens again. But is the financial system really any safer today? Undoubtedly, banks have improved balance sheet and funding resilience, and have significantly reduced their involvement in complex financial activities. The propensity for contagion among banks has diminished and there has been a dramatic decline in the volume of complex structured credit securities. The bad news is that the level of global debt has increased at an alarming pace. The third part of this report highlights that elevated levels of debt could cause instability in the global financial system. Choking debt levels boost the vulnerability to negative shocks. The number and probability of potential shocks appear to have increased since 2007, including extreme weather events, sovereign debt crises, large-scale migration, populism, water crises and cyber & data attacks. The lack of a fiscal buffer in most countries means that it will be difficult or impossible to provide any fiscal relief in the event of a negative shock. Moreover, the end of the Debt Supercycle means that the monetary and fiscal authorities will find it difficult to encourage the private sector to spend more in most cases. For EM, deleveraging has not even started and more financial fireworks seem inevitable in the context of a strong dollar and rising global yields. China may avoid a crisis, but the adjustment to a less credit-driven economy is already proving to be a painful process. The Great Recession and Financial Crisis cast a long shadow that will affect economies, policy and financial markets for years to come. Rather than reviewing the roots of the crisis, the first of our two-part series examined the areas where we believe structural change has occurred related to the economy or financial markets. We covered the changing structure of the corporate bond market, the inflation outlook, central bank policymaking and equilibrium bond yields. We highlighted that the financial crisis transformed the corporate bond market in several ways that heighten the risk for quality spreads in the next downturn. We made the case that the prolonged inflation undershoot is sowing the seeds of an overshoot in the coming years, in part related to central bank policymakers that are doomed to fight the last war. Finally, we argued that the forces behind the structural and cyclical bull market in bonds reached an inflection point in 2016/2017. In Part II, we examine the theory that the financial crisis has permanently lifted market correlations among risk assets. Next, we look at whether regulatory changes implemented as a result of the financial crisis have made the global financial system safer. Finally, we highlight the implications of the continued rise in global leverage over the past decade in the context of BCA’s Debt Supercycle theme. The bottom line is that the global financial system still faces substantial risks, despite a more highly regulated banking system. (1) Are Risk Asset Correlations Permanently Higher? Correlations among financial assets shifted dramatically during the financial crisis and the after-effects lingered for years. For example, risk assets became more highly correlated, suggesting little differentiation within or across asset classes. Chart II-1 presents a proxy for U.S. equity market correlations, using a sample of current S&P 100 companies. The average correlation was depressed in the 1990s and 2000s relative to the 1980s. It spiked in 2007 and fluctuated at extremely high levels for several years, before moving erratically lower. It has jumped recently and is roughly in the middle of the post-1980s range. Chart II-1Two Factors Driving Correlation Correlations will undoubtedly ebb and flow in the coming years and will spike again in the next recession. But a key question is whether correlations will oscillate around a higher average level than in the 1990s and 2000s. The consensus seems to believe that the underlying level of correlation among risk assets has indeed shifted higher on a structural basis for two main reasons: Market Structure Changes: Many investors point to trading factors such as the increased use of index products (exchange-traded funds for example), and high-frequency/algorithmic trading as likely culprits. Macro “theme” investing has reportedly become more popular and is often implemented through algorithms. The result is an increase in stock market volatility and a tendency for risk-asset prices to move up and down based on momentum because they are all being traded as a group. These factors would likely be evident today even if the financial crisis never happened, but the popularity of algorithm trading may have been encouraged by the fact that the macro backdrop was so uncertain for years after Lehman collapsed. Risk On/Off Trading Environment: Trading has become more binary in nature, due to the sharp rise in policy uncertainty, risk aversion and risk premiums in the aftermath of the Great Recession. Even after the recession ended, the headwinds to growth were formidable and many felt that the sustainability of the recovery hinged largely on the success or failure of unorthodox monetary policies. The general feeling was that either the policies would “work”, the output gap would gradually close and risk assets would perform well, or it would fail and risk assets would be dragged down by a return to recession. Thus, markets traded on an extreme “risk-on/risk-off” basis, as sentiment swung wildly with each new piece of economic and earnings data. While the market structure thesis has merit on the surface, the impact should only be short term in nature. It is difficult to see how a change in the intra-day microstructure of the market could have such a fundamental, wide-ranging and permanent impact on market prices. Previous research suggests that any impact on market correlation beyond the very short term is likely to be small. For the sake of brevity, we won’t present the evidence here, but instead refer readers to two BCA Special Reports.1 The risk on/off trading environment thesis is a more plausible explanation. However, we find it more useful to think about it in terms of the equity risk premium (ERP). A higher ERP causes investors to revalue cash flows from all firms, which, in turn, causes structural shifts in the correlation among stocks. A lower ERP results in less homogenization of the present value of future cash flows, and raises the effect of differentiation among business models. A rise in the ERP could occur for different reasons, but the most obvious are an increase in the perceived riskiness of firms, a shift in investor risk aversion, or both. Shifts in the ERP are sometimes structural in nature, but there is also a strong cyclical element in that persistent equity declines historically have had the effect of temporarily raising the ERP and correlations. A simple model based on the ERP and volatility explains a lot of the historical variation in equity correlation, including the elevated levels observed in the years after 2007 (Chart II-2).2 The shift lower in correlations after 2012 reflects both a lower equity risk premium and a dramatic decline in downside volatility. Chart II-2Simple Model Explains Correlation It is tempting to believe that the lingering shell-shock related to the financial crisis means that the underlying equity risk premium has shifted permanently higher. The ERP is still elevated by historical standards, but this is more reflective of extraordinarily low bond yields than an elevated forward earnings yield. Investors evidently believe that the U.S. and other developed economies are stuck in a “secular stagnation”, which will require low interest rates for many years just to keep economic growth near its trend pace. In other words, the equilibrium interest rate, or R-star, is still very low. The ERP and correlations among risk assets will undoubtedly spike again in the next recession. Nonetheless, in the absence of recession, we expect fears regarding secular stagnation to fade further. If the advanced economies hold up as short-term interest rates and bond yields rise, then concerns that R-star is extremely low will dissipate and expectations regarding equilibrium bond yields will shift higher. The ERP will move lower as bond yields, rather than the earnings yield, do most of the adjustment. The underlying correlations among risk asset prices should correspondingly recede. This includes correlations among a wide variety of risk assets, such as corporate bonds and commodities. While this describes our base case outlook, there is a non-trivial risk that the next recession arrives soon and is deep. This would underscore the view that R-star is indeed very low and the economy needs constant monetary stimulus just to keep it out of recession (i.e. the secular stagnation thesis). The ERP and correlations would stay elevated on average in that scenario. What About The Stock/Bond Correlation? Chart II-3 shows the rolling correlation between monthly changes in the 10-year Treasury bond yield and the S&P 500. The correlation was generally negative between the late-1960s and the early-2000s. Bond yields tended to rise whenever the S&P 500 was falling. Over the past two decades, however, bond yields have generally declined when the stock market has swooned. Chart II-3Structural Shifts In The Stock/Bond Correlation Inflation expectations can help explain the shift in stock/bond correlation. Expectations became unmoored after 1970, which meant that inflationary shocks became the primary driver of bond yields. Strong growth became associated with rising inflation and inflation expectations, and the view that central banks had fallen behind the curve. Bond yields surged as markets discounted aggressive tightening designed to choke off inflation. And, given that inflation lags the cycle and had a lot of persistence, central banks were not in a position to ease policy at the first hint of a growth slowdown. This was obviously a poor backdrop for stocks. When inflation expectations became well anchored again around the late 1990s, investors no longer feared that central banks would have to aggressively stomp on growth whenever actual inflation edged higher. Central banks also had more latitude to react quickly by cutting rates at the first sign of slower economic growth. Fluctuations in growth became the primary driver of bond yields, allowing stock prices to rise and fall along with yields. The correlation has therefore been positive most of the time since 2003. Bottom Line: A negative correlation between stocks and bond yields reared its ugly head in the last quarter of 2018. The equity correction reflected several factors, but the previous surge in bond yields and hawkish Fed comments appeared to spook