In This Issue
The unprecedented coronavirus panic and economic shutdown have precipitated extreme financial-market reactions only seen in previous crises such as in 2008–2009 and 1929–1933. For example, return correlations of large-capitalization stocks (Exhibit 1) as measured by analysts at Empirical Research Partners rose to levels a bit above those seen in the Great Financial Crisis (GFC) of 2008–2009 and only matched by those observed at the beginning of the Great Depression. In short, the coronavirus selloff compares with the other two greatest financial panics of the past 100 years. When correlations are so high, it means there's indiscriminate selling of everything—good and bad—as in "throwing out the baby with the bathwater."
Exhibit 1: Correlations Reveal Full Blown Panic.
Note: Data represents detrended daily return correlations, measured over 21-day windows for large-capitalization stocks. Computed as the deviation off the 10-year rolling average; capitalization-weighted data. Grey bars are recessions. Sources: National Bureau of Economic Research; Empirical Research Partners. Data as of March 2020.
Inevitably, the enormity of what has happened in financial markets has prompted comparisons with the aftermath of the Great Depression collapse. This has led some prominent money managers and strategists to warn that "a prolonged period of hard times and new lows in equity prices is ahead." This is possible, but not necessarily inevitable. The outcomes ahead will depend on the particular circumstances of our current problems and, importantly, on how policymakers respond to them. When we assess both the nature of the problem and the mitigation policies being applied, we conclude that the coronavirus bear market will eventually turn out to be more comparable to the 1987 bear market than that of the Great Depression.
The S&P 500 peaked at about 336 on August 25, 1987, and was relatively flat until early October, when it began to fall and then collapsed on October 19, 1987, in the biggest one-day percentage drop ever seen. Almost all the 33% decline occurred between October 5 and October 19. Similarly, the S&P 500 recently peaked just shy of 3,400 in the third week of February 2020 and fell to a low near 2,200 on March 23 for a decline also of about 33%. Both declines were unusually fast, and, as shown in Exhibit 1, the high correlation observed in 1987 also almost matches the correlation observed in the current bear market, a key metric of widespread panic.
Most economists predicted a recession would occur after the 1987 drop in equity prices. Instead, a quick policy response and a strong underlying economy defied their predictions and the secular bull market resumed, although it took the better part of the next two years to set a fresh S&P 500 record high.
The 1987 experience was unusual in a couple of respects. First, most bear markets are preludes to recessions. However, the 1987 drop, especially its speed, was caused by new financial technology, in particular the use of "portfolio insurance," which the Brady Commission investigation of the crash primarily blamed for the unprecedented nature of the October 19 collapse.
The 2020 stock-market collapse is similarly the result of a one-off factor rather than reflecting a buildup of pre-recession weaknesses. The U.S. economy was strong when the government response to the pandemic brought it to its knees. Indeed, the Atlanta Fed's GDPNow estimate for the first quarter was tracking around 3% up until the middle of March. As in 1987, there was a specific, exogenous cause for the extremity of the selloff. Moreover, the cause is clearly identifiable, and the conditions to stop its negative impact are well understood: flatten and turn the case curve down and prevent a recurrence through social distancing, testing, therapeutics and, eventually, a vaccine.
There is growing confidence about the timeline for cases in more countries, and schedules for reopening are starting to emerge. As timeline uncertainty recedes, policies to tide over household and small-business incomes have already been tailored to address the labor-income and business-revenue shortfalls, essentially extending a bridge to the other side. Without this bridge, defaults and bankruptcies would likely rise to systemic levels, causing the kind of scenario that created the Great Depression.
Those looking for a Great-Depression scenario this time around are essentially predicting that these policies will fail to prevent a proliferation of defaults and an extended recession or even depression. These concerns are evident in the relative performance of the equities of companies perceived to have higher bankruptcy risk. Empirical Research analysts find that so far this year the relative returns of companies with high and low bankruptcy risk showed the biggest differential in performance compared to any year since 1965.
The answer to the question of whether there will be a lower low and an extended recession largely depends on the effectiveness of policies to bridge the financing gap over the shutdown period and allow the economy to reopen in a still favorable environment for renewed growth. Interestingly, in the Blue Chip Survey of Economic Forecasters, 87% of economists responded right before the passage of the Coronavirus Aid, Relief, and Economic Security Act ("CARES Act") that they were doubtful that these policies would be effective. Yet, when we analyze the policy response, it is clear that it has benefited from the lessons learned both from the Great Depression and the GFC. Indeed, it's fair to say that the GFC, bad as it was, was not as bad as the Great Depression, precisely because of the lessons learned from that traumatic experience. Similarly, we doubt that the current recession will be as bad as the GFC, precisely because of the lessons learned from that experience.
