Factors in Stock Market Crashes and Portfolio Recoveries

Authors: Damian Handzy, Chief Commercial Officer & James Monroe, Client Consultant

Executive Summary

Equity Factor Investing matters when positioning for a portfolio recovery from a stock market crash as specific factors have a strong tendency to either outperform or underperform the market in both crashes and recoveries. As the COVID crash unfolds, equity investors who properly position their portfolios for the subsequent recovery can significantly outperform the broader market. Positioning for a portfolio recovery is best accomplished by overweighting specific factors, like Value and High Volatility, and has historically outperformed the market by as much as 8% annualized!

The COVID crash is worldwide: the US Russell 3000 is now down 19% from its high (it was down 35%), the UK FTSE 100 is down 25% (max drop was 33%) and the MSCI World is down 19% (max drop was 34%). To gain insights into what the future may hold we examined three historical crashes and their recoveries.

The 1987 crash lasted only three months, and the Global Financial Crisis (GFC) lasted 16 months.  Both were similar because their recoveries were mirror-images of their crashes: factors that did poorly in the downturn posted the best performance in the recovery. The Tech Bubble focused on one industry and heavily hit the US and some other developed markets and had a jumble of factor performances with no discernible pattern.

Our research shows:

• If we have yet to feel the depths of the COVID crash, then Low Volatility, Quality and Large Cap stocks will outperform as we continue to sink.
• If we have seen the bottom already then Value, Small Cap, Yield and High Volatility stocks should outperform.
• The COVID crash most closely resembles the GFC – correlations between factor returns in the current crash and the GFC range from 0.71 to 0.94 in the US and Europe.

But equity markets have been rallying since 23 March. Because Value and Small Cap – normally two of the most outperforming factors in recoveries – are the two worst underperformers during the current rally it, it raises questions about whether the recovery has begun or whether this rally is just a momentary rise in the middle of a larger overall market crash. One possibility is that this recovery truly is unique as it is fueled not by market demand and consumer spending but rather by governments and central banks flooding companies with cash.

The Wall Street Journal examined this rally in a recent article that was partially based on our Q1 2020 Factor Performance Analysis. Whether the market has bottomed out or not, once it does start its recovery, history would tell us that Value stocks should do very well. But this crash has not been like any other crash we’ve seen. Who would have predicted, just 2 months ago, that volatile Growth stocks would be defensive?

The PDF version of this article can be downloaded here.

Research Approach

We examine how different equity sub-factors behaved in stock market crashes and in subsequent portfolio recoveries to shed light on how investors should position their portfolios in the near-term.

We examined the 1987 crash, the Tech Bubble and the Global Financial Crisis (GFC) by measuring the returns posted by factor-specific portfolios made of stocks with the highest scores of each of Style Analytics’ standard 23 sub-factors. To qualify for entry into a factor portfolio, a stock must rank in the top 20% of the factor score. Portfolios are rebalanced monthly to ensure that the highest scoring stocks are always included. The 23 sub-factors are grouped into seven investment styles: Value, Yield, Growth, Quality, Size, Momentum and Volatility. For ease of identification these styles are consistently color-coded:

 

Figure 1: Style Analytics’ Factor Group color codes used throughout this paper. 

For each sub-factor, we sorted the constituents of our universe of stocks for the US, Europe and Emerging Markets to measure the average return of that sub-factor on a sector and country adjusted basis (to remove those biases). We then ranked the sub-factors in each of the market crashes and subsequent recoveries from most outperforming to most underperforming. Crashes were defined as market peak to trough and recoveries from trough back to original peak:

Event Name	Crash	Recovery
1987	-30% over 3 months	+34% over 16 months
Tech Bubble	-45% over 33 months	+80% over 48 months
Global Financial Crisis	-49% over 16 months	+100% over 36 months

Table 1: Performance of each crash and recovery analyzed in this paper. Source: Style Analytics.

Stock Market Crashes

The 1987 stock market crash was the shortest of the crashes, taking only 3 months to go from peak to trough, losing 30% in the US markets.

The GFC was the most widespread stock market crash, impacting all markets over a 16-month period with a 50% drop.

The Tech Bubble was the most concentrated of the stock market crashes, centered on overvalued technology and growth stocks in developed markets and especially in the US.

Each stock market crash had its own particular sub-factor signature but there are common trends:

• Low volatility stocks consistently did well across the crashes/regions. The most volatile stocks did poorly;
• Quality stocks work well with the exception of the 1987 crash in Emerging Markets;
• Avoiding Small Cap stocks, which is conventional wisdom during a downturn, worked well in the US in 1987 and in the GFC but not elsewhere;
• Generally avoiding both Growth and Value was the right move;
• The current COVID stock market crash has most closely resembled the GFC from a factor perspective.

Here are the sub-factors sorted from highest outperforming to worst underperforming, together with the average monthly return over the entire peak-to-trough period.

Figure 2: The sub-factor profiles of the four crashes in the US, grouped into color-coded investment styles. Average monthly returns are shown next to each sub-factor. Source: Style Analytics.

Figure 3: The sub-factor profiles of the four crashes in Europe. Average monthly returns are shown next to each sub-factor. Source: Style Analytics

Figure 4: The sub-factor profiles of the four crashes in Europe. Average monthly returns are shown next to each sub-factor. Source: Style Analytics

Visually, Figures 2, 3 and 4 show that the COVID Crash resembles the GFC and 1987 crashes more closely than it does the Tech Bubble, especially in the US and Europe. To better define this, we measured the cross-sectional correlations between the individual factor returns and the correlations between the average color-grouped styles. For the style calculations, we computed the equally weighted factor average of all sub-factors belonging to each style before computing the correlations between these ‘style returns’.

