Are measures of stock market valuation useful?

Many analysts think that the United States stock market is now overvalued.

A Bloomberg poll showed last week that forty-seven percent of financial professionals surveyed said that the equity market is close to unsustainable levels while fourteen percent already saw a bubble.

Federal Reserve Chair Janet Yellen told the Senate Banking Committee last week that equity valuations “remain generally in line with historical norms” even as the Fed's Monetary Policy Report accompanying her testimony noted that valuations for smaller companies in social-media and biotech industries appear “substantially stretched”.

However, on Friday, Cullen Roche at Pragmatic Capitalism wrote a post questioning the value of valuation measures for the stock market.

In the case of the stock market we’ve now seen a 20+ year period where stocks are “overvalued” by several metrics (Shiller CAPE, Tobin’s Q, Market cap to GDP, etc). So I think it’s worth asking yourself how useful all of these metrics really are. Can you afford to go through a 20 year period relying on a rear view mirror dataset assuming that the market is overvalued when the other participants might not be using the same gauge of “beauty” as you are?

John Hussman thinks that valuation measures can be useful though. In a post earlier last week, he showed a chart of various valuation measures for the stock market over the past few decades and correlated them with subsequent 10-year returns for the S&P 500.

He concluded that valuation measures have “provided clear guidance about expected market returns across a century of market history” and have “not failed at all even in recent decades”.

In any case, by most of the valuation measures commonly used, Hussman's chart showed that the stock market has not been obviously overvalued throughout the whole of the past 20 years. It was only so around 2000, when valuations reached a point which correlated with negative return for the next ten years.

Indeed, in late 2008 and early 2009, the stock market was actually undervalued, as Hussman noted in real time back in October 2008. From the latter post:

Stocks are now at the same valuations that existed at the 1990 bear market low. Relative to 30-year Treasury yields, the S&P 500 is priced to deliver the highest excess return since the early 1980's.

The current bull market in stocks took off from the base of undervaluation back in late 2008 and early 2009.

In any case, “overvaluation” is actually a subjective term, as Hussman explained last week.

A widespread misunderstanding comes in when people start using the phrase “fair value.” For any given set of expected future cash flows, if you tell me the price, I can tell you the long-term return, period. If you tell me the long-term return, I can tell you the price, period. Nothing changes this. If you want to say that lower interest rates “justify” a low expected return, and therefore justify a higher price, that’s fine. Just understand that the low expected return will still follow that higher price. If you want to say that in a zero interest rate world, stocks should be priced for zero expected returns over the next 8 years, I have no problem with the conclusion that under that assumption, stocks are at “fair value” here. Just understand that under that conception of “fair value” stocks can still be expected to return nothing over the next 8 years. What is emphatically not true, and not mathematically consistent, is to say that low interest rates “justify” a low expected return, and therefore justify a higher price, but then to turn around and say that since stocks are “fairly valued” under that assumption, they can be expected to achieve normal returns in the future.

It is useful to remember that when Hussman talked about expected return, he was referring to long-term return. In the short term though, valuation measures alone cannot determine when an “overvalued” market corrects.

Indeed, Hussman had written back in December 2006 that bull markets can end in a “speculative blowoff”. That turned out to be prescient. While Hussman noted that the stock market by then had already become overvalued, that cycle's bull market went on for almost another year before finally crumbling.

So while long-term returns correlate quite well with valuation measures, if the short term is the concern, Roche is probably correct in doubting whether valuation measures alone can be relied on to determine one's investment stance.