markets. Investors became nervous that the fed funds rate had already entered restrictive territory, at a time when the global economy was cooling off. We expect more of these episodes as the Fed normalizes short-term interest rates over the next couple of years. Nonetheless, we see no evidence that inflation expectations have become unmoored. This implies that the stock-bond correlation will generally be positive most of the time over the medium term. In addition, the average level of correlation among risk assets has probably not been permanently raised, although spikes during recessions or growth scares will inevitably occur. (2) Is The Global Financial System Really Safer Today? The roots of the great financial crisis and recession involved a global banking and shadow banking system that encouraged leverage and risk-taking in ways that were hard for investors and regulators to assess. Complex and opaque financial instruments helped to hide risk, at a time when regulators were “asleep at the switch”. In many countries, credit grew at a much faster pace than GDP and capital buffers were dangerously low. Banking sector compensation skewed the system toward short-term gains over long-term sustainable returns. Lax lending standards and a heavy reliance on short-term wholesale markets to fund trading and lending activity contributed to cascading defaults and a complete seizure in parts of the money and fixed income markets. A vital question is whether the financial system is any less vulnerable today to contagion and seizure. The short answer is that the financial system is better prepared for a shock, but the problem is that the number of potential sources of instability have increased since 2007. Since the financial crisis, regulators have been working hard to ensure that the financial crisis never happens again. Reforms have come under four key headings: Capital: Regulators raised the minimum capital requirement for banks, added a buffer requirement, and implemented a surcharge on systemically important banks. Liquidity: Regulators implemented a Liquidity Coverage Ratio (LCR) and a Net Stable Funding Ratio (NSFR) in order to ensure that banks have sufficient short-term funds to avoid liquidity shortages and bank runs.3 Risk Management: Banks are being forced to develop systems to better monitor risk, and are subject to periodic stress tests. Resolution Planning: Banks have also been asked to detail options for resolution that, hopefully, should reduce systemic risk should a major financial institution become insolvent. Global systemically-important banks, in particular, will require sufficient loss-absorbing capacity. A major study by the Bank for International Settlements,4 along with other recent studies, found that systemic risk in the global financial system has diminished markedly as a result of the new regulations. On the whole, banks have improved balance sheet and funding resilience, and have significantly reduced their involvement in complex financial activities. Lending standards have tightened almost across the board relative to pre-crisis levels, particularly for residential mortgages. Additional capital and liquid assets provide a much wider buffer today against adverse shocks, allowing most banks to pass recent stress tests (Chart II-4). Financial institutions have generally re-positioned toward retail and commercial banking and wealth management, and away from more complex and capital-intensive activities (Chart II-5). The median share of trading assets in total assets for individual G-SIBs has declined from around 20% to 12% over 2009-16. Moreover, the propensity for contagion among banks has diminished. The BIS notes that assessing all the complex interactions in the global financial system is extremely difficult. Nonetheless, a positive sign is that banks are focusing more on their home markets since the crisis, and that direct connections between banks through lending and derivatives exposures have declined. The BIS highlights that aggregate foreign bank claims have declined by 16% since the crisis, driven particularly by banks from the advanced economies most affected by the crisis, especially from some European countries (Chart II-6). It is also positive that European banks have made some headway in diminishing over-capacity, although problems still exist in Italy. Finally, and importantly, there has been a distinct shift toward more stable sources of funding, such as deposits, away from fickle wholesale markets (Charts II-7 and II-8). Chart II-6Less Cross Border Lending (Until Recently)   Outside of banking, many other regulatory changes have been implemented to make the system safer. One important example is that rules were adjusted to reduce the risk of runs on money market funds. What About Shadow Banking? Of course, more could be done to further indemnify the financial system. Concentration in the global banking system has not diminished, and it appears that the problem of “too big to fail” has not been solved. And then there is the shadow banking sector, which played a major role in the financial crisis by providing banks a way of moving risk to off-balance sheet entities and securities, and thereby hiding the inherent risks. Shadow banking is defined as credit provision that occurs outside of the banking system, but involves the key features of bank lending including leverage, and liquidity and maturity transformation. Complex structured credit securities, such as Collateralized Debt Obligations, allowed this type of transformation to mushroom in ways that were difficult for regulators and investors to understand. A recent study by the Group of Thirty5 concluded that securitization has dropped to a small fraction of its pre-crisis level, and that growing non-bank credit intermediation since the Great Recession has primarily been in forms that do not appear to raise financial stability concerns. Much of the credit creation has been in non-financial corporate bonds, which is a more stable and less risky form of credit extension than bank lending. Other types of lending have increased, such as corporate credit to pension funds and insurance companies, but this does not involve maturity transformation, according to the Group of Thirty. There has been a dramatic decline in the volume of complex structured credit securities such as collateralized debt obligations, asset-backed commercial paper, and structured investment vehicles since 2007 (Chart II-9). While the situation must be monitored, the Group of Thirty study concludes that the financial system in the advanced economies appears to be less vulnerable to bouts of self-reinforcing forced selling, such as occurred during the 2008 crisis. Chart II-9Less Private-Sector Securitization One exception is the U.S. leveraged loan market, which has swelled to $1.13 trillion and about half has been pooled into Collateralized Loan Obligations. As with U.S. high-yield bonds, the situation is fine as long as profitability remains favorable. But in the next recession, lax lending standards today will contribute to painful losses in leveraged loans. The Bad News That’s the good news. The bad news is that, while the financial system might have become less complex and opaque, the level of debt has increased at an alarming rate in both the private and public sectors in many countries. Elevated levels of debt could cause instability in the global financial system, especially as global bond yields return to more normal levels by historical standards. We discuss other pressure points such as Emerging Markets and China in the next section, although the latter deserves a few comments before we leave the subject of shadow banking. The Group of Thirty notes that 30% of Chinese credit is provided by a broad array of poorly regulated shadow banking entities and activities, including trust funds, wealth management products, and “entrusted loans.” Links between these entities and banks are unclear, and sometimes involve informal commitments to provide credit or liquidity support. The study takes some comfort that most of Chinese debt takes place between Chinese domestic state-owned banks and state-owned companies or local government financing vehicles. Foreign investors have limited involvement, thus reducing potential direct contagion outside of China in the event of a financial event. Still, the potential for contagion internationally via global sentiment and/or the economic fallout is high. The other bad news is that, while regulators in the advanced economies have managed to improve the ability of financial institutions to weather shocks, potential risks to the financial system have increased in number and in probability of occurrence. The Global Risk Institute (GRI) recently published a detailed comparison of potential shocks today relative to 2007.7 The report sees twice the number of risks versus 2007 that are identified as “current” (i.e. could occur at any time) and of “high impact”. The most pressing risks today include extreme weather events, asset bubbles, sovereign debt crises, large-scale involuntary migration, water crises and cyber & data attacks. Any of these could trigger a broad financial crisis if the shock is sufficiently intense, despite improved regulation. The GRI study also eventuates how the risks will evolve over the next 11 years. Readers should see the study for details, but it is interesting that the experts foresee cyber dependency rising to the top of the risk pile by 2030. The increase is driven by the importance of data ownership, the increasing role of algorithms and control systems, and the $1.2 trillion projected cost of cyber, data and infrastructure attacks. Our computer systems are not prepared for the advances of technology, such as quantum computing. Climate change moves to the number two risk spot in its base-case outlook. Space limitations precluded a discussion of the rise of populism in this report, but the GRI sees the political tensions related to income inequality as the number three threat to the global financial system by 2030. Bottom Line: Regulators have managed to substantially reduce the amount of hidden risk and the potential for contagion between financial institutions and across countries since 2007. Banks have a larger buffer against stocks. Unfortunately, the number and probability of potential shocks to the financial system appear to have increased since 2007. (3) Implications Of The Global Debt Overhang The End of the Debt Supercycle is a key BCA theme influencing our macro view of the economic and market outlook for the coming years. For several decades, the willingness of both lenders and borrowers to embrace credit was a lubricant for economic growth and rising asset prices and, importantly, underpinned