It Doesn’t Matter All That Much How Long It Takes for Return Predictions to Work

Updated on

 Valuation-Informed Indexing #406 on investor emotion and the impact on long-term S&P 500 returns

By Rob Bennett

Get The Full Warren Buffett Series in PDF

Get the entire 10-part series on Warren Buffett in PDF. Save it to your desktop, read it on your tablet, or email to your colleagues

Robert Shiller made a prediction in a paper published in 1996. He wrote: ““The market is quite likely to decline substantially in value over the succeeding ten years; it appears that long run investors should stay out of the market for the next decade.” The annualized real return for the ten years following the day that Shiller made that prediction was 5.9 percent. Shiller’s prediction did not prove out.

Does it matter?

It obviously matters to some extent. Shiller showed us that the market is not efficient, that valuations affect long-term returns. That finding opens up lots of exciting possibilities for stock investors. It shows that stock investing risk is not static but variable. Investors who employ the same lifetime asset allocations as Buy-and-Holders but go with higher stock allocations when stock valuations are low and lower stock allocations when stock allocations are high will over the course of an investing lifetime earn higher returns while taking on less risk. Being able to predict future returns gives a stock investor a big edge.

But Shiller is a Nobel-prize-winning economist. If he can not use his research findings to make effective long-term stock predictions, no one can. And Shiller’s prediction failed. What hope do any of the rest of us have?

Shiller really did get it wrong. But I am not sure that it matters all that much. The entire historical record shows that it is hard to make effective return predictions. Shiller knows this. That’s why he put his return prediction so far out into the future. The historical record shows that short-term predictions rarely work but that long-term predictions usually do. Shiller went out a full ten years because that had always worked in the past. But there’s obviously no guarantee that what worked in the past will work in the future. Going out 10 years did not work in 1996.

I think we focus too much on when the prediction pays off rather than on why it pays off. Shiller says that stock price changes are caused by shifts in investor emotion. If that’s so, it’s easy to understand why it would be hard to get the timing of a prediction correct. Emotion is an irrational phenomenon. So who can say when it will shift? But, if price changes are caused by shifts in emotion, we can know for certain that there will at some time be a price movement in the opposite direction. It is the entire purpose of a market to get prices right. So, if they go wrong in one direction for a time, they have to reverse. The direction of the price movement is certain, only the timing of it is unpredictable.

Stock returns have been poor from 2000 forward (the return for the past 18 years has been 3.2 percent real, half of the ordinary long-term return of 6.5 percent real). And, even after those 18 years of sub-par returns, stocks are still priced at two times fair value. If we see a 50 percent price crash sometime over the next few years (something we should expect to see, if Shiller is right that it is investor emotion that drives stock price changes rather than economic realities), investors who listened to Shiller in 1996 and lowered their stock allocations as a result will be glad they did. $10,000 invested in stocks in 1996 would be worth $40,000 today. A 50 percent price crash would bring that down to $20,000. Treasury Inflation-Protected Securities and IBonds were both paying long-term returns of 4 percent real at the top of the bubble. An investor who switched a portion of his portfolio to a far safer asset class in response to Shiller’s prediction could have earned an equal or greater return despite the fact that Shiller ended up being wrong.

Buy-and-Holders of course do not believe in this sort of mathematical exercise. They think it is silly to assume that a 50 percent crash is coming because they believe that price changes are caused by economic developments, which cannot be foreseen. That’s the issue that matters. If Shiller is right that price changes are caused by shifts in investor emotion, he is telling us something very important regardless of whether his predictions turn out or not. If he is wrong about price changes being caused by shifts in investor emotion, then of course he is telling us nothing of any significance at all and the Buy-and-Holders are right.

Those of us who believe that valuations affect long-term returns need to be cautious when making predictions. It is obviously hard to get the timing of price shifts right. But I think we need to continue to point out how different a world it is in which all investors live now that we have Shiller’s research showing that valuations affect long-term returns and that making long-term return predictions is one way to do that. We will never get the timing of all of our predictions exactly right. But we will get more right than wrong if we show other investors how important it is to take the price at which stocks are selling into account when setting their stock allocations.

Rob’s bio is here.

Leave a Comment