Debates abound on the relative importance of the economic cycle to investment success. Peter Lynch famously said, "If you spend more than 13 minutes analyzing economic and market forecasts, you've wasted 10 minutes.” On the flip side, macro investment houses have constructed intricate frameworks to understand the “economic machine.” The challenge with economic data is that it is notoriously prone to revisions, significant lags, and adjustments in how it is measured through time—all of which are not suitable for timely and reliable investment signals.

As factor investors, we believe that certain fundamental characteristics, as distinct from economic variables, drive stock returns. Our job as researchers is to distill the hundreds of investment factors into cohesive themes that can serve as foundational building blocks for equity strategies. Among the litmus tests for those themes is persistence. The purpose of this piece is to identify investment themes that deliver persistent outperformance in multiple different economic environments. By doing so, the burden of market timing—an often subjective and error-prone exercise—is lessened.

Key Findings:

• The persistent outperformance of certain investment factors across the economic cycle argues against passive equity allocations.
• Investment factors are not rewarded linearly across the economic cycle.
• Factor spreads tend to expand in and around recessions, suggesting that greater potential exists for disciplined active managers to deliver outperformance in those periods.

Parsing the cycle

The National Bureau of Economic Research (NBER) has designated seven recessionary periods from 1964 through 2015. The average length of recession was a relatively short 12 months. The average length in between recessions—“expansions”—was a much longer 73 months. In fact, of the 624 months in the study from 1964-2015, expansions represented the vast majority of observations at 541 (87%), versus recessions at 83 (13%).

Because recessions are notoriously challenging to predict, and only truly viewable in the rearview mirror, we further designate four phases of the economic cycle in our analysis.

Recession: NBER defines recession as “a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales.”  [1] For the purposes of this study, we use the official NBER recession beginning and ending dates.

Pre-recession: The 12 months prior to a NBER designated recession. This period would be analogous to the late stage of an economic expansion, just before economic growth trails off and turns negative.

Post-recession: The 12 months after a NBER designated recession. These periods represent the early stages of recovery when economic growth bottoms out and begins to turn positive.

Pure expansion: Anything that is not included in the pre-recession, recession, or post-recession periods, we call pure expansion. This represents periods in which economic growth is robust.

The goal is to identify which factor themes are most effective in the different regimes, and draw inference for portfolio allocations. Before diving into the four regimes, I discuss overall factor performance from 1964-2015 as a baseline for the analysis.

Level-setting: overall factor performance

The pantheon of potential investment factors can be distilled into six distinct themes that deliver strong absolute, risk-adjusted, and consistency of return over time. Themes can be used for both positive and negative screening within strategies. While important to focus in on themes for selection, it is equally important to actively eliminate stocks through a quality overlay.  [2] The themes used primarily for selection are: Shareholder Yield, Dividend Yield, Value, and Momentum. The themes used to avoid stocks are Financial Strength, Earnings Quality, and Earnings Growth.  [3]

To level-set the analysis, I rank each stock within a Large Stocks [4] universe based on the selection themes mentioned above. Hypothetical decile portfolios [5]  are formed and excess returns are observed across the full 1964-2015 period in each regime. Below are the excess returns for the High and Low deciles on each theme for the entire 52 year evaluation period. The high deciles are considered to be those most favorably positioned from a characteristic perspective. The high Value decile, for example, represents the cheapest 10% of stocks in the universe. Conversely, the low Value decile includes stocks considered to be in the most expensive 10%.

The Spread of excess return between the high and low deciles is also included. The high-low spread is a great robustness check and will become important as we compare factor performance in economic regimes. Think of the spread between high and low as a proxy for the value the market places on differentiation by that factor. Using Shareholder Yield as an example: the market rewards an investor’s emphasis on stocks that reward shareholders via dividends and buybacks (high decile) by delivering an 8.9% higher annualized return than for stocks that do not pay dividends and dilute shareholders (low decile). Spread serves as a quantitative proxy for an active managers’ ability to generate excess return by deviating from the market.