the effectiveness of monetary policy. During times of economic and/or financial stress, it was relatively easy for the Federal Reserve and other central banks to improve the situation by engineering a new credit up-cycle. However, since the 2007-09 meltdown, even zero (or negative) policy rates have been unable to trigger a strong revival in private credit growth in the major developed economies, except in a few cases. The end of the Debt Supercycle has severely impaired the key transmission channel between changes in monetary policy and economic activity. The combination of high debt burdens and economic uncertainty has curbed borrowers’ appetite for credit while increased regulatory pressures and those same uncertainties have made lenders less willing to extend loans. This has severely eroded the effectiveness of lower interest in boosting credit demand and supply, forcing central banks to rely increasingly on manipulating asset prices and exchange rates. On a positive note, the plunge in interest rates has lowered debt servicing costs to historically low levels. Yet, it is the level, rather than the cost, of debt that seems to have been an impediment to the credit cycle, contributing to a lethargic economic expansion. The Bank for International Settlements (BIS) publishes an excellent dataset of credit trends across a broad swath of developing and emerging economies. Some broad conclusions come from an examination of the data (Charts II-10 and II-11):7 Chart II-10Advanced Economies: Some Deleveraging Chart II-11EM: Deleveraging Has Not Even Started Private debt growth has only recently accelerated for the advanced economies as a whole. There are only a handful of developed economies where private debt-to-GDP ratios have moved up meaningfully in the past few years. These are countries that avoided a real estate/banking bust and where property prices have continued to rise (e.g. Canada and Australia). The high level of real estate prices and household debt currently is a major source of concern to the authorities in those few countries. Even where some significant consumer deleveraging has occurred (e.g. the U.S., Spain and Ireland), debt-to-income ratios remain very high by historical standards. In many cases, a stabilization or decline in private debt burdens has been offset by a continued rise in public debt, keeping overall leverage close to peak levels. This is a key legacy of the financial crisis; many governments were forced to offset the loss of demand from private sector deleveraging by running larger and persistent budget deficits. Weak private demand accounts for close to 50% of the rise in public debt on average according to the IMF. Global debt of all types (public and private) has soared from 207% of GDP in 2007 to 246% today. The Debt Supercycle did not end everywhere at the same time. It peaked in Japan more than 20 years ago and has not yet reached a decisive bottom. The 2007-09 meltdown marked the turning point for the U.S. and Europe, but it has not even started in the emerging world. The financial crisis accelerated the accumulation of debt in the latter as investors shifted capital away from the struggling advanced economies to (seemingly less risky) emerging markets. Both EM private- and public-sector debt ratios have continued to move up at an alarming pace. The lesson from Japan is that deleveraging cycles following the bursting of a major credit bubble can last a very long time indeed. One key area where there has been significant deleveraging is the U.S. household sector (Chart II-12). The ratio of household debt to income has fallen below its long-term trend, suggesting that the deleveraging process is well advanced. However, one could argue that the ratio will undershoot the trend for an extended period in a mirror image of the previous overshoot. Or, it may be that the trend has changed; it could now be flat or even down. Chart II-12U.S. Household Deleveraging... What is clear is that U.S. attitudes toward saving and spending have changed dramatically since the Great Financial Crisis (GFC) (Chart II-13). Like the Great Depression of the 1930s that turned more than one generation off of debt, the 2008/09 crisis appears to have been a watershed event that marked a structural shift in U.S. consumer attitudes toward credit-financed spending. The Debt Supercycle is over for this sector. Chart II-13...As Attitudes To Debt Change Developing Countries: Debt And Economic Fundamentals BCA’s long-held caution on emerging economies and markets is rooted in concern about deteriorating fundamentals. Trade wars and a tightening Fed are negative for EM assets, but the main headwinds facing this asset class are structural. Excessive debt is a ticking time bomb for many of these countries. EM dollar-denominated debt is now as high as it was in the late 1990s as a share of both GDP and exports (Chart II-14). Moreover, the declining long-term growth potential for emerging economies as a group makes it more difficult for them to service the debt. The structural downtrend in EM labor force and productivity growth underscores that trend GDP growth has collapsed over the past three decades (Chart II-14, bottom panel). Chart II-14EM: High Debt And Slow Growth... The 2019 Key Views8 report from our Emerging Markets Strategy team highlights that excessive capital inflows over the past decade have contributed to over-investment and mal-investment. Much of the borrowing was used to fund unprofitable projects, as highlighted by the plunge in productivity growth, profit margins and return on assets in the EM space relative to pre-Lehman levels (Chart II-15) Decelerating global growth in 2018 has exposed these poor fundamentals. Chart II-15...Along With Deteriorating Profitability As we highlighted in the BCA Outlook 2019, emerging financial markets may enjoy a rally in the second half of 2019 on the back of Chinese policy stimulus. However, this will only represent a ‘sugar high’. The debt overhang in emerging market economies is unlikely to end benignly because a painful period of corporate restructuring, bank recapitalization and structural reforms are required in order to boost productivity and thereby improve these countries’ ability to service their debt mountains. China’s Debt Problem Space limitations preclude a full discussion of the complex debt situation in China and the risks it poses for the global financial system. Waves of stimulus have caused total debt to soar from 140% of GDP in 2008 to 260% of GDP at present (Chart II-16). Since most of the new credit has been used to finance fixed-asset investment, China has ended up with a severe overcapacity problem. The rate of return on assets in the state-owned corporate sector has fallen below borrowing costs (Chart II-17). Chinese banks are currently being told that they must lend more money to support the economy, while ensuring that their loans do not sour. This has become an impossible feat. Chart II-16China's Overinvestment... Chart II-17Has Undermined The Return On Assets The previous section highlighted that much of the debt has been created in the opaque shadow banking system, where vast amounts of hidden risk have likely accumulated. Whether or not the central government is willing and/or able to cover a wave of defaults and recapitalize the banking system in the event of a negative shock is hotly debated, both within and outside of BCA. But even if a financial crisis can be avoided, bringing an end to the unsustainable credit boom will undoubtedly have significant consequences for the Chinese economy and the emerging economies that trade with it. Interest Costs To Rise Globally, many are concerned about rising interest costs as interest rates normalize over the coming years. In Appendix Charts II-19 to II-21, we provide interest-cost simulations for selected government, corporate and household sectors under three interest-rate scenarios. The good news is that the starting point for interest rates is still low, and that it takes years for the stock of outstanding debt to adjust to higher market rates. Even if rates rise by another 100 basis points, interest burdens will increase but will generally remain low by historical standards. It would take a surge of 300 basis points across the yield curve to really ‘move the needle’ in terms of interest expense. This does not imply that the global debt situation is sustainable or that a financial crisis can be easily avoided. The next economic downturn will probably not be the direct result of rising interest costs. Nonetheless, elevated government, household and/or corporate leverage has several important long-term negative implications: Limits To Counter-Cyclical Fiscal Policy: Government indebtedness will limit the use of counter-cyclical fiscal policy during the next economic downturn. Chart II-18 highlights that structural budget deficits and government debt levels are higher today compared to previous years that preceded recessions. The risk is especially high for emerging economies and some advanced economies (such as Italy) where investors will be unwilling to lend at a reasonable rate due to default fears. Even in countries where the market still appears willing to lend to the government at a low interest rate, political constraints may limit the room to maneuver as voters and fiscally-conservative politicians revolt against a surge in budget deficits. This will almost certainly be the case in the U.S., where the 2018 tax cuts mean that the federal budget deficit is likely to be around 6% of GDP in the coming years even in the absence of recession. A recession would push it close to a whopping 10%. Even in countries where fiscal stimulus is possible, the end of the Debt Supercycle means that the monetary and fiscal authorities will find it difficult to encourage the private sector to spend and take on more debt. Growth Headwinds: The debt situation condemns the global economy to a slower pace of trend growth in part because of weaker capital spending. From one perspective this is a good thing, because spending financed by the excessive use of credit is unsustainable. Still, deleveraging has much further to go at the global level, which means that spending will have to be constrained relative to income growth. The IMF estimates that deleveraging in the private sector for the advanced economies is only a third of historical precedents at this point in the cycle. The IMF also found that debt overhangs have historically been associated with lower GDP growth even in the absence of a financial crisis. Sooner or later, overleveraged sectors have to retrench. Vulnerability To Negative Shocks: If adjustment is postponed, debt reaches levels that make the economy highly vulnerable to