The heady rise in this year’s stock market has been powered by the tech sector, which now accounts for nearly one-quarter of the entire S&P 500 Index by market capitalization. The four so-called FANG stocks (Facebook, Amazon, Netflix, Google/Alphabet) alone generated 2 percentage points of the S&P 500’s 14 percent gain this year through September. Indeed, large-capitalization growth stocks are on quite a roll. In the 10 years from Sept. 1, 2007, through Aug. 31, 2017, large-cap growth returned 9.9 percent annualized, compared to just 7.2 percent for small-cap value stocks, which historically have been the best-performing U.S. equity asset class. In other words, during the past decade, the “small-cap premium” has become a deficit of 2.7 percentage points. Is it time to throw in the towel on small caps?

Before I address that question, let’s review a bit of market history. From January 1927 through August 2017, small-cap value returned 14.7 percent annualized versus 9.6 percent for large-cap growth, a wide 5.1 percentage-point premium. It was in 1993 that Kenneth French published a landmark research paper with Eugene Fama (who was awarded a Nobel Prize in Economics in 2013) on the three-factor model, which asserted that investors could earn a higher return by being exposed to the extra risks of value stocks (e.g., those with a low price-to-book ratios), smaller-cap companies, and the market itself (as opposed to “safe” assets such as Treasury bills).

I should emphasize that by being different from a benchmark (which is the investment approach of factor-based investors), portfolios may outperform for periods of time — or underperform; investors must accept the downside possibility as well as the upside opportunity. And as we saw over the past decade, the downside can go on for a while. But although a 10-year investment horizon is reasonably long, any specific decade is bound to be idiosyncratic and not necessarily indicative of what to expect going forward (for example, the past 10 years happened to encompass the 2008–2009 global financial crisis and epic stock market crash). Still, the math continues to favor portfolios that tilt toward “small” or “value” factors.

With that in mind, we conducted several studies of premiums over time periods that are more useful than any one specific trailing 10-year period. The research makes clear that small-cap value outperformed large-cap growth in an overwhelming majority of the 10-year periods. During the past 30 years, the 10-year gap between small value and large growth oscillated between an annualized premium of 14 percentage points and a deficit of six points.

If we extend our analysis back to 1927 (when reliable data became available), we can see that the longer the investment period, the greater the likelihood of a small-value strategy outperforming a large-growth counterpart. So, for example, over one-year rolling periods, small-cap value and large-cap growth each won close to half the time, but over 10-year periods, value dominated with almost a 90 percent win ratio.

I have no idea when the cycle will turn again toward small-cap value, but I do note a recent pattern of investor herding, which often is a prelude to a sharp correction, in the large-cap growth space. In fact, our research shows that investors’ collective appetite for higher valuations has grown over time, and that the gap between the lowest- and highest-valuation securities in the U.S. market, using price/book ratios as a touchstone, has never been larger than it is right now.

In closing, I should note that, while the argument for small-cap value is historically compelling, I do not recommend that investors abandon growth stocks and load up on small-value shares. Particularly due to the possibility of pronounced factor cycles and the behavior biases of investors, I believe that some combination of multiple equity asset classes is a better solution for most investors, not least because it will likely smooth out short-term portfolio volatility.

Gregg Fisher is founder and head of quantitative research and portfolio strategy at Gerstein Fisher.