This article examines the historical relationship between price-earnings ratios and subsequent stock market performance and discusses why history might not repeat itself this time. The article finds strong historical evidence that high priceearnings ratios have been followed by disappointing stock market performance in the short and long term. Specifically, high price-earnings ratios have been followed by slow long-run growth in stock prices. Moreover, when high price-earnings ratios have reduced the earnings yield on stocks relative to returns on other investments, short-run stock market performance has suffered as well. Despite this evidence, however, we cannot rule out the possibility that these historical relationships are of little relevance today due to fundamental changes in the economy.
This paper appears to have been written with a year or so of the 2000 crash.
Here I’ll jump to the conclusion for you.
IV. CONCLUSION Some analysts view the current high priceearnings ratio of the stock market as a sign that the stock market may be headed for a downturn. This view receives some support from historical evidence that very high price-earnings ratios have usually been followed by poor stock market performance. When price-earnings ratios have been high, stock prices have usually grown slowly in the following decade. Moreover, at times such as the present when high price-earnings ratios have reduced the earnings yield on stocks relative to interest rates, stock prices have also tended to grow slowly in the short run. Forecasts based on such evidence are subject to much uncertainty, however, because history may not repeat itself. Specifically, the possibility cannot be ruled out that this time will be different due to fundamental changes in the economy that will allow high price-earnings ratios to persist and thus stock prices to continue growing both in the near term and in the coming decade.
Which is so funny b/c of the book by the same name
which is apparently based on some bad data.
Anyhow, let’s look at the meaning of the PE Ratio and the Schiller PE Ratio (smoothed PE ratio over 10 years).
In June 2000, the P/E ratio was slightly above 29. While this value was lower than a year earlier, when the ratio was close to 36, it was still high by historical standards.
Currently at 26.57
A key question is this. Is the P/E ratio high because earning are temporarily down due to some historical anomaly? Or is the P/E ratio high because the prices have risen, and the underlying companies are not likely to change their earnings much in the near term?
For example this article looks at the PE ratio of the FTSE (British)
The author argues the currently high P/E ratio is due to exceptional circumstances in the resources and energy industries.
Mark Richards, a strategist at JP Morgan Asset Management, suggested that investors could largely ignore the current high p/e values.
“What some of these numbers are reflecting is that two of the biggest sectors in the UK market – energy and resources – have had very weak earnings and a lot of exceptional circumstances,” he said.
And goes on to argue that a smoothed P/E ratio does not show the same kind of inflation:
If valuation ratios that look beyond the past 12 months are looked at, a very different picture emerges.
A variant of the basic p/e number called the “cyclically adjusted price to earnings” or Cape ratio is calculated in the same way as the p/e but uses average earnings over a 10-year period – adjusted for inflation – to smooth out the ups and downs of a typical business cycle.
Note: the Cape ratio for the S&P 500 is more like 29 right now, so the same argument can’t be made there.
Another question — if teh PE ratio drops, does it drop b/c of higher earnings or lower stock prices? Higher earnings can’t be bad. What usually happened?
A decline in the P/E ratio back to its longterm average could occur in two ways—through slower growth in stock prices or faster growth in earnings. Of these two possibilities, only slower growth in stock prices would imply a negative outlook for the stock market.9 One way to determine which outcome is likely to prevail is to examine the historical record and see whether movements in the P/E ratio back to its longterm average have occurred mainly through slower growth in stock prices or mainly through faster growth in earnings.10 This is the approach taken by Campbell and Shiller in a widely cited study published in 1998.
For each year from 1880 to 1989, Campbell and Shiller calculated three measures: the P/E ratio of the S&P 500 index at the beginning of the year, the annualized changes in real stock prices over the following ten years, and the annualized changes in real earnings over the following ten years. The measure of earnings used in the P/E ratio was the average of realized earnings over the previous ten years.11 Stock prices were measured in real terms because what matters to investors is the purchasing power of their investment.12 If movements in the P/E ratio back toward the average occurred through changes in stock price growth, years with high P/E ratios should be years with low subsequent growth in stock prices. On the other hand, if movements in the P/E ratio back toward the average occurred through changes in earnings growth, years with high P/E ratios should be years with high subsequent growth in earnings. Campbell and Shiller use both simple graphs and statistical analysis to determine which outcome tended to prevail over the sample period.