negative shocks as defaults rise and lenders demand a higher return or withdraw funding altogether. IMF work shows that economic downturns are more costly in terms of lost GDP when it is driven or accompanied by a financial crisis. This is particularly the case for emerging markets. Bottom Line: Although credit growth has been subdued in most major advanced economies, there has been little deleveraging overall and debt-to-GDP is still rising at the global level. Elevated debt levels are far from benign, even if it appears to be easily financed at the moment. It acts as dead weight on economic activity and makes the world economy vulnerable to negative shocks. It steals growth from the future and, in the event of such a shock, the lack of a fiscal buffer in most countries means that it will be difficult or impossible to provide fiscal relief. The end of the Debt Supercycle means that the monetary and fiscal authorities will find it difficult to encourage the private sector to spend in most cases. For EM, deleveraging has not even started and more financial fireworks seem inevitable in the context of a strong dollar and rising global yields. China may avoid a crisis, but the adjustment to a less credit-driven economy is already proving to be a painful process. Mark McClellan Senior Vice President The Bank Credit Analyst APPENDIX Chart II-19Corporate Interest Cost Scenarios   Chart II-20Government Interest Cost Scenarios   Chart II-21U.S. Household Sector Interest Cost Scenarios​​​​​​​​​​​​​​​​​​​​​ 1      Please see BCA U.S. Investment Strategy Special Report "The Bane Of Investors’ Existence: Why Is Correlation High And When Will It Fall?" dated January 4, 2012, available at usis.bcaresearch.com. Also see BCA Global ETF Strategy Special Report "The Passive Menace," dated September 13, 2017, available at etf.bcaresearch.com 2       We use only below average returns in the calculation of volatility (downside volatility) because we are more concerned with the risk of equity market declines for the purposes of this model. 3       The LCR requires a large bank to hold enough high-quality liquid assets to cover the net cash outflows the bank would expect to occur over a 30-day stress scenario. The NSFR complements the LCR by requiring an amount of stable funding that is tailored to the liquidity risk of a bank’s assets and liabilities, based on a one-year time horizon. 4       Structural Changes in Banking After the Crisis. CGFS Papers No.60. Bank for International Settlements, January 2018. 5       Shadow Banking and Capital Markets Risks and Opportunities. Group of Thirty. Washington, D.C., November 2016. 6       Back to the Future: 2007 to 2030. Are New Financial Risks Foreshadowing a Systemic Risk Event? Global Risk Institute. 7       For more details on public and private debt trends, please see BCA Special Report "The End Of The Debt Supercycle: An Update," dated May 11, 2016, available at bca.bcaresearch.com 8       Please see BCA Emerging Markets Strategy Weekly Report "2019 Key Views: Will The EM Lost Decade End With A Bang Or A Whimper?" dated December 6, 2018, available at ems.bcaresearch.com

Overweight Though it may be hard to see in the top panel of our chart amidst a spectacular long-term fall from grace, the S&P industrial conglomerates index has been outperforming for the past week. At first glance, much of the credit for this bounce can be given to GE which has seen even the longest term sell-side bears throwing in the towel and turning positive. However, GE is no longer the dominant name in industrial conglomerates; at only 21.5% of the market cap weight of the S&P industrial conglomerates index, it ranks below both MMM and HON which together account for more than two-thirds. As such, any outperformance would have to be broad based, as it has been since the beginning of December. Considering how valuations have contracted, this is of little wonder. The S&P industrial conglomerates index is a full standard deviation below fair value according to our Valuation Indicator (second panel) while our Technical Indicator (bottom panel) shows the index to still be exceptionally oversold. Bottom Line: The pessimism baked in to the S&P industrial conglomerates index is out of proportion. We reiterate our valuation-driven overweight recommendation. The ticker symbols for the stocks in this index are: BLBG: S5INDCX - GE, MMM, HON, ROP. ​​​​​​​

Underweight The S&P soft drinks index popped in late October, driven by better than expected Q3 results, mostly at Coke. In truth, pricing power has been staging a fairly steady recovery since falling off a cliff in 2016, though it has recently rolled over (second panel). More important to the index is the structural underperformance in earnings growth (third panel). While pricing improvements seem to be helping close the gap, the industry has nearly a decade of uninterrupted earnings deficit relative to the broad market. As such, the recent recovery in the S&P soft drinks index without an accompanying EPS lift has driven sector valuations to a 40% premium to the S&P 500 (bottom panel). Considering the stalling pricing efforts, U.S. dollar strength and a generalized global trade slowdown, this seems overly optimistic. Stay underweight. The ticker symbols for the stocks in this index are: BLBG: S5SOFD - PEP, KO, MNST. ​​​​​​​ ​​​​​​​

  Overweight Household product stocks have typically performed well as retail gasoline prices have contracted; this iteration has proven no different (gasoline prices shown inverted, top panel). This boon to consumers has supported an uptick of the consumer’s wallet being deployed to household products which should translate into top line support for these equities (second panel). Curiously, the end of the bear market for this sector coincided with a moderating of S&P household products profit margins from their historically high level (third panel). The market is likely seeing ahead to a return to margin expansion. As noted above, the demand environment appears robust and, with commodity and labor costs well contained (bottom panel), things should continue looking up for the sector, especially given the recent success constituents have had in raising selling prices. Bottom Line: Earnings growth looks set to reaccelerate in the S&P household products index; we reiterate our overweight recommendation. The ticker symbols for the stocks in this index are: BLBG: S5HOPR – PG, CL, KMB, CLX and CHD.  

Overweight In this week’s Weekly Report, we highlight three macro factors that, should they sustain their recent trajectories, would serve to catalyze the semi equipment group. First, trade policy uncertainty has dealt a blow to this tech subindex (trade policy uncertainty shown inverted, top panel). Not only are 90% of sales foreign sourced, but a large chunk is also China-related sales. Second, emerging market manufacturing PMIs are showing some signs of life, underscoring that semi equipment demand may turn out to be marginally less grim than currently anticipated (second panel). Third, while global semi sales will continue to decelerate for the next three-to-six months, the semi market is functioning as if the inventory liquidation cycle is in the later innings (third and bottom panel). Net, we think the drubbing in the S&P semi equipment index is overdone and even a modest improvement on either the trade policy, industry demand or currency fronts could result in a reflex rebound. Bottom Line: We lifted the S&P semi equipment index from underweight to overweight on Monday, as a tactical move for the next three-to-six months; please see Monday’s Weekly Report for more details. The ticker symbols for the stocks in this index are: BLBG: S5SEEQ – AMAT, LRCX and KLAC.

Highlights Portfolio Strategy The drubbing in the S&P semi equipment index is overdone and even a modest improvement on either the trade policy, industry demand or currency fronts could result in a reflex rebound, warranting an above benchmark allocation. An oil price rebound on the back of a more balanced supply/demand backdrop, a continuation in the global capex energy upcycle and compelling relative valuations all suggest that the path of least resistance is higher for oil majors. Recent Changes Boost the S&P Semiconductor Equipment index to overweight today, on a tactical three-to-six month time horizon. Table 1 Feature Equities attempted to stage a recovery last week and are in a triple bottom technical formation, still consolidating the October tremor. The Fed meeting later this week will likely prove a catalyst on the monetary policy front, especially if the closely watched FOMC median dots decrease for 2019 as the bond market has been expecting. As we mentioned in our 2019 High-Conviction calls Report two weeks ago,1 the Fed will dominate markets next year and any dovish change in interest rate expectations will breathe a sigh of relief into the SPX. Given the heightened volatility and violent recent equity market oscillations, it is important to separate the noise from the actual signal. While distinguishing between the two is hard at times, we are relying on a few key indicators to aid us in this process. First, our S&P 500 EPS growth model is still expanding near the 10% mark for next year as clearly 25% EPS growth is not sustainable. While the risk is that this growth rate decelerates further, as long as EPS do not contract next year, stock prices should recover (Chart 1). From a macro perspective, at this stage of the cycle with nominal GDP growth between 4-5%, organic EPS growth should at least mimic nominal output growth. Tack on a 2% buyback yield or artificial EPS growth and attaining a 7% EPS growth rate is likely next year. Second, while the 5/2 and 5/3 yield curve slopes have inverted and we heed these signals, the 10/2 and the Fed’s spread (2-year yield minus the fed funds rate) have yet to invert. Historically, the most significant yield curve signals for the equity market are when simultaneously all the different yield curve slopes are inverted. While everyone is infatuated with the yield curve inversion implications of recession, we are laser focused on the interplay between the yield curve and stock market peaks. Importantly, typically the 10/2 yield curve inversion occurs before stock market peaks. Going back to the mid-1960s there has been only one time when the stock market peaked prior to the yield curve inversion, in 1973: the SPX crested on January 11 and the yield curve inverted on January 16 (due to lack of data we use the effective fed funds rate instead of the 2-year yield prior to 1976). In all the other iterations, the yield curve inverts prior to the stock market top. Even in 1998 the yield curve inverted in late-May and the SPX peaked in mid-July