Larry Swedroe has recently written a great summary of a paper investigating the sources of the 12-year underperformance of value vs. growth for The Evidence-Based Investor. The paper was written by Rob Arnott and Vitaly Kalesnik of Research Affiliates, Campbell Harvey of Duke University and Juhani Linnainmaa of Dartmouth. It is available here. To get a feel for how bad the underperformance of value vs. growth was, it is worthwhile to take a look at the performance of the 20% cheapest stocks in the US vs. the 20% most expensive stocks in the US-based on the data available on Ken French’s website. The chart below shows that the underperformance of value vs. growth was worse during the Great Depression but not as long. In fact, using this metric, value stocks have underperformed growth stocks now for 13 years. Before that, the longest period of underperformance was the 12-year period from January 1951 to January 1963.
The underperformance of value vs. growth
Source: K. French website.
In the paper, the authors argue that what drove the outperformance of growth vs. value over the last 12 years was mostly the shift in relative valuations with price/book-ratios for growth stocks growing faster than price/book-ratios for value stocks. Hence, growth stocks are now overvalued relative to value and we can expect growth to underperform value going forward.
I agree with all of these conclusions, but I thought I’d bait Larry and others into a discussion. You see, one of the things that I find unsatisfactory is to explain performance differences with changes in valuation. Valuations are about the most important driver of future returns but if you want to explain past returns, it is insufficient to explain them by changes in valuations because valuation is not a primary variable. As I have explained in an article for the Enterprising Investor, asset class returns are driven by the fundamental drivers of the economy and markets. Hence, if you want to explain why value underperformed growth, you need to explain why relative valuation changed between value and growth stocks. And you cannot say that this is because the performance of value was worse than the performance of growth…
In my view, valuations are a reflection of expected inflation, real rates, economic growth and investor sentiment. If inflation and real rates decline, the discount rate drops and as a result, valuations become more expensive. Growth companies typically have a higher duration because they have lower dividends than value stocks and more of their present value is derived from earnings far in the future. Hence, if interest rates decline, the valuation difference between growth and value stocks should increase as Drew Dickson has pointed out on Twitter to me. Given the low inflation and real rates since the financial crisis, it should be no surprise that growth stock valuations have outpaced value stock valuations.
But if this is the main driving force behind the valuation difference, then going back to these fundamental drivers also gives us an idea why and under what circumstances valuation differences should revert to the mean. Mean reversion would happen if interest rates normalise and inflation and/or real rates rise. While I think that this is entirely possible, it is by no means certain as I have argued in this blog post.
Another reason why the relative valuation between growth and value stocks have shifted might be economic growth and as a result earnings growth. Given the anemic economic growth in many developed markets over the last decade, one would expect value stocks to have done better than growth stocks, though, so this seems unlikely to be an explanation. Furthermore, Rob Arnott and his co-authors argue with price-book-ratios which have a weaker link to earnings growth than the price/earnings-ratio, for example.
This leaves us with the third possibility that growth stocks have higher valuations than value stocks because of investor sentiment. In other words, growth stocks are in a bubble. While that is possible, this greed-driven valuation difference seems much smaller today than it was in the late 1990s indicating that growth stocks may have a long time of outperformance ahead of them because the current valuation of growth stocks can easily rise to mich higher levels.
Finally, I would like to admit that about eight years ago, I analysed the cyclically adjusted PE-ratio (CAPE, aka Shiller PE) and its dependence on the fundamental drivers of the economy. The results can be found here and here. Back then, I argued that the CAPE in the United States is way too high to be explained by inflation, real rates, etc. at the time and thus, US stocks should underperform the regions with better value like Europe. This view was informed by my conviction back then that interest rates and inflation had to normalise within a couple of years. How wrong I was. And how wrong investors could be this time around if they argue for mean reversion in the relative valuation of growth and value.
I hope mean reversion kicks in soon because I am a value investor at heart and can’t wait to see growth stocks falter, but if you track valuations back to their fundamental drivers then you can see that mean reversion in valuation requires a big shift in interest rates or a massive recession. Otherwise, we will not see a reversion to the mean or only a mild reversion to a mean that is higher than the current historic mean. By just arguing that valuations will eventually revert to the mean without understanding the drivers of valuations you run the risk of making the same mistake I made many years ago in advocating for the underperformance of US stocks vs. Europe.