Value, Momentum, and Shareholder Yield are dominant themes, with strong excess return of the High decile and a wide spread (10.8% in the case of Value) between the High and Low deciles. Each provide a different “take” on an individual stock that is useful in stock selection.

Valuation tends to be a surrogate for market expectation for a stock. Cheap stocks are often characterized by low investor expectations, and therefore, discounted valuations. As we dive into the research, we will find that differentiation based on valuation is most attractive during times of maximum pessimism. This makes perfect sense, when overall market sentiment is poor, investors tend to only see red. Quantitatively, this is the time for active managers to shine.

Momentum exploits a number of behavioral biases—underreaction, overreaction [6] , and disposition effects [7] —in order to identify stocks having consistent and upwardly trending price movement. As we will see, Momentum, which is a trend following signal, tends not to work well at inflection points in market cycles. However, it seems to work exceptionally well in high return and trending markets.

Shareholder Yield represents a total return of capital to shareholders. It consists of a stock’s dividend yield and net buyback activity over the previous 12 months. Stocks with high Shareholder Yield tend to be conservative capital allocators and good shepherds of shareholder assets. Similar to Value, the factor theme does well during challenging times (in and around recessions), but also produces strong risk-adjusted return throughout the different regimes.

Dividend Yield is an interesting factor as it tends to move in and out of favor over time to a greater extent that the other selection themes. I include dividend yield because income generation holds a special place in the hearts of millions of baby boomer retiree investors. In implementation, dividend yield is actually most effective when paired with Value and Quality to smooth the gyrations over time.

The Quality themes of Financial Strength, Earnings Quality, and Earnings Growth tend to offer lower excess return in the High deciles, but interact well with the other selection themes when the Low deciles are avoided.

Pre-recession: own expensive stocks at your peril

Since 1979, the NBER has tracked the formal announcement date of their designation of a turning point in the economic cycle. Announcements typically occur 6-12 months after a recession begins, but can be longer. Even though official determinations are lagging, the market seems to have a sense for recession well in advance. In the 12 months preceding recessions, the Large Stocks universe was effectively flat, having delivered an annualized return of -0.5%. In economic cycle parlance, these periods would be analogous to late stage expansion. Given that we are 85 months (as on June) into the current expansion, one could readily argue we are closest to this type of pre-recession environment—though not necessarily in it quite yet.

Striking in the pre-recession periods is the strong high-low spread for Value of 17.7%, which suggests that investors shun expensive stocks when economic growth is in question. In fact the low Value (expensive) decile excess return of -12.4% (meaning the market returned on average -0.5% and expensive stocks did -12.4% worse) is the worst in our study across regimes. The trend of expensive stocks underperforming starts well in advance of an official recession and extends through the early recovery post-recession regime. Not only do these stocks deliver poor excess and absolute return, they tend to exhibit extreme volatility. While high Value (cheap) stocks had a standard deviation of just 14.8% annualized, low Value (expensive) stocks had a standard deviation of 30.3%. As I have written about here, volatility is highly correlated with valuation. Expensive stocks are volatile, while cheap stocks tend to be less so.

Momentum and Shareholder Yield also exhibit strong high-low spreads and deliver solid excess return. High Momentum stocks do well in these environments as the last vestiges of the bull market runs its course. Similarly, Low Earnings Growth—unprofitable stocks with poor earnings trends—are severely punished by the market having excess of -11.5%.

Investors also seem to prefer stocks with strong balance sheets during the pre-recession regime. The high-low spread for Financial Strength more than doubles the 1964-2015 average (from 5.3% to 11.5%). The theme of financial strength encompasses a firm's reliance on outside sources of capital—debt and equity. In this environment, preference is placed on firms that organically finance their balance sheets through operational cash flow.