before suffering a 20% drawdown. Similarly, on February 2, 2000 the yield curve inverted and on March 24, 2000 the SPX topped out for the cycle. Chart 3… And Then The SPX Peaks In other words, the yield curve inversion is a leading indicator and once the curve inverts, it signals that the stock market highpoint will follow soon thereafter (Charts 2 & 3). The broad market tops on average 248 days (median 77 days) following the yield curve inversion (Table 2), though the large variability in each iteration limits the usefulness of this average as an accurate predictor. Nevertheless, the implication remains that the SPX has yet to peak for the cycle. Table 2Yield Curve Inversions And S&P 500 Peaks Third, a slew of economically sensitive indicators have troughed. Sweden’s PMI and Swedish stock market relative performance have been in a V-shaped recovery. As we highlighted earlier this week,2 Sweden is a small open economy and it is likely sniffing out an improvement in global export volume growth and a likely de-escalation in the U.S./China trade tussle. EM FX, the CRB raw industrials commodities index, the Baltic Dry Index and semi equipment stocks (see more details in the next section) all suggest that the worst is over, and global trade will likely resume its advance in the coming months (Chart 4). Chart 4Hyper-sensitive Indicators Sniffing Out A Trough? Finally, inflation is coming off the boil and will likely decelerate in the months ahead courtesy of the fall in WTI crude oil prices. Were oil to move sideways from here, headline inflation would decelerate further, likely overwhelming core CPI (Chart 5). This is significant, as it could serve as a monetary policy catalyst. Put differently, decelerating inflation may cause the Fed to reconsider the pace of its interest rate hikes. A pause in the tightening cycle in March 2019 would be a welcome development for stocks, especially if the fed funds rate is nearing the terminal rate as we recently highlighted in our trough-to-peak fed funds rate tightening cycle analysis.3 Chart 5Inflation Will Decelerate Adding it all up, our still expanding SPX EPS growth model, a lack of a 10/2 yield curve inversion, a trough in a number of economically sensitive indicators and the potential for a temporary Fed hike pause in March next year, all signal that the equity bull market is not over and fresh all-time highs are looming in 2019. This week we are upgrading, on a tactical basis, a bombed out tech subgroup, and updating our view on a deep cyclical index. Semi Equipment: Enough Is Enough We are lifting exposure in the niche S&P semi equipment index from underweight to a modest overweight. Putting this in perspective, this small index comprises only 1.5% of the tech universe and commands a mere 0.3% weight in the S&P 500. There are high odds that most of the carnage in semi equipment stocks is already reflected in the violent swing of the sell side community from extreme bullishness up until August of this year to the current extreme bearishness. As a reminder, the S&P semi equipment index was part of U.S. Equity Strategy’s high-conviction underweight call revealed in November 27, 2017 when the sell-side could not have enough of semi equipment stocks as analysts were also mesmerized last winter by the near $20,000/bitcoin related mania.4 This timing coincided with the peak in performance of this hypersensitive early-cyclical tech index (Chart 6). Chart 6Extreme Bearishness... To get a sense of how far the pendulum has swung on the bearish camp, we note the following: The relative 12-month forward EPS growth has deflated from positive 60% to negative 20% (Chart 6). The index’s forward P/E is trading at a 40% discount to the SPX, relative 5-year EPS growth estimates are near previous troughs and even compared to the overall tech sector; semi equipment long-term EPS growth is now forecast to trail their tech brethren (Chart 7). Even forward sales growth has collapsed, falling to a multi-year low. Analysts now expect an outright contraction in revenues to the tune of 4% or 10 percentage points below the S&P 500 (Chart 6). Net EPS revisions have also been sinking like a stone, approaching the 2012 nadir (Chart 6). Technical conditions are oversold with cyclical momentum as bad as it gets (Chart 7).  Chart 7...Reigns Beyond this overly pessimistic backdrop, there are some macro indicators that, were they to sustain their recent budding recoveries, would serve to catalyze the chip equipment group. First, trade policy uncertainty has dealt a blow to this tech subindex (trade policy uncertainty shown inverted, top panel, Chart 8). Not only are 90% of sales foreign sourced, but a large chunk is also China-related sales. Table 3 highlights the excessive sensitivity these stocks have to the U.S. dollar. In fact, the correlation with J.P. Morgan’s EM FX index is an almost perfect one (Chart 8). If President Trump is serious about striking a deal with China, then this group would enjoy a relief rally. Chart 8Potential Positive Catalysts Table 3U.S. Semi Equipment Geographical Sales Breakdown Second, emerging market manufacturing PMIs are showing some signs of life, underscoring that semi equipment demand may turn out to be marginally less grim than currently anticipated (Chart 9). Chart 9EM Green Shoots? Third, while global semi sales will continue to decelerate for the next three-to-six months, the semi market is functioning as if the inventory liquidation cycle is in the later innings, with our industry pricing power proxy plummeting 180 percentage points from peak-to-the-recent trough, just below the contraction zone (Chart 10). Chart 10Inventory Liquidation Is In Late Stages Finally, any bounce in cryptocurrencies may also serve as a positive catalyst for additional demand for the semi equipment companies that enjoy monopolies in their respective manufacturing niches (Chart 10). In sum, the drubbing in the S&P semi equipment index is overdone and even a modest improvement on either the trade policy, industry demand or currency fronts could result in a reflex rebound. Bottom Line: Lift the S&P semi equipment index from underweight to overweight today, as a tactical move for the next three-to-six months. The ticker symbols for the stocks in this index are: BLBG: S5SEEQ – AMAT, LRCX and KLAC. Oil Majors Are Holding Firm In early-February we upgraded the heavyweight integrated oil & gas energy subindex to an above benchmark allocation. Our thesis centered on a capex upcycle recovery and firming oil price backdrop that would unlock excellent value in this key energy subgroup. Since then, the relative share price ratio has moved laterally. Interestingly, this defensive energy subindex neither kept up with the steep oil price advance until the end of September, nor with the recent drubbing in crude oil prices (Chart 11). Put differently, oil majors never discounted sustainably higher oil prices, and are also refraining from extrapolating recent oil prices weakness far into the future. Chart 11Defensive Oil Equities While the Trump Administration’s flip-flop on the Iranian sanctions has injected extreme volatility into oil prices, some semblance of normality has returned to the crude oil markets as last week OPEC and Russia agreed to a production cut in order to help balance the market. Another key factor that has contributed to the recent fall in oil prices at the margin has been U.S. shale oil supplies. Roughly 30% per annum growth in U.S. crude oil production is unsustainable and, were production to remain near all-time highs and move sideways in 2019, then the growth rate would fall back to the zero line. Such a paring back in the growth rate, along with OPEC/Russia discipline, would likely balance the oil market and pave the way for an oil price recovery (oil production shown inverted, Chart 12).   Chart 12U.S. Supply Response Is Looming Given that BCA’s Commodity & Energy Strategy service continues to forecast higher oil prices into 2019, the S&P integrated oil & gas index should stage a sustainable rebound next year. While the recent swift drop in oil prices is jeopardizing the still recovering capital expenditure cycle, we doubt $50/bbl oil would make current projects uneconomical and result in mothballing or outright canceling of ongoing oil exploration projects (Chart 13). Granted, a big assumption is that oil prices at least hold near the current level and do not suffer a relapse to the early-2016 lows. Historically, rising oil exploration outlays and integrated oil & gas share prices move in lock step and the current message is to expect a rebound in the latter (Chart 14). Chart 13Low Odds Of A Total... Chart 14...Capex Collapse Finally, sell-side analysts are throwing in the towel. Net earnings revisions have taken a beating of late, which is positive from a contrary point of view (second panel, Chart 15). Relative valuations are extremely compelling on a number of metrics including relative price-to-book, price-to-sales and relative forward price-to-earnings (third panel, Chart 15). Tack on a near 200bps positive delta in the dividend yield versus the broad market and yield hungry investors will also seek the relative safety of this defensive energy subindex (bottom panel, Chart 15). Chart 15Integrated Stocks Are On Sale Netting it all out, an oil price rebound on the back of a more balanced supply/demand backdrop, a continuation in the global capex energy upcycle and compelling relative valuations all suggest that the path of least resistance is higher for oil majors. Bottom Line: Stay overweight the heavyweight S&P integrated oil & gas energy subindex. The ticker symbols for the stocks in this index are: BLBG: S5IOIL – XOM, CVX and OXY. Anastasios Avgeriou, Vice President U.S. Equity Strategy anastasios@bcaresearch.com Footnotes 1      Please see BCA U.S. Equity Strategy Weekly Report, “2019 Key Views: High-Conviction Calls,” dated December 3, 2018, available at uses.bcaresearch.com. 2      Please see BCA U.S. Equity Strategy Sector Insights, “Can Sweden Lead The SPX?” dated December 12, 2018, available at uses.bcaresearch.com. 3      Please see BCA U.S. Equity Strategy Weekly Report, “2019 Key Views: High-Conviction Calls,” dated December 3, 2018, available at uses.bcaresearch.com. 4      Please see BCA U.S. Equity Strategy Weekly Report, “2018 High-Conviction Calls,” dated November 27, 2017, available at uses.bcaresearch.com.   Current Recommendations Current Trades Size And Style Views Favor value over growth Favor large over small caps

Special Report

