The good news is that U.S. stocks have had a great run – and especially so since the November election. The more worrying thought is that they look more expensive.
We know that when stock valuations are high, returns in future years are lower -- and vice verse. From over 100 years of stock data, we have good evidence that the highest returns come from buying when stocks are cheapest.
So what is one to do today, when the stock market trades somewhere between 17 and 19 times expected earnings versus a historical average of 15?
A 2013 paper by Dimson, Marsh, and Staunton called “Mean Reversion” says probably nothing. A high PE relative to historic standards may make us think the market is too expensive – but it’s not a good enough signal to try timing the market.
The paper looked at a strategy of entering the stock market when valuation multiples were below the historic mean and exiting when they were above the historic mean in 20 different countries. In each case, using valuation signals to time the market did worse than a simple buy-and-hold strategy.
The problem is not with the idea of buying into stock markets when they are cheap or selling when they are expensive – that is something the authors agree is sound. The problem is that it is hard to turn this into a realistic strategy. Because market valuations can drift into different regimes for long periods of time, it is impossible to determine if a market is over- or undervalued just by comparing current valuation signals to past ones.
When we look at past data, we have the benefit of hindsight and can divide the data into quintiles or deciles to clearly identify periods of high and low valuation. That makes it easy to believe that we’ll be able to execute a strategy of buying low and selling high. But it’s an illusion to believe it can be done in real time: “Prices may look cheap compared to recent years, and simultaneously expensive versus their long-run average. Or they may look cheap in one country, and expensive in another. We cannot know in advance what valuation level is going to prevail at some point in the (possibly very distant) future.”
So yes, stocks may look somewhat expensive relative to the past. But we don’t know enough to take action – especially when the cost of being un-invested or underinvested can be large.