Time periods matter. Sure, for 1972-2019 so far, mid-cap beats the U.S. stock market by compounding at about a 1.6% higher rate. But consider this:
From January 1972 to December 1981,
mid-cap compounded at 9.25 and U.S. stocks at 7.24.
From January 1982 to December 1997,
mid-cap compounded at 16.88 and U.S. stocks at 16.36. Note that mid-cap was also more volatile.
From January 1998 to December 2004,
mid-cap compounded at 10.82 and U.S. stocks at 5.13.
From January 2005 to December 2019,
mid-cap compounded at 9.12 and U.S. stocks at 8.71. Mid-cap stocks were also notably more volatile.
I would submit that the entire "premium" exhibited by mid-cap (and small-cap) under this kind of backtesting happened in two historical periods: 1972 to 1981 and 1998 to 2004. Although mid-cap outperformed during the other periods, the outperformance is not significant in the context of the higher standard deviation (no risk-adjusted outperformance). What happened in 1972 to 1981 and 1998 to 2004? Here is a
historical P/E chart of the S&P 500:
The S&P 500 started the 1970s at a healthy P/E ratio in the 17-20 range. It then crashed after the 1973 Arab oil embargo, reaching as low as 7ish by 1980. This covered the exact period of mid-cap outperformance. I don't know the exact story how this happened--perhaps at the time publicly traded oil companies were large (Exxon, Chevron, etc.) and there were fewer independents publicly traded as mid-caps. But it was a definite historical event, and there's an explanation why the pain was felt by big companies but not mid-caps or small-caps.
Another strange event happened in the late 1990s. Many of you lived through it--the dot-com bubble. People bid up stocks to enormous P/E ratios. The S&P 500 crossed a P/E of 30 two years before a recession even started in 2001. I don't know why people bid up large-caps that then went down, but there is an explanation why the "irrational exuberance" was felt for large companies that didn't ultimately grow that fast without the same problem for small-caps or mid-caps as a whole.
So, in my mind, betting on mid-cap or small-cap stocks based on this data is betting that there will be another significant historical event like the dot-com bubble or the Arab oil embargo with the pain concentrated on large-cap stocks. On the one hand, it seems like this bet wouldn't hurt since over long periods of time mid-cap and small-cap stocks generally do at least as well as large-cap stocks, so the stakes of being wrong aren't that bad. On the other hand, I don't see any reason for pain to be concentrated on large-cap stocks in the future unless there's a very specific industry concentration problem or something (like IT or media conglomerates being large-caps).