Dear Client, I have been on the road visiting clients in St. Louis, Minneapolis, and Chicago this week. Instead of our regular Weekly Report, we are sending you a Special Report on European bank stocks written by my colleague Xiaoli Tang from our Global Asset Allocation service. In advance of the holiday season, we will be publishing next week’s report summarizing our key views for 2019 on Tuesday morning. Best regards, Peter Berezin, Chief Global Strategist   Highlights Euro area bank profits are driven more by economic growth than monetary factors. This growth link explains the close correlation between the relative performance of banks within the euro area and the relative performance between euro area and U.S. equities. It also highlights the importance of euro area banks to global asset allocators. Euro area banks now have attractive valuations, which are offset however by a lackluster profit outlook. Long-term investors should avoid banks in the region. Investors with a more tactical mandate and much nimbler style could use our valuation indicators to “time” their entry and exit into banks as a short-term trade. Feature Banks in the euro area have underperformed the region’s broader market by about 50% since March 2009, when global equities reached their financial crisis lows. In the same period, the overall euro area equity index also underperformed U.S. equities by about 50% in common-currency terms. In fact, the relative performance of euro area banks to the euro area broad market has been joined at the hip with the relative performance of euro area equities vs. U.S. equities over the past decade (Chart 1, panel 1). Getting the bank view right in the euro area is therefore an important input into our country allocation decision between U.S. and euro area equities. Chart 1Is It Time To Buy Euro Area Banks? With a more than 50% discount to the broad market in terms of price-to-book (P/B), banks are now looking very cheap. However, banks in the euro area have always traded at a discount to the broader market on an absolute basis. Currently the relative P/B reading of 0.45 is only slightly lower than the 3-year average of 0.47 – still higher than the lower band of the valuation range (Chart 1, panel 2). The relative dividend yield also gives similar information (Chart 1, panel 3). Historically, when the relative P/B discount hits the lower band and the relative dividend yield hits the upper band, a rebound in relative return performance could be expected. In order to support sustainable outperformance, however, banks need to have sustained profitability. In this Special Report, we delve into the fundamental factors that affect a bank’s profit outlook such as capital position, loan growth and non-performing loan situation to determine if banks in the euro area are cheap for a reason, or are about to embark on a period of sustainable outperformance. What Drives Bank Share Performance? According to research published in BCA’s Global Asset Allocation Special Report on July 27, 2017,1 it is clear that return on equity (ROE) has historically been closely correlated with the performance of bank shares, especially on a relative-to-the-broad-market basis (Chart 2, panel 1). Chart 2Euro Area Bank Performance Drivers The recovery of ROE has so far been tepid. This is largely a result of deleveraging in the banking system and very low asset utilization, because both return on assets and net profit margins have recovered strongly (Chart 2, panels 2 and 3). Since the Global Financial Crisis (GFC), euro area banks have steadily reduced leverage to a multi-decade low, while asset utilization has been in a downtrend since the 1990s – even though this ratio seems to have been stabilizing over the past few years. Profit margins reached a historical high of 12.7% in Q4/2006, then collapsed during the GFC and reached a low of 0.34% in Q3/2009. The subsequent rebound in profit margins was short-circuited by the euro debt crisis, causing net profit margins to plummet into negative territory, reaching a historical low of -7.6% in Q3/2012. They have recovered strongly since, reaching 9.8% in Q3/2018, not far from the 2006 peak margin level. As such, banks have to increase their leverage and asset utilization in order to generate higher ROE. This also means they need to increase their asset base and take on more risk. Do banks in the euro area have the ability to do so? Capital Adequacy Vs. Deleveraging The capital adequacy ratio (CAR), the ratio of a bank’s regulatory capital to its risk-weighted assets, measures a bank’s ability to absorb shocks. As shown in Chart 3, banks in all countries have steadily increased this ratio since the GFC. Banks in Ireland, the Netherlands and Finland have the highest CAR values, but they have all come down from their respective peak levels. On the other hand, Spanish, Italian and Portuguese banks have the lowest CAR readings, though they are still improving. French banks stand out because their capital adequacy ratio has been in a steady uptrend with the least volatility. Chart 3Improving Capital Position, But... Looking at CAR alone, however, could be misleading when trying to gauge a bank’s capital situation. In fact, the generally rising capital adequacy ratio has mainly been achieved through the reduction of risk-weighted assets in all countries except France (Chart 4). Chart 4...With Massive Leverage French banks’ risk-weighted assets have been more or less stable since 2006, with a small decline into 2015 and a gradual increase since. Belgian banks have also experienced similar asset growth as French banks over the past few years, though that is after massive deleveraging occurred between 2007 and 2014 (Chart 4, panel 1). Both Spanish and Italian banks tried to grow assets in 2014 after several years of deleveraging, but the attempt was short-lived as both resumed asset reduction, starting in 2015 (Chart 4, panel 2). Dutch banks seem to have stabilized their asset base since 2014, while Irish banks, which cut half their asset base between 2010 and 2014, have continued to deleverage, albeit at a much slower pace (Chart 4, panel 3). The deleveraging process in Portuguese and Finish banks has been ongoing since 2010, and it seems that the painful deleveraging process may have come to a stage of stabilization (Chart 4, panel 4). In terms of regulatory capital, the numerator of the capital adequacy ratio, French banks again stand out with a steadily increasing regulatory capital base, while Dutch banks have also grown their regulatory capital base at a similar pace. The regulatory capital bases in Spanish, Italian and Belgian banks, however, have been oscillating over the past decade, while Portuguese and Irish banks’ regulatory capitals have declined significantly (Chart 5). Chart 5Regulatory Capital Growth: No Synchronization Another indicator of bank resilience, the ratio of non-performing-loans (NPLs) net of provision relative to capital, measures if a bank can write off all of its bad loans and remain solvent. How do all the banks measure up in this aspect? Even though banks in all countries now have good readings (less than 100%), both Italy and Portugal were under severe stress until only a few years ago. Despite significant improvement since, banks in these two countries still have high levels of bad loans relative to capital compared to banks in other countries in the region (Chart 6). Chart 6Bad Loans Are Well Provisioned Loan Quality Vs. Quantity The ratio of NPLs-to-gross loans provides potentially useful insights into the quality of assets. NPL ratios in France, Germany, Belgium, Austria, and Finland are all less than 5%, while those in Italy, Portugal and Ireland are higher than 10%, and Spain is in between (Chart 7). Since the peak around 2015, the NPL ratios in all countries other than Finland have come down. Compared to levels before 2006, however, bad loan ratios are still high. Chart 7NPL Ratio In addition, despite the improvement in asset quality, banks have not aggressively grown their loan books. Only banks in France and Finland have been consistently lending to their respective private sectors – along with German banks, albeit at a lesser pace. Lending to the private sector in Spain, Portugal and Ireland has in fact contracted by 40%-50% since 2008, while loan growth from banks in Italy, Austria and the Netherlands has basically been flat since the GFC, as shown in Chart 8. Chart 8Bank Loans To Private Sector Exposure To Emerging Markets Banks in the euro area are known to have a strong presence in the emerging markets. As shown in panel 1 of Chart 9, Spanish banks have more than doubled their lending to emerging markets (EM) since 2006; even after a reduction over the past two years, loans to EM still account for over 16% of total lending. This stands in stark contrast to their domestic lending, which has contracted sharply since peaking in early 2009 (Chart 8, panel 3). Portuguese banks share similar patterns to Spanish banks in terms of loan growth to EM and domestically, however, their absolute amounts have been much smaller (Chart 8, panel 3 and Chart 9, panel 2). Dutch banks shrank their loan books to EM right after the GFC but have been gradually building them back up since 2011, while Austrian banks have been steadily reducing the pace of their lending to EM (Chart 9, panels 3 and 4). Chart 9Bank Exposure To EM After the turbulence earlier this year in Turkey and Argentina, BCA’s Global Investment Strategy and Foreign Exchange Strategy services identified six countries (Argentina, Turkey, Colombia, Brazil, Mexico, and Chile) as the most vulnerable to catching the “Turkish Flu,” based on the following factors: current account balance, net international investment position, external debt, external debt-service obligation, external funding requirements, private-sector savings/investment, private-sector debt, government budget balance, government debt, foreign ownership of local-currency bonds, and inflation2 (Table 1). The vulnerability of Latin America highlights the poor performance of Spanish banks, given their heavy exposure to the region. For example, Banco Santander, the largest Spanish bank and also the largest component in the euro area bank index, has aggressively expanded into Latin America to beef up asset utilization and return on assets. However, loan quality from Latin America has been much lower, as evidenced by the much-higher percentage of bad loan provisions from the region compared to its share of loans. Currently, loans to Latin America account for about 18% of total lending, yet bad loan provisions account for about 42% of total provisions (Chart 10). Chart 10Banco Santander: More Like An EM Bank Exposure To Italian Government Debt The fiscal budget saga in Italy has been a negative factor impacting euro area assets, especially Italian banks. Italian banks have been large buyers of Italian government debt securities, reaching over 400 billion euros