Though High Dividend Yield is not one of the stronger factors preceding a recession, the avoidance of Low Dividend Yield seems to be extremely important (-10.0% excess).

Recession: seek companies returning capital to shareholders

Across the 83 months within the seven recessions, the Large Stocks universe declined an annualized -4.7% as the market rightly anticipated economic uncertainty. Yield was the clear winner from an excess return perspective in this regime. Stocks in the high decile of Shareholder Yield produced 7.0% annualized excess return, while high Dividend Yield generated 4.9% of excess. The high-low spread widens significantly for both themes as compared to their 1964-2015 average.

As believers in the virtue of strong shareholder-orientation, it makes sense that the market would reward stocks engaging in buyback and dividend activity, as opposed to share issuance and no dividends. Buybacks and dividends necessitate cash, and the ability to distribute cash in times of economic stress is a clear signal of firm strength. Investors also appear to be more than happy to accept dividend payments to cushion the blow of stock price depreciation. Further supporting this hypothesis is the continued performance of Financial Strength in recessions as a carry-over from the pre-recession regime. This dovetails with earlier comments on the preference for organically funded balance sheets. Excess return of high Financial Strength remains elevated (3.4%) versus the full period trend (1.1%) and the high-how spread remains wide at 7.2%.

In addition to strong balance sheets and shareholder-orientation in recessions, investors continue to shun expensive stocks. The low Value (expensive) excess of -11.0% is one of the worst of all the environments evaluated and the high-low spread of 15.6% is the second only to the Value high-low spread in pre-recessions.

Momentum, normally a very robust factor loses some of its luster in recessions. Given though that recessionary periods usually signal a change in leadership, it makes sense that Momentum would struggle around an inflection point. These “Momentum inversions” exist extensively in post-bear market rallies, which periodically coincide with economic recession.  [9] To the extent that recessions correlate with bear markets, this may explain the decrease in excess return to high Momentum in the table below. It further, argues for the pairing of Momentum based strategies with other factors like Value or Quality to increase the efficacy of the factor.

Post-recession: shareholder-orientation continues to deliver

Post-recession periods tend to encapsulate the early stage recovery from economic troughs. These are one of the stronger periods for absolute return. The Large Stocks universe delivered 14.8% annualized returns. Yield and Value again dominate the other themes in these environments. The high-low Value spread remains high at 14.7%, but is down from its pre-recession and recession highs of 17.7% and 15.6%, respectively. High Financial Strength falls to the back of the pack, delivering marginally negative excess return, as investors place little emphasis on high balance sheet quality and appropriate leverage in the early stages of a recovery.

The Momentum theme again begins to assert itself with the spread widening from 6.6% in recession to 10.3% post-recession. It is during these periods that the precursors for steadily trending markets can be established.

Pure Expansion: momentum leads the pack

Pure expansions, which lie in between post-recession recoveries and pre-recession late stage expansions, is the most common at about 60% of monthly observations from 1964-2015. Notable in this period is that the high-low spreads compress to half that of the other regimes while the overall market return is highest at 15.8%. Momentum, a trend-following factor, rises to the top of the pack based on excess return and high-low spread.

As economic growth takes hold and consumer confidence rises, the market places less emphasis on differentiation, driving spread compression.  [10] Intuitively, this makes sense. In periods when market returns are on average negative or the future is uncertain, the ability to generate positive return is predicated on an investor’s skill of choosing wisely. For example, the average market return in recession is -4.7%. High Shareholder Yield delivers excess of 7.0%, which translates to a positive absolute return of 2.3%. In a raging bull market, when risk concerns slowly fade away (Pure Expansion), differentiation is less important as investors can generate double digit positive return with simple passive beta exposure to the market.

This helps explain why active managers tend to struggle in raging bull markets similar to the one we have experienced since 2009—differentiation is simply not rewarded to the same extent. This helps explain why active managers have struggled so much over the last several years.