The GFC unfolded in stages over more than a year, as did the policy response. The collapse of Lehman Brothers was the catalyst for the deepest phase of the crisis, and it occurred nine months into the recession. The coronavirus pandemic effect manifested in just two months. In record time, the biggest ever financial response was also set in motion to bridge the shutdown period. A second response is likely soon as the shortcomings of the first response and the need for extended benefits become apparent. This is designed, and will be bolstered as needed, to preclude the wave of bankruptcies and defaults that would create the bearish outcome some expect. Unlike the fiscal response to the GFC, this intervention is not spread out over a decade. It is bigger, quicker, and designed to directly help the specific parts of the economy that are the source of the problem. In the GFC, housing was never directly addressed, and banks were left to recover over many years even though they were at the epicenter of the crisis. In this case, money is being put directly where it is needed to bridge the gap. If more is needed, more will come.
Finally, the Fed learned from the Great Depression and then learned more from the GFC. As a result, it had emergency response programs on the shelf and brought them out immediately for this crisis. The specifics of the current crisis mean that problems were more concentrated outside of the banking sector compared to the GFC. The Fed, together with a strong banking sector, makes it possible to address the problems, and the results are encouraging. Pro-growth economic policies in place prior to the onset of the pandemic are also a positive for a quick return to growth compared to the post-GFC period.
This crisis will also provide lessons for dealing with the next crisis. Importantly, fiscal policy can be more directly and efficiently applied using the private financial system to get money directly to people instead of looking for shovel-ready projects. Businesses can use direct support to maintain operations instead of creating government make-work projects and bridges to nowhere.
These differences between prior crises and the current one suggest those waiting for the other shoe to drop may be "waiting for Godot." If the money in the CARES Act gets distributed over the next few weeks and the economy begins to reopen as the pandemic gets under control, the train will "leave the station" and the coronavirus bear market will turn out to be just a pause in a secular bull market, just as the 1987 crash proved to be.
Finally, there is the old expression "don't fight the Fed." As the expansion reignites, the Fed has pledged to maintain an accommodative policy for the indefinite future. Thus, until inflation becomes a problem, fiscal and monetary policies can work together to support economic growth. The deflationary shock from coronavirus has pushed that day of reckoning off into the future. In the meantime, the power of well-applied policy should help the economy return to growth sooner rather than later.
The severe uncertainty and acute economic effects of the coronavirus are reflected in major market moves unfolding at blinding speed. Valuation multiples, for example, nearly made a round trip from lofty pre-crisis levels to severely depressed levels and back within just one month. From its February peak, the forward price-to-earnings (P/E) multiple for the S&P 500 fell nearly 30% to a low of 13.3x, cratering from the 90th percentile to the 17th percentile. But valuations have sauntered back, and, as of April 8, the forward multiple trades at 17.6x, sitting above the historical average of 15.7x. The sheer speed of these moves reflects just how vulnerable these markets are to new information.
What's priced into the market? The answer is imprecise but the substantial and abrupt decline in global equities is consistent with an abrupt recession brought about by the exogenous shock of economies shutting down in an effort to contain the medical crisis. Negative headlines piled up as Saudi/Russian tensions contributed to multidecade lows for crude oil, unemployment claims surged beyond historical highs, and analysts rushed to mark down forecasts. At the market low on March 23, the equity risk premium hit its 99th percentile, and other measures of valuation sat very close to extreme historical ranges, illustrating a preponderance of negativity.
The S&P 500 has since retraced nearly 50% of its lows, and valuation metrics have likewise drifted higher but still remain attractive compared to bond yields. The twin efforts of coordinated fiscal and monetary policy and some green shoots for containment have improved the backdrop. With that said, there is no telling if the final bottom is in. To draw from history, trailing P/Es troughed at wide ranges (11x-18x) during prior bear markets, while markets bottomed at an average multiple of 13x to 14x to their actual trough earnings, according to BofA Global Research. While the S&P 500 has touched the range for trough multiples, it's the uncertainty of corporate earnings this year, which makes it near impossible to call a bottom, although we'd suggest lower price levels represent an opportunity to "re-risk" allocations.
Exhibit 2: Valuation Multiple Percentiles Fell Close To Their Historic Lows But Have Since Moved Up Closer To February.
Given the large uncertainty associated with this unprecedented health crisis, any signs that case numbers are climbing or that post-containment waves of infections loom could lead investors to assign lower valuation multiples to equity indices once again. Alternatively, positive developments on the medical front coupled with significant stimulus measures could shift investor focus away from plunging economics in the near term and toward 2021, where expectations for a U-shaped recovery form on the horizon and should support higher multiples and earnings. Going forward, certain secular trends including lower for longer interest rates, which decrease borrowing costs, and the technological revolution across sectors, which can help to minimize volatility within earnings and margins, help support valuation premiums relative to past periods.
What is sentiment telling us?
Valuation powers stock returns over the long run, with predictive power approaching 90% over a 10-year time period, but over the short-term markets may deviate significantly away from "fair value" and are often influenced by momentum and sentiment, both powerful forces that reflect underlying market psychology. In the midst of a major correction prompted by a medical and economic crisis, it comes as no surprise that sentiment is currently very weak.