Campbell and Shiller found that higher P/E ratios are usually followed by lower stock price growth during the following decade.13
As a check on these results, Campbell and Shiller also calculated the statistical correlation over the period between the P/E ratio and subsequent growth in stock prices and earnings.14 They found that the P/E ratio was negatively correlated with subsequent stock price growth but uncorrelated with subsequent earnings growth. They also found that the negative correlation between the P/E ratio and subsequent stock price growth was statistically significant, in the sense that the probability that this correlation was due to pure chance was very small.15 Thus, the statistical results confirm the conclusion from Charts 2 and 3 that movements in the P/E ratio back toward the long-term average have occurred mainly through changes in stock price growth rather than changes in earnings growth.
The Campbell-Shiller study was published in 1998, at which time they predicted “substantial declines in real stock prices, and real stock returns close to zero, over the next ten years.” As the current P/E ratio of the S&P 500 index is actually higher than that at the time of their study, their bearish conclusion would presumably still apply today.16
Note that the Schiller PE ratio isn’t perfect
Why would highly intelligent investors like Rob Arnott and Jeremy Grantham assign so much credence to a metric that contains major flaws? It’s a metric that flashed a sell signal in May 2009, perhaps the best time to buy stocks since the early 1980s.
We’re talking about Robert Shiller’s cyclically adjusted price-to-earnings (CAPE) ratio, for which the Yale University economics professor won a Nobel economics prize. Speaking at the ninth annual Inside ETFs conference, Wharton School of Finance professor Jeremy Siegel analyzed equity market valuations in depth and gave particular attention to Shiller’s CAPE ratio.
It’s clearly an input that should’t be considered in isolation.
Siegel is not the only savant to question whether this metric should be viewed with the reverence some treat it, but he cited some powerful evidence why it can be misleading. Between 1981 and 2015, the CAPE ratio signaled that equities were overvalued in no fewer than 416 of 422 months.
And here’s a tricky part of the analysis.
Look at this graph. Obviously, the average value in 1990 was not the current average of 16 – it was lower. Since 1982, the ratio has been higher than most of the time in the previous century. If you’d used this as your buy/sell guide, you wouldn’t missed out on all the gains during the super profitable 1982-2000 run.
On the other hand, if you’re looking at a 100 year investment cycle, it could be that we’re waiting for a big crash, a massive correction. We see big drops in 2000 and 2008, making out way back down to the mean.
Say in 1982 you knew you were going to die in 2000, and you could see the future. Well, then fuck the Schiller PE ratio, because you planned on cashing out your inflated stocks on hookers and blow in 1999 anyway. Long term reversion to the mean doesn’t matter for you, because your meat bag is expiring in less than 20 years.
As much as the long term reversion to the mean is an attractive and clean scheme to predict the behavior of the market, if relevant timespans are that much greater than a human life, then it’s likely *your* investment portfolio will be dominated by the economic quirks / black swans of the time period in which you happen to live. (e.g. the timing of the Internet Bubble and the housing crash).
Here’s another quote about the next 20 years in the stock market will be meh:
Perceived by some as a raging bull, Siegel concurred with many observers that investors should expect lower equity returns going forward over the next decade. But he predicted equities could generate real returns of 5.0 to 5.7 percent, not dramatically less than the 6.7 percent they have produced over the last two centuries.
What is driving his diminished expectations? Besides slightly high valuations, slow economic growth and the risk aversion associated with an aging population are two of the key determinants.
Flagging productivity doesn’t help, either. One of the biggest questions perplexing economists at present is how the U.S. could add more than 500,000 jobs in November and December while GDP is expected to expand at a rate of 0.7 to 1.7 percent in the fourth quarter. “The collapse of productivity baffles economists,” he said.
Looking out over the next decade, the Congressional Budget Office projects annual GDP growth will struggle to reach 2.0 percent. But Siegel noted that the big decline in expected real returns can be found in the bond market, where investors are likely to see 1.0 or 1.5 percent after inflation.
Ok, so the bond market is crap, and a person definitely should aim to stay there indefinitely. But in theory, if you can predict the next crash, clearly it’s better to be in bonds than take a 20%+ loss.
I should probably start researching more obscure investment opportunities.