Implications for investors

Though unpredictable, the economic cycle has significant implications for investors. Given the vast amount of evidence that certain themes consistently outperform, and underperform, throughout the cycle, an allocation to passive market cap-weighted indexes seems almost naive.

The information presented here supports an assertion that active management is, at the least, more likely to outperform during turbulent times when factor spreads are wide. Remember though that investing in public equities is largely a zero-sum game, less transaction costs. For every active manager that takes advantage of wide factor spreads to their own benefit, another is investing in the wrong end of the spectrum to their clients' detriment. Whether any associated outperformance is based on luck or skill is up for debate. In either case, differentiation based on fundamental factor themes has been historically rewarded in a non-linear fashion over the economic cycle. Spread compression in certain environments (pure expansion) suggests that generating excess return may be more difficult, while spread expansion in other environments (pre-recession, recession, and post-recession) suggests greater opportunity for excess return.

Unfortunately, it just so happens that investors tend to emotionally drawdown equity allocations in the periods when high-low spreads are highest, and pile into equities when the spreads are compressing. Rather than attempt to time allocations based on economic indicators that are often outdated, volatile, and revision-prone, it seems diversifying equity exposure to multiple key selection factors and staying invested throughout the economic cycle may be the most prudent course of action.

Realistically though, not every investor can allocate to the high decile of Value, Momentum, or Yield while avoiding low Earnings Quality, Financial Strength, and Earnings Growth. Large asset aggregators have recognized this problem in their attempts to build scalable highly-liquid products that are broadly accessible. In an era of continued fee compression, product sponsors sacrifice potential alpha for scale. The result is usually a neutered implementation of true factor-based investing that uses factor tilting instead of factor concentration. As such, most widely-available ETF’s and factor implementations will struggle to capture the excess returns discussed in this piece. My colleagues view this as the battle of Smart Beta versus Factor Alpha. [12]  Together, we are confident that in the end, concentrated lower-capacity factor exposures—Factor Alpha—will prevail.

[1] National Bureau of Economic Research. US Business Cycle Expansions and Contractions. http://www.nber.org/cycles.html

[2] “Stocks You Shouldn’t Own” (2016). http://osam.com/pdf/Commentary_StocksYouShouldntOwn_Jan-2016.pdf

[3] Shareholder Yield is the combination of a stock’s dividend yield and the percentage of shares repurchased over the trailing 12 months. Financial Strength evaluates the appropriateness of a stock’s overall leverage. Earnings Quality measures the conservatism of a stock’s earnings. Earnings Growth reviews profitability and the trend in earnings.

[4]  Data is sourced from Compustat via Thompson QA. Robust quarterly corporate financial statement data is available starting in 1964. Factor returns are generated from our proprietary Large Stocks universe, which includes all stocks listed on the NYSE, AMEX and NASDAQ with a market-cap greater than the inflation-adjusted database average and excluding Utilities. All excess returns are relative to the equal-weighted Large Stocks universe.

[5] Decile portfolios are formed on a rolling annual basis and refreshed monthly to avoid the impact of seasonality.

[6] Hong and Stein (1999).

[7] Disposition effect is a behavioral phenomenon whereby investors tend to hold on to losers and sell winners quickly.

[8] For Dividend Yield, High and Low were formed from the Large Stocks universe restricted to dividend-paying stocks only.

[9] “The Same Old Bear” (2010). http://www.osam.com/pdf/Commentary_Jan10.pdf. “Momentum Crashes” (2013). http://ssrn.com/abstract=2371227.

[10]  I find spread compression to be statistically significant at the 1% level for all factor themes across all regimes.

[11] Recession dates: NBER. Returns: Thompson QA, Compustat. Consumer Confidence: University of Michigan.

[12] “Alpha or Assets? Factor Alpha vs. Smart Beta” (April 2016). http://osam.com/Pdf/Commentary_AlphaOrAssets_FactorAlphaVersusSmartBeta_April-2016.pdf