Trends in factor investing in the 20th and 21st centuries

Key takeaways

• “The trend is your friend,” is a common investment expression, that makes sense, right up until it reverses.
• In this Angle, we’ll explore how value trends and cycles of value, momentum, quality, low size, and low volatility have changed over the 20^th and 21^st
• We’ll investigate the reason for the shrinking value premium in the 21^st century (January 2001 to today) vs. the higher premium experienced during the 20^th century (1965 to 2000).
• Investigating the trends and cycles in factors can provide deeper insight into the long-term trends for factors like value going forward.

In finance, nothing goes straight up. 

Take the last 10-15 years of value. The average annual return of value in the 21st century (January 2001 to today), has been a modest 1.5%, compared to a healthy 7.1% per year in the 20th century (1965 to 2000).¹ I examine the long-run returns and shorter-term trends of value, and other factors, in detail in a recently published paper in the Journal of Portfolio Management. 

Value’s drawdown and recovery

Figure 1 illustrates the start of the current value drawdown from 2017 through December 2022 (in orange) contrasted with the other worst five value drawdowns from the 20th century using the HML value factor constructed by Kenneth French.²

^{Figure 1: Value drawdowns}

We’re still living through the longest and deepest value drawdown in history.

There are three parts to our most recent, and worst-ever, drawdown. First, we were in a late up market cycle until 2019. Value firms typically have a lot of physical capital, and because it’s hard to change that capital to other purposes when things start slowing down, value firms tend to underperform when the economy starts slowing.³ Conversely, we show in Factor Timing with Cross-Sectional and Time-Series Predictors, that value tends to outperform coming out of recessions. Large holdings of physical capital of value firms gives them operating leverage, so it’s relatively easy to increase manufacturing on a physical production line when demand ramps up.

Second, if we stopped at the end of 2019, the value drawdown would have been bad. But there would have been other drawdowns, like the dot-com bull market of the late 1990s that were worse. What made the drawdown notable was COVID in 2020. As the world turned virtual, companies with physical capital that relied on physical interactions were challenged.

The third part is the recovery. The trough, November 2020, corresponds to losing approximately 55% from peak. That’s the same month as when Pfizer announced positive results from its first vaccine trial against the coronavirus.⁴ With some hope that things would return to normal, value started to outperform. Value beat growth decisively in 2021 and had a ripper of a year in 2022. The path wasn’t straight up, as variants like delta and omicron came and went, but since then value has been ascendant.

Note that we’re now six years into the value drawdown and still 30% from the previous peak—my money is on that value has a long way to run.

Value trends and cycles

It’s useful to decompose the returns of value into trends and cycles. Short run cycles run from around 6 months to 1.5 years, while long run trends span more than a decade. How have value trends and cycles changed over the 20th and 21st centuries?

Figure 2: Trends and cycles of Value

Value trends strong upward until the late 1980s, flattens out but is still positive over the 1990s, and then in the 21st century exhibits an inverse-U shape. What I find interesting is that the cycles around that trend (shown in the red dashed lines) are approximately the same in the 20th and 21st centuries (which can be proved by some fancy statistical techniques like spectral analysis and durations of regime-switching models). Cycles for all factors, in fact, are similar across centuries.

You’ll observe that the late 1990s (the second worst drawdown for value) is associated with a pronounced downturn of short-term cyclical components, but the long-run trend is flat and if anything, increasing as the 20th century approaches. The loss of value in the most recent drawdown is particularly nasty because it is associated with both negative trends AND cycles. Furthermore, value trends have changed, but the length of the cycles have remained fairly constant. 

Trends and cycles of other factors

While the trend for value has changed direction so far in the 21st century (I am confident it will turn), some factor trends have increased. Figure 3 shows the trend for the quality factor, which becomes stronger in the 21st century.

Figure 3: Trend of quality

The trend for the size factor has also increased in the 21st century (not shown in this Angle, but you can see in the full paper).

I’ll close with minimum volatility. The ability of minimum volatility to reduce equity market risk is even larger in the 21st century than the 20th, partly because there are some very large movements for the most volatile stocks over the last 20 years. These large reductions in risk have translated to higher Sharpe ratios (a measure of risk-adjusted return) for low volatility stocks in the 21st century.⁵ Minimum volatility tends to generate active returns best when the market is going down—like last year, when our minimum volatility USMV ETF outperformed the S&P 500 Index by 876 basis points.⁶

Performance data represents past performance and does not guarantee future results. Investment return and principal value will fluctuate with market conditions and may be lower or higher when you sell your shares. Current performance may differ from the performance shown. For most recent month-end performance and standardized performance, click here

Future short-term and long-term trends for factors

Factors do vary over time—but even though the trends and cycles vary, I believe the premiums for holding factors will persist over the long run because they arise from economic rationales of a reward for bearing risk, structural impediments, or investors’ behavioral biases. The fact that factors’ trends and cycles are different means there is diversification in holding a well-balanced portfolio of multiple factors and certain investors might consider tilting into different factors at different points in the business cycle. Factors are here to stay.

Performance data represents past performance and does not guarantee future results. Investment return and principal value will fluctuate with market conditions and may be lower or higher when you sell your shares. Current performance may differ from the performance shown. For most recent month-end performance and standardized performance, click here