"Kee Points" with Jim Kee, Ph.D.

Is it too late to get into stocks? Should I take some money off the table? Those are always common questions after a pretty big stock market run-up like the one we’ve been experiencing since the November elections (up about 11%). The response should be equally common, which is that there is no credible, peer-reviewed research that suggests anyone can answer those questions that really imply the ability to time the market. When the market looks expensive it can stay that way for a long time (years), and when it looks cheap it can do the same. That’s because valuation tends to be “period specific,” meaning that different periods involve different levels of interest rates, inflation rates, taxes rates, and regulatory regimes (all of which influence valuations levels). That makes valuing the market difficult, as it involves forecasting a raft of variables. But there are a few data-driven ways to get your arms around the issue of stock market valuation levels that don’t require a huge amount of mathematical modelling and forecasting.

For example, I have mentioned in prior Kee Points the insightful research published some years ago by the Financial Analysis Lab at the Georgia Institute of Technology titled, “Seeking Guidance for the Dow? Try GDP.” The piece follows the Dow Jones Industrial average and nominal GDP (in billions of dollars) since 1916, and demonstrates that, over time, the Dow tends to move with GDP. The Dow exceeded GDP by 200% on the eve of the 1929 stock market crash, but by 1932 the discrepancy had pretty much been erased. And in the early 1980s the Dow was over 70% below GDP, but that discrepancy too was erased over time. The Dow has traded above GDP (1920s, 1950s-1960s) and below GDP (1940s-1950s) for multi-year periods, but for the past 100 years those two episodes – 200% above and 70% below - have been the valuation extremes. Currently, the Dow at 20,921 is about 8% above GDP, not cheap but certainly not an extreme.

Another approach to market valuations and subsequent stock market performance is to look at the correlation between valuation levels today and future shareholder returns. Crandall, Pierce, & Company performs this analysis on a regular basis using the Standard & Poor’s 500 price-to-earnings (p/e) ratio as a measure of valuation. Going back 45 years and correlating the market’s p/e ratio (high is expensive, low is cheap) with subsequent total returns (dividends plus price appreciation) over the next month to 15 years, they find the following: There is little correlation between valuation levels and stock market returns over the next year or so. But there is a very strong correlation between valuation levels and returns over the next 11 years. In other words, the more expensive stocks are right now, the lower the annual returns to investors over the next 11 years (the highest correlation). But there is significant variation, depending upon the period under consideration. If we applied these historical relationships to the p/e ratio for stocks today (i.e. at the beginning of the current quarter), investors could expect an average annual return of about 7.75% over the next 11 years. That would be below average but still doubling your money about every 9 years. Variation is huge, with the lowest annualized returns (given today’s valuation levels) of about 2% per year, and the highest being about 16% per year.

These two data-driven approaches to market valuation levels and expected investor returns should cause you to question anyone who tells you that “now is the time to buy,” or “now is the time to sell.” Current valuation levels are well within historical norms, so such statements imply an ability to outguess the market’s current estimate of economic and earnings growth, interest and inflation rates, tax rates, and regulatory developments. I’ve been a professional economist for over 20 years and I’ve never met anybody who could do that!