Stocks

Shiller Should Return to Making Long-Term Return Predictions

Valuation-Informed Indexing #400

By Rob Bennett

I wrote here recently about how Robert Shiller made a prediction in July 1996 that did not come close to proving out. He said in a paper titled “Price-Earnings Ratios as Forecasters of Returns: The Stock Market Outlook in 1996.”  that: “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.” Shiller couldn’t have been more wrong. The annualized real return for the ten years following the day that Shiller made that prediction was 5.9 percent. From the time that he made the prediction through today, the return has been 6.5 percent.

Shiller has been reluctant to offer such predictions in recent years even though valuation levels are now higher than the ones that caused his alarm back in 1996. In several interviews, he has disdained market timing. He indicated in one interview that, while he once believed that some forms of return predictions were possible, he no longer does. He did not say what caused the change in heart but I think that the most likely explanation is that seeing his own prediction fail so spectacularly convinced him that it would be a mistake to try again.

I think that Shiller did a good thing to advance the return prediction he made in 1996. I think he made a mistake when he decided to refrain from making further predictions.

I have personally partaken in the humbling experience of seeing a public prediction fail in spectacular fashion. I wrote a column at this site a number of years back saying that I thought we would see another price crash prior to the end of President Obama’s second term. Things obviously did not turn out that way.. There are a number of critics of my investing work who to this day appear on the comments section of my blog to taunt me about the failed prediction. There have been moments when I wondered whether I should have held back from committing the words of that column to the permanent record of my words that follows me everywhere I go on Planet Internet. So I think I am well positioned to feeling some sympathy for Shiller’s reasons for wanting to get out of the return prediction game.

There’s another side to the story.

I believe that it is possible to make long-term return predictions. I believe that because of the research that Shiller published in 1981 showing that valuations affect long-term returns. If valuations affect long-term returns, the market is not efficient and price changes are not determined by economic realities but by shifts in investor emotions. Those shifts have always caused overpricing to disappear in the long term. So the risk of owning stocks is not static but variable. Stocks are more risky when priced higher and investors need to know how much more risky. Return predictions quantify the added risk. They provide investors who believe that valuations matter guidance on what sort of stock allocation they should be going with at times when prices are temporarily out of control.

Shiller is wrong to suggest that timing doesn’t work. His entire life’s work shows that it must work (at least that’s my opinion). I certainly do not believe that short-term predictions work and I of course fully understand why Shiller would not want to advance predictions for how stocks will perform over the next year or two or three. But, if even long-term predictions don’t work, I don’t see how his work (which caused him to be awarded a Nobel prize in Economics) has value. If valuation levels do not provide guidance as to how much we should change our stock allocations at times of high prices, I don’t see how there is any practical significance to our new-found knowledge that valuations affect long-term returns. Making even long-term predictions is clearly a risky business. But if even long-term predictions do not work at all, the only alternative is to join our Buy-and-Hold friends in sticking with the same stock allocation at all valuation levels.

There are two reasons why it is important for those of us who believe in valuation-informed strategies to offer predictions for how the market is going to perform over the next 10 years or so.

One, it is the fair thing to do. Our Buy-and-Hold friends do not believe that predictions work. Shiller’s research suggests that they should. The Buy-and-Holders are right to seek accountability from those of us speaking from the other side of the table. If we truly believe that Valuation-Informed Indexing is a better strategy, we should demonstrate the courage of our convictions needed to put our names on publicly voiced return predictions. When those predictions fail, it helps the Buy-and-Hold case. And it should. If effective predictions cannot be made, Buy-and-Hold is the best strategy and investors need to know that.

Two, the humbling experience of seeing a prediction fail forces those of us who believe that at least long-term return predictions should be possible back to the drawing board. Shiller got it wrong back in 1996. I got it wrong when I made my prediction after the price recovery of March 2009. As painful as it is to make public mistakes, doing so can bring on a learning experience if we dig deep to discover why we got it wrong. None of us today knows all there is to know about how stock investing works. We all need to be digging deeper all the time. We possess a human inclination to become complacent in our beliefs. Seeing those beliefs fail a public test shakes us out of our complacency.

Rob’s bio is here.