at the peak and accounting for about a quarter of total debt securities. Following the European Central Bank’s quantitative easing program (QE) that started in March 2015, Italian banks’ share of government debt holdings subsequently dropped to about 18% by the end of 2017. In 2018, however, Italian banks purchased more government bonds to a level of 393.8 billion euros as of September 2018, or about 20% of the overall debt securities outstanding – only a tad lower than the peak level before the QE program (Chart 11). Chart 11Italian Debt By Type Of Investor Now the ECB’s QE program is expected to come to an end soon. With government debt securities holdings accounting for 24% of tier 1 capital in Italian banks, (Chart 12), investors should pay close attention to the “Doom Loop,” i.e. when weakening government bonds threaten to topple the banks that own those bonds, the banks are forced to unload the bond holdings, which in turn pushes the government into additional fiscal stress. Chart 12The Doom Loop Moreover, Italian banks are not the only banks in the euro area which are exposed to Italian government debt. According to the European Banking Authority’s 2017 Transparency Exercise, French and Spanish banks held 44 billion euros and 29 billion euros of Italian debt, respectively. For example, the largest French bank, BNP Paribas (BNP), which is the second-largest component by market cap in the euro area bank index, has gradually added more Italian government debt securities since 2015 (when the ECB started buying Italian bonds) following a large reduction in 2011 (Chart 13). Investment Implications The euro area banks and diversified financial sector indices are currently mostly dominated by Spain (30%), France (25%) and Italy (15%), which all have grown at the expense of the German banks over the past two decades (Chart 14). Chart 14Euro Area Bank Index: High Concentration From a fundamental perspective, only French banks have both good-quality assets with decent and steady loan growth; the largest weight – Spanish banks – has experienced negative loan growth domestically while expanding aggressively to emerging markets up until 2017. Some may argue that exposure to Italian debt and emerging markets may have already been fully priced in, given the massive underperformance of the banks. This may well be true, and there could be a short-term bounce in bank stocks, given the attractive valuation metrics. For long-term investors, however, such a bounce may not be captured easily. We suggest long-term investors stay away from euro area banks, in line with our regional equity view of favoring the U.S. over the euro area. Why? Because cheap valuations are offset by lackluster profit outlook at a time when growth is slowing and monetary policy is becoming less accommodative (Charts 15A and 15B). Relative earnings growth for both banks and diversified financials are closely tied to the euro area PMI, the leading indicator for economic growth (Charts 15A and 15B, panel 2). This growth link explains why the banks’ relative performance in the euro area has such a close correlation with the performance of euro area equities relative to their U.S. peers. Chart 15APoor Profit Outlook For Banks Chart 15BPoor Profit Outlook For Diversified Financials For investors with a more tactical mandate and much nimbler style, however, Chart 1 could be used as a guide to “time” an entry and exit to the industry: go overweight when the relative price-to-book reaches the lower band and relative dividend yield reaches the upper band, and vice versa.   Xiaoli Tang, Associate Vice President xiaoliT@bcaresearch.com Appendix 1 Euro Area Bank Indexes Different index providers have different classifications and compositions for banks, based on their different respective index methodologies.3, 4 GAA uses the MSCI All Country Equity index as its global equity benchmark. As such, whenever possible, we use the MSCI indexes in our research work. When data is not available from MSCI, however, we also use the Datastream Thomson Reuters (Datastream) index. In this Special Report, we have combined the MSCI “Bank Index” and “Diversified Financials Index” into one Aggregate Bank Index for one reason: MSCI reclassified Deutsche Bank as a “diversified financial” from a “bank” in 2003. Appendix Table 1 and Appendix Table 2 show the comparisons between the Datastream Bank Index and the MSCI Aggregate Bank Index. Even though Datastream includes 16 countries and MSCI includes only eight countries, both indexes are quite concentrated in Spain, France, Italy and the Netherlands. These four countries account for 77.4% of the Datastream Bank Index with 34 stocks, while they account for 78.8% of the MSCI aggregate bank index with 19 stocks. What’s more, the top five stocks are the same in both indexes, but they account for half of the MSCI Aggregate Bank Index and only 42% of the Datastream Bank Index.   Consequently, while the two indexes are quite similar, users should be aware of the differences. For example, since March 2009, the MSCI Aggregate Bank index has underperformed the broader index by 48%, but Datastream banks have underperformed the broad index by 55%, as shown in Appendix Chart 1. Footnotes   1 Please see Global Asset Allocation Special Report, “What Drives Bank Share Performance?” dated July 27, 2017 available at gaa.bcaresearch.com. 2 Please see Global Investment Strategy Weekly Report, “Hot Dollar, Cold Turkey,” dated August 17, 2018, available at gis.bcaresearch.com. 3 Please see https://www.msci.com/eqb/methodology/meth_docs/MSCI_GIMIMethodology_Nov2018.pdf 4 Please see http://www.datastream.jp/wp/wp-content/uploads/2017/02/DatastreamGlobalEquityIndicesUGissue05.pdf  

Special Report

Highlights On the bright side, Malaysia’s structural backdrop is improving notably, especially in the semiconductors segment. Yet the cyclical growth outlook remains downbeat. While we are maintaining a market-weight allocation to Malaysian equities within an EM equity portfolio, we are putting this bourse on our upgrade watch list. As a play on the ameliorating structural outlook, we recommend an overweight position in Malaysian small-cap stocks relative to the EM universe – both the small-cap and overall equity benchmarks. Feature Malaysian stocks have performed quite poorly in recent years: the equity index, in U.S. dollars, is close to its 2016 lows in absolute terms, and relative to the emerging markets (EM) benchmark, it is not far from the lows of last decade (Chart I-1). Chart I-1Malaysian Stocks & Commodities Prices: Tight Relationship Odds are that a structural bottom in this bourse’s relative performance versus the EM index may have been reached. Hence, we are putting Malaysian equities on our upgrade watch list while maintaining a market-weight allocation due to tactical considerations. On the negative side, the past credit excesses have not been recognized and provisioned for by Malaysian commercial banks. The latter account for a notable 34% of the MSCI Malaysia index, and they will be a drag on this bourse's performance. Absolute performance also still hinges on global growth, commodities prices and the overall direction of Asian/EM markets. We are still negative on these parameters. Critically, there are various signs indicating an ameliorating structural backdrop in Malaysia. The country is undergoing notable improvements in the semiconductor sector, thereby reducing its dependence on commodities and increasing its exposure to a high-value industry. To capitalize on this theme of an improving structural backdrop, we are recommending an overweight position in Malaysian small-cap stocks relative to the EM universe – both the small-cap and overall equity benchmarks. Shifting Away From Commodities And Toward Electronics Parting Ways With Commodities Malaysia and its financial markets have been very exposed to commodities prices over the past 15 years or so (Chart I-1, top panel). Nevertheless, the country seems to be shifting away from its considerable reliance on the resource sector and moving into other value-added segments: in particular, semiconductors and technology. Such a structural shift – if successful – would be an extremely positive development as it would lead to rising productivity gains and higher per capita income growth. In short, the country would be able to achieve higher rates of sustainable non-inflationary growth, feeding into a sustainable bull market in Malaysian equities. Several points are noteworthy in this regard: The real output of crude and petroleum products as well as palm oil are declining sharply relative to the economy’s real total output (Chart I-2). Chart I-2Malaysia's Commodities Output Is Falling In Importance Exports volumes of palm oil, crude oil and natural gas have all been falling relative to Malaysia’s total overseas shipment volumes (Chart I-3). Chart I-3Commodities Export Volumes Are Declining In Relative Terms Crude oil, gas, and palm oil now account for 4%, 5%, and 7% of total exports in value terms, respectively. Crucially, not only is the importance of commodities in the overall Malaysian economy diminishing in volume terms, it is also falling in nominal terms due to low resource prices. For instance, net export revenues from fuel (i.e. crude oil, petroleum and natural gas) have fallen from US$18 billion in 2013 to US$5 billion today (Chart I-4, top panel). Chart I-4Commodities' Net Export Revenues Are Also Diminishing Meanwhile, net exports of palm oil (and other plant-based fats) have dropped from US$20 billion to US$10 billion (Chart I-4, bottom panel). Improvement In High-Value-Added Manufacturing There are also some positive structural signs taking place in the Malaysian economy that are signaling an improvement in productivity and competitiveness: Malaysian export volumes of machinery and transport equipment are expanding in absolute terms as well as relative to overall export volumes (Chart I-5, top and middle panels). Chart I-5Malaysia's Machinery Exports Are Rocking Remarkably, Malaysian aggregate export volumes are quickly regaining lost global market share (Chart I-5, bottom panel). Further, the ratio of exports to imports has hit a structural bottom and is slowly picking up in volume terms (Chart I-6). Chart I-6Malaysian Overall Exports Are Regaining Lost Market Share This suggests some improvements in the competitiveness of domestic industries is slowly underway. Meanwhile, Malaysian high-skill and technology intensive exports as a share of global high-tech exports seem to have made a major bottom in U.S. dollar terms and will begin to rise (Chart I-7). Chart I-7Bottom In Malaysia's High-Tech Global Share? Advanced education enrollment is high and improving – and is only outpaced by