The media has been sharpening its focus on cycles lately, but not in anticipation of the Tour de France. Rather, the financial media’s spotlight is on business cycles.
A common refrain from the industry is that the performance of certain groups of stocks is tied to the business cycle. For example, it has been argued that some factors or investment strategies are “defensive” because they perform well late in the cycle when economic growth begins to slow, while other premiums should be avoided because historically they have underperformed during this phase. Investors who are pessimistic about the direction of the economy may be tempted to wonder if it’s time to reevaluate how their portfolios are positioned.
It is important to be cautious when interpreting research conducted on a small number of observations. The National Bureau of Economic Research (NBER) has identified 15 recessions in the US since the start of stock market data coverage in 1926. Small sample sizes can be especially susceptible to the effect of a handful of outliers.
We assessed the behavior of stock market premiums—size, value, and profitability1—and found no reliable evidence that their performance fares differently deep into an economic expansion. Our results suggest that pinpointing where we are in the business cycle is unlikely to yield useful inputs for one’s asset allocation.
The NBER retrospectively classifies the US economy through time by identifying peaks and troughs that define recessions. The last trough was determined to be June 2009, marking the end of a recession beginning December 2007.
While the NBER effectively provides a binary historical classification of the economy (recession or not), many in the industry describe fluctuations in the economy as business cycles with four phases: expansion, peak, contraction, and trough. The nearly 10 years that have elapsed since the last NBER trough have prompted many to suggest we are nearing the end of the current expansion prior to the next recession, a stretch some refer to as “late-cycle.”
For the purposes of our study, we define late-cycle using the chronological midpoint between NBER trough and peak dates, as shown in the illustrative example of the business cycle in Exhibit 1. Late-cycle months therefore compose the second half of each expansion period.
Late-cycle periods defined for each NBER expansion using the halfway point between each trough and peak.
While the average premiums are similar between the late-cycle months and all months, an investor may be wondering how the range of outcomes compares. Specifically, are the premiums more prone to especially bad outcomes late in the business cycle? The distributions illustrated in Exhibit 3 suggest not. The distributions of premiums in late-cycle months are quite similar to the overall distribution in all months. In other words, there is no reason to expect different size, value, or profitability premiums just because we are many years into the current economic expansion.
Includes monthly observations through December 2018 and beginning from June 1927 for small caps vs. large caps, July 1926 for value vs. growth, and July 1963 for high vs. low profitability. Months classified as late-cycle marked in teal.
Small Minus Large
Value Minus Growth
High Prof Minus Low Prof
What we know is that the US expansion is in its 10th year. What we don’t know is how much longer this cycle will last. With historical US expansions lasting an average of about four years, many in the industry would agree we are late in the cycle. What does that imply about the distribution of size, value, and profitability premiums? Not much, according to the historical evidence. While the premiums are not assured over any time frame, research suggests the most reliable way to capture them is through a disciplined approach that maintains consistency through time. For investors still craving some news about cycles, better to focus on ones with handlebars.
Wes Crill is a Vice President in Dimensional’s Research group. He holds a PhD from North Carolina State University.
Marlena Lee is Co-Head of Research at Dimensional. She holds a PhD from the University of Chicago Booth School of Business.