The results, shown in Table 1, show whether measured by investment style or by individual sub-factor, the COVID crash’s profile is closest to the GFC profile.

Region	COVID to GFC Correlation		COVID to 1987 Correlation		COVID to Tech Bubble Correlation
	By sub-factor	By style group	By sub-factor	By style group	By sub-factor	By style group
USA	0.71	0.75	0.24	0.51	-0.57	-0.49
Europe	0.85	0.94	-0.28	-0.31	0.17	0.22
Emerging Markets	0.44	0.47	-0.38	-0.36	-0.62	-0.65

Table 2: Correlations between factor and style returns for the three crashes and three regions examined in the text. The COVID crash most closely resembles the GFC. Source: Style Analytics.

Portfolio Recoveries

Unlike stock market crashes, which have different factor characteristics, portfolio recoveries tend resemble each other in all regions:

• Value and Dividend Yield are strong winners followed by Small Cap;
• The most volatile stocks are almost as good as Value stocks;
• Momentum is lacklustre tending towards negative;
• Growth and Low Volatility are clear losers.

These tables below show sub-factors sorted from highest outperforming to worst underperforming together with the average monthly return over the entire recovery (trough-to-prior-peak).

These growth figures are much smaller than in the corresponding crashes because recoveries last two to three times as long as crashes, so monthly average returns are smaller.

While crashes are all idiosyncratic with different factor signatures, recoveries all look rather similar, strongly favoring Value, Small Cap, and High Volatility Stocks.

Figure 5: Portfolio recoveries in the US favored Value, Small Cap and Dividend Yield. Source: Style Analytics.

Figure 6: Portfolio recoveries in Europe also favored Value, Small Cap and Dividend Yield. Source: Style Analytics.

Figure 7: Portfolio recoveries in Emerging Markets also favored Value, Small Cap and Dividend Yield. Source: Style Analytics.

What does this mean for the COVID stock market crash?

The COVID crash most resembles the GFC in that it is global, hitting all industries and has induced governments and central banks around the world to respond with strong stimulus packages. But it is unlike the GFC in that its cause was not systemic separation of financial risk and reward but is a global health crisis that radically cuts economic activity, employment, consumer spending and corporate profitability. Regardless, if history is any guide, the recovery from this crash is very likely to favor Value, Yield, Small Cap and High Volatility stocks around the world.

We assessed highest volatility stocks to see how well they did as a separate group. Shown in Table 1 are the average monthly returns of the three volatility sub-factors we use for each recovery in the US market:

Sub-factor	1987 recovery	Tech bubble recovery	GFC recovery
Volatility – 3 year	23 bps	50 bps	60 bps
Daily Volatility 1 year	18 bps	60 bps	70 bps
Market Beta	29 bps	40 bps	70 bps

The strong average monthly returns for these volatility measures, especially in the GFC, place High Volatility stocks at the top of the sub-factor list in expected outperformance in the eventual COVID recovery. High volatility delivered, on average in the GFC recovery, 60 to 70 basis points excess return per month.

For the remainder of this paper, we will exclude the Tech Bubble and focus on the crises more similar to today’s downturn: 1987 crash and especially the GFC.

 

Portfolio Recovery from a Stock Market Crash 

The GFC and 1987 crashes, in the US, show a strong reversal tendency: those factors which do worst in the crash do best in the subsequent recovery and vice-versa. Remember the recoveries took longer than the crashes, so average monthly returns are lower.

Figure 8: US GFC factor performances in the crash and subsequent recovery, showing a strong mirror-image effect. Source: Style Analytics.

Figure 9: The 1987 Crash/Recovery again shows a general mirror-image of factor performance. Source: Style Analytics.

What have factors done in the recent rally?

Factor performances since the 23 March bottom don’t resemble their behavior in any historical recovery we examined. The only factor to beat the overall market in the current rally is momentum, indicating that investors are doing little more than chasing returns (mostly in tech stocks). Quality and Growth factors have done well but two of the normally most outperforming factors – Value and Small Cap – are the two worst underperformers during this rally.

While this may partially be explained by the fast cash stimulus propping up previous winners, it raises questions about whether the recovery has begun or whether this rally is part of an overall larger market decline yet to materialize.

Conclusions

The GFC and the 1987 crash are both useful guides when positioning for portfolio recovery from a stock market crash like the one induced by the COVID pandemic.

If the downturn resumes:

• Portfolios that avoid high volatility and small cap stocks will likely outperform;
• Investors who shy away from both Value and Growth strategies should do well;
• Quality, low volatility and large cap stocks work best for defensive posturing.

During a full-fledged recovery:

• High Volatility / Small Cap and especially Value strategies should outperform;
• Growth and low volatility both lag the market;
• Momentum is a mixed bag.

Another possibility is that this recovery will be unlike any previous recovery in that it will come in two phases, the first of which we’ve already entered:

• Phase 1: fueled by government stimulus and ‘easy cash’, designed to prevent a market crash, which rewards companies that have already done well in the past (momentum) and those with strong growth prospects along with technology stocks.
• Phase 2: a more typical recovery fueled by consumer spending and confidence, starting after we reduce the social distancing policies, that rewards Value, Small Cap, Yield and High Volatility stocks.

While we all look forward to a return to less volatility and rising markets, we hope everyone stays safe and healthy.