At the moment, nearly all global equity markets are trading below their 200-day moving averages, and investors pulled a combined $348 billion out of bonds and equities while moving $658 billion into cash during the month of March, according to BofA Global Research. In addition, average fund manager cash holdings have popped to 5.1%, and various investor surveys, including Investors' Intelligence Bull/Bear Ratio, indicate a surging proportion of negative views. These represent extreme indications of bearishness. The massive uncertainties pertaining to the coronavirus outbreak underpin these concerns, but the concentrated buildup of bearishness could quickly unwind if positive developments emerge.
Diversification can be a powerful tool, but in times of severe market stress, correlations across some assets tend to spike as investors seek out liquidity. For example, correlations between U.S., European and Emerging Market (EM) equities surged through the market calamity of the first quarter. The year-to-date correlation between the S&P 500 and the MSCI Emerging Markets hit 0.97 on April 6, after averaging 0.65 in 2019. A similar spike has occurred across market caps and within indices. According to Empirical Research Partners, the three-week rolling daily return correlation for large-cap U.S. equities is near a 90-year high. Greater diversification benefits have been found across broader asset class exposures, as less risky assets including the U.S. dollar, gold and Treasuries have somewhat counterbalanced equity movement. The ICE BofA Treasury Index has enjoyed a -0.86 average correlation to the S&P 500 this year, while gold has offered no patterned relationship.
Correlations illustrate the patterns of directional price movement, but dispersion describes the degree of price movement. Dispersion thus far in 2020 has been massive and could present a major opportunity for skilled active managers. The standard deviation between S&P 500 sectors is approximately 11% this year, meaning that there's been a wide disparity in average performance amongst sectors, and the performance spread between the best-performing sector (Tech) and worst performing sector (Energy) in Q1 was almost equal to the largest sector ranges over each of the past three full years. Market cap and style differentials have also stood out. The year-to-date dispersion of Russell 1000 equities over Russell 2000, and Russell 3000 Growth relative to Value, amounts to annualized (make actualized instead of annualized) performance differentials of approximately 51% and 48%, respectively, favoring Large Cap and Growth. Finally, valuation spreads between individual large-cap equities, intra-sectoral, and across regions all reside near historic highs. These many measures of difference illustrate a wider range of performance outcomes and a greater opportunity set for active managers.
In addition to active management, gaining exposure to the right factors can add to active returns. Quality, Minimum Volatility and Momentum factors have outperformed the broader market in each month this year, whereas the Value factor has lagged. The performance durability of these factors across both periods of market stress and surge is notable (Exhibit 3). Quality and Momentum have significantly led the broad benchmark since 2017, while High Dividend and Value have lagged. We expect Quality could likely continue to benefit as companies with sustainable earnings growth and strong balance sheets should better manage heightened periods of uncertainty.
Exhibit 3: Quality, Min Vol, and Momentum Have Outperformed the Broader Index Each Month.
Source: Chief Investment Office. Indexes Represented: MSCI USA, MSCI USA Quality, MSCI USA High Dividend, MSCI USA Min Vol, MSCI USA Momentum, MSCI USA Enhanced Value, MSCI USA Low Size. Peak date represented as 2/19/20. Trough date represented as 3/23/20. Data as of April 6, 2020. Past performance is no guarantee of future results.
Market bottoms are impossible to predict, but valuation levels are attractive for investors with a longer time horizon. Bearish investor sentiment suggests more upside than downside for stocks. Active managers can potentially take advantage of the wider opportunity set, and we believe tilting portfolios toward High Quality and Growth will be accretive in the current environment. As such, our conviction in U.S. Large Cap's has risen relative to international equities recently.
We believe that the largest U.S. banks will likely not have to suspend dividends for a few reasons. First and foremost is the Fed's annual stress test, which a foreign bank is yet to pass. Under the severely adverse scenario, banks need enough capital to absorb a 10% unemployment rate; -8% real gross domestic product (GDP); a -50% decline in equity markets; a 70-handle on the VIX; a-25% shock to residential real estate; and a -35% hit to commercial real estate while continuing to pay dividends on its common stock. That equates to $410 billion in aggregate losses and appears that bank dividends should not be suspended. Second, dividends represent only about a third of the capital returned to U.S. shareholders—approximately two-thirds comes in the form of share repurchases.
U.S. banks shutting off share buybacks (as they have through the second quarter) makes a much bigger impact on capital levels vs. suspending dividends, which would be largely symbolic. By contrast, for European banks, dividends are the lion's share of capital return, so it's more accretive to capital to turn off dividends in their case. Lastly, given the large amount of pensions and retirement funds that make up the banks' investor base, shutting off the dividend would really be harmful to the most vulnerable when they are already hurting and in need of income. Politically, the issue of share buybacks has been much more of a "hot button" than dividends, even for the president. We expect any change to capital return policy will be addressed, as it generally has, through this year's stress test. Submissions were due on April 6, and we typically get results by mid-to-late June.
Exhibit 4: Banks 2019 Comprehensive Capital Analysis and Review (CCAR) Capital Return Plans.