Horses for Courses

Some styles of investing do better in certain environments than others

By: Verdad Research

The horse racing track in Brighton and the track in Epsom are quite similar, and horses that win at Brighton are often therefore favored to win at Epsom, noted the English writer Alfred Watson in an 1890s book on the sport. “A common phrase on the turf is ‘horses for courses,’” he noted, and “there is a good deal in the expression.”

There’s a good deal in the expression to to be learned about the stock market.
Small-cap value stocks beat large-cap growth stocks in October. This was an unusual phenomenon given the recent growth rally: this was only the third month in two years when we’ve seen small-cap value outperform.

But while small-value outperformance has been a rarity in recent months, it’s been more common over longer periods. Below, we show the percentage of months small-cap value has beaten large-cap growth over 3-, 5-, 10- and 50-year horizons.

Figure 1: Percentage of Months Small Value Beat Large Growth

Source: Ken French Data Library

Over the past decade, but especially over the past 3–5 years, the course favored large growth.

Over the past 50 years, environments that favored large growth have been very distinctive. The below chart shows average annualized monthly returns divided into a 10x10 grid with size on the x-axis and value on the y-axis. The right horse to bet on for this course was very clearly the largest and the most expensive.

Figure 2: What Worked When Large Growth Beat Small Value 1970-2020

Source: Ken French Data Library

On an annualized basis, the most extreme large growth portfolio beat the most extreme small value portfolio by 44% per year when it was winning. Any portfolio skew toward smaller companies or cheaper companies was actively punished in these environments. This is, notably, why market-cap weighted indices do so well in these periods, because they tend to have the largest weight on the largest, most expensive stocks, and most active managers prefer not to own such large concentrations in such large and expensive stocks.

But what about in months when small-cap value beat large-cap growth? Below we show the same 10x10 grid but only for months from 1970-2020 when small value won.

Figure 3: What Worked When Small Value Beat Large Growth 1970-2020

Source: Ken French Data Library

Here we see the exact opposite. The most extreme small-cap value portfolio beat large-cap growth by 62% on an annualized basis when it was winning, with any move toward smaller size or cheaper companies being significantly rewarded by the market.

Small-cap value worked about 50% of the time, and when small-cap value worked, the margin of outperformance was significantly larger than the margin of underperformance when it didn’t work. More importantly, the more extreme the bet on small value, the better the performance over the full period, according to our research. Yet that is exactly the strategy that does the worst during large growth rallies.

We call this factor inversion. In both the Nifty 50 bubble of the early ’70s and the dot-com bubble of the early 2000s, small value had underperformed large growth in the majority of 3-, 5- and 10-year trailing months.

And these concentrated, consecutive strings of losses for small value resulted in relative returns to large growth that approached or exceeded the double digits on a five-year look back about five times since the 1950s. This is the blue line in the figure below.

But in each of these periods when large growth looked like it was at its best from its trailing returns, we feel these were precisely the times to bet on the other horse. This is the red shaded area in the figure below. These are the annualized relative returns over five years for the capitalization-weighted large growth index versus the equal-weighted small value index.

Figure 4: Trailing Relative Returns to Large Growth and Forward Relative Returns to Small Value

Source: Ken French Data Library. 10x10 matrix relative returns.

Had you bet on large growth in the early ’60s after it had built up a lead on small value over half a decade, you would have missed one of the best opportunities to bet on the opposite horse. Had you done it again in 1974 when large growth had really pulled ahead by double digits, you would have missed out again. Same for the early ’90s and, of course, March of 2000. Today, large growth is way ahead of small value.

But this was not just a curious migratory cycle of a “comeback” horse accompanied by no logic. Each time large growth had built up a sizable lead, large growth valuations increased. And the greater the valuation spikes that accompanied large growth rallies, the greater the future headwind.

Today’s growth valuations look extreme relative to history. This isn’t a quirk of the price-to-book metric, as shown in the chart below including cash-flow multiples.

Figure 5: Growth Stock Multiples

Source: Ken French Data Library. 10th decile multiples for extreme growth and extreme large growth.

Most invested dollars in the stock market are currently crowded into the extremely liquid, cap-weighted indexes, which are disproportionately concentrated in large growth stocks. It’s an easily executable and surprisingly uncontroversial bet these days. But as in decade after decade of past market history, we believe those investors that can switch horses would have to switch (before others do) to win, place or even show.