Korea and China in emerging Asia (Chart I-8). Importantly, Malaysia has among the best demographics of mainstream developing countries. The working age population as a share of the total population will continue to be high all the way to 2040. Chart I-8Malaysians Like Going To School Malaysian expenditures on R&D have also been on the rise, outpacing a lot of other countries in the region (Chart I-9, top panel). R&D expenditures in Malaysia could also be catching up to Singapore’s (Chart I-9, bottom panel). Chart I-9Malaysia's Expenditure On R&D Is Rising In line with these positives, net FDIs into Malaysia have been rising briskly (Chart I-10). Importantly, these investments have been driven by European companies, meaning the latter are transferring valuable technological know-how to Malaysia. Chart I-10Net FDIs Are Rising The Malaysian ringgit is cheap (Chart I-11) and has reached almost two-decade lows against many Asian currencies. This makes Malaysia increasingly more competitive. Chart I-11The Ringgit Is Cheap Finally, our colleagues from the Geopolitical Strategy team believe that the recently elected Pakatan Harapan government will improve governance and transparency, which had significantly deteriorated under Najib Razak’s rule. A Marriage To Electronics Malaysia is attempting to reestablish itself as a major semiconductor hub in the region. Remarkably, after declining for 15 years, semiconductor exports are finally rising as a share of GDP (Chart I-12) and Malaysian semiconductor exports are outperforming those of its neighbors. Chart I-12Malaysian Semiconductor Exports Are Booming The Malaysian government since 2010, has identified the semiconductor sector as a key area for development and prosperity. In turn, it has been introducing programs and setting up institutions to support the industry. The 2019 budget reinforces the government’s priority to develop the sector. Several anecdotal observations confirm that Malaysia is moving up the value chain in the semiconductor industry, and is going beyond simple testing and assembly: Growing the semiconductor cluster: The Malaysian Institute of Microelectronic Systems (MIMOS) has established a shared services platform for advanced analytical services in the semiconductor industry to provide support to Malaysian semiconductor companies. The Economic Industrial Design Centre (EIDC) is also providing support to SMEs in order to enhance their efficiency. Similarly, the Semiconductor Fabrication Association of Malaysia (SFAM) has been partnering with local universities to enhance their engineering programs and offer training, internships and research opportunities for students. Developing home-grown semiconductors: In 2015, Malaysian public institutions in partnership with private companies developed the Green Motion Controller (GMS), an integrated circuit that reduces energy consumption. This semiconductor is an energy efficient controller that carries applications in hybrid cars and air conditioners, among other things. Nanotechnology: NanoMalaysia – a nanotechnology commercialization agency – is providing services to SMEs and start-ups to help increase their competitiveness by enabling them to upgrade to more efficient production methods. Light-emitting Diode (LED) manufacturing: Malaysia is becoming a hub for the manufacturing of energy efficient LED chips. This is the result of OSRAM’s – a German light manufacturer – large investment in a high-tech production facility. There are early signs already that the above developments are beginning to bear results. Chart I-13 shows that the difference between exports and imports of semiconductors (in U.S. dollars) have been surging. This shows Malaysia is able to add greater value to the semiconductors it imports and then re-exports. Chart I-13Malaysia Adds Value To The Semis It Imports Bottom Line: Commodities are declining in importance to the Malaysian economy. Meanwhile, Malaysia’s structural backdrop is improving as the semiconductor and hardware technology segments are rising in prominence. Cyclical Weakness Despite the positive structural backdrop, Malaysia’s cyclical outlook remains challenging. Our view on commodities and global trade continues to be negative. Not only are commodities prices deflating but semiconductor prices are also falling, and their global shipments are weakening (Chart I-14). Chart I-14Cyclical Weakness In Global Semiconductor Cycle The epicenter of the global trade slowdown, however, will be in Chinese construction activity. Consequently, industrial resources prices are more vulnerable than electronics in this global growth downturn. The above deflationary forces would negatively shock Asia’s growth outlook, and consequently Malaysian growth as well: The top panel of Chart I-15 shows that Malaysian narrow money growth has already rolled over decisively and is foreshadowing weaker bank loan growth. Chart I-15Malaysian Domestic Growth Set To Weaken Slower bank loan growth will weaken purchasing power and impact domestic consumption. The middle panel of Chart I-15 shows that car sales – having surged this summer because of the abolishment of the GST – are weakening anew. Malaysian companies and banks have among the largest foreign currency debt loads (Table I-1). We expect more currency depreciation in Malaysia, as we do in EM overall. This will make foreign currency debt more expensive to service, and consequently dampen companies’ and banks’ appetites for expansion. Table I-1Malaysia's External Debt Breakdown Finally, the real estate sector remains depressed. Property volume sales are contracting and have dropped to 2008 levels, and housing construction approvals are slumping (Chart I-16). Chart I-16Malaysia's Property Sector Is Depressed While this means that cleansing has been taking place in the property sector, the banking sector has not recognized NPLs and remains the weakest link in the Malaysian economy. Specifically, the top panel of Chart I-17 illustrates that the NPLs in the banking system still stand at a mere 1.5%. This is in spite of the fact that since 2009, non-financial private sector credit to GDP has risen significantly. Therefore, the true level of NPLs is probably considerably higher. Chart I-17Malaysian Banks Are Under-Provisioned Further, Malaysian banks have been lowering provisions to boost profits (Chart I-17, bottom panel). This is unsustainable. As growth weakens, Malaysian banks will see their NPLs rise and will need to raise provisions. Chart I-18 demonstrates that if provisions rise by 20%, bank operating earnings will contract and bank share prices would fall. Chart I-18Malaysian Banks' Share Prices Will Fall Bottom Line: Malaysia’s cyclical growth outlook is still feeble, with the banking system being the weakest link. Banks’ large weight in the equity index makes this bourse still vulnerable in the coming months. Optimal Macro Policy Mix Fiscal Consolidation… Fiscal policy is set to be tighter as per the Malaysian government budget announced on November 2 and its preference to pursue fiscal consolidation to reduce the deficit. The budget projects only a slight increase in expenditures in 2019, which means it will likely slowdown from 8% currently (Chart I-19). Chart I-19Government Expenditure Growth Will Soften The government will also recognize public-sector liabilities not presently shown on its balance sheet and strengthen both transparency and administrative efficiency. Critically, the budget also includes strategies to support the entrepreneurial part of the economy. Overall, this budget bodes very well for the structural outlook. Yet it will not support growth cyclically. …To Be Offset By Easy Monetary Policy Despite continued currency weakness, the Malaysian monetary authorities will not be in a hurry to raise interest rates to defend the ringgit. This is in contrast with other central banks in the region like Indonesia and the Philippines. This is presently an optimal policy mix for Malaysia and is positive for the stock market’s relative performance versus its counterparts in many other EMs. Malaysia’s structural inflation is low: core inflation hovers around zero. Therefore, the central bank will neither raise interest rates nor sell its foreign exchange reserves to defend the currency. Both currency depreciation and low interest rates are needed to mitigate the downturn in exports as well as offset fiscal consolidation. In the meantime, the ringgit is unlikely to depreciate in a sudden and vicious manner but rather will likely fall gradually. First, the current account will remain in surplus, even as global trade contracts. The basis is that if Malaysian exports fall, imports will simultaneously follow. The country imports a lot of intermediate goods to then process and re-export. Second, Malaysia is unlikely to witness pronounced capital flight as occurred in 2015. The new government has increased confidence in the economy among both locals and foreigners. In addition, net portfolio investments have been negative for a while. This means that a large amount of foreign capital has exited already, reducing the risk of further outflows. What’s more, foreign ownership of local currency bonds has fallen from 33% in June 2016 to 24% today. Moreover, at 28% of market cap, foreign ownership of equities is among the lowest in EM. These also limit potential foreign selling. Bottom Line: Policymakers are adopting a wise policy mix for the economy at the current juncture: tight fiscal and easy monetary policies. This is structurally positive, even if it does not preclude cyclical weakness. Investment Conclusions Weighing structural positives versus the cyclical growth weakness and the unhealthy banking system, we are maintaining a market-weight allocation to Malaysian stocks within the EM universe, but are placing this bourse on our upgrade watch list. We need to see a selloff in bank stocks before we upgrade it to overweight. Within Malaysian equities, we recommend shorting/underweighting banks and going long/overweighting small cap stocks. To capitalize on Malaysia’s improving structural growth outlook, we recommend buying Malaysian small caps, but hedging positions by shorting the EM aggregate or small-cap indexes. The ringgit is poised to depreciate further versus the U.S. dollar along with other EM/Asian currencies. We continue to short the ringgit versus the greenback. With respect to sovereign credit and local government bonds, dedicated portfolios should currently have a market-weight allocation. The negative cyclical growth outlook is offset by the right macro policy mix and improving growth potential.   Ayman Kawtharani, Associate Editor ayman@bcaresearch.com​​​​​​​ Equity Recommendations Fixed-Income, Credit And Currency Recommendations