Investors should treat indices’ Cape ratios with caution, says Max King
The conventional tool for assessing the valuation of stockmarkets is the price/earnings (p/e) ratio, calculated by dividing the level of the index by its aggregate earnings. If these profits are in the past, you have a historic p/e ratio; if they represent analysts’ forecasts for the year ahead, it is a prospective p/e ratio. Earnings normally rise in a period of economic growth and fall in a recession, often sharply. Markets decline, but not as far as earnings, as investors discount a rebound in the future.
Similarly, investors should be sceptical about an economic upswing going on forever and therefore ascribe a lower p/e ratio to their holdings. At the top of a market, equities often appear to be reasonable value judging by their p/e, and they seem expensive at the bottom of a market cycle.
To enable investors to look through the cycle, Robert Shiller of Yale University devised the “cyclically adjusted p/e ratio” (Cape). This divides the index level by average earnings over the last decade, which is assumed to iron out the economic cycle.
The resulting ratio, which will tend to look high because of the upward trend of earnings over time, is then compared to the history of the ratio to judge whether the market is cheap or expensive.
The problem is that it has not turned out to be a very useful forecasting tool. The Cape ratio can remain depressed or elevated for long periods of time; the S&P 500 was only fleetingly reasonable value in 2008 before looking expensive again. “Mean reversion,” the return to the average, looks credible on the 100-year chart, but not on any realistic timescale.
There are three key reasons for this. Firstly, accounting conventions have become more stringent over time. Expenditures and write-offs once regarded as “extraordinary items” are now included in earnings. Secondly, economic cycles are not regular, so a decade may include two recessions or none at all.
Thirdly, Cape takes no account of yields on government bonds, the risk-free alternative to equities. When yields are structurally low, investors should be prepared to regard a higher p/e ratio as normal, but when they are high, as in the 1970s, the p/e ratio should be lower.
How oil prices affect the ratio
In a recent report for investment management group Gavekal, Charles Gave provides two other convincing explana
