When people use the word “bubble,” they are suggesting that prices are about to pop and head downward rapidly. If it were possible to know when that is about to happen, it would be possible to engage in short-term market timing. I don’t think short-term timing works. I think it is possible to say when stocks are overpriced. But I don’t think it is possible to know when a price drop is imminent, or when there is a bubble that is about to pop.
People want to know this. It amazes me how much mental energy people put into the project of knowing how stocks are going to perform over coming weeks or coming months when efforts to make assessments just do not seem to produce results any better than what would be produced by random guesses. Has anyone ever shown an ability to identify bubbles? Has anyone ever even demonstrated that such a thing as a bubble has ever been known to exist?
Instead of trying to distinguish risky bubble conditions from non-risky non-bubble conditions, it seems to me to make more sense to view stock investing risk as a continuum. Stock risk increases gradually as valuations rise. I don’t think there is any exception to that rule. But there is no price at which stocks can definitely be said to have entered bubble territory and no level of overvaluation that can be said to be entirely free of risk. You want to avoid investing heavily in stocks at times of high prices But you can never know in advance whether or not a certain level of overvaluation is going to produce a bad result in a specified amount of time.
Say that you were concerned about drunk driving and you wanted to warn drivers never to drink enough beer to put them in extreme danger, a level of alcohol consumption analogous to what we think of as bubble territory in the stock investing realm. Could it be done?
You certainly could say that it is best not to drink at all when driving. Zero alcohol consumption does not permit a risk-free driving experience. Accidents happen all the time to drivers who did not take a single drink before starting out on their journeys. But the only way to protect yourself entirely from the risk of alcohol-impaired driving is not to drink at all. Even small amounts of alcohol consumption increase risk. So it is in the investing realm. Even P/E10 levels only a few points above the fair-value P/E10 levels present small levels of risk. But those risk levels are indeed limited.
What about high levels of alcohol consumption? Someone who drinks six beers and then gets behind the wheel of a car is taking on an insane level of risk. It is a foolish thing to do. However, the reality is that there are many cases in which someone has had six beers and then driven a car and did not get in an accident as a result. If we applied the terminology used in the investing realm to the drunk driving situation, we could say that driving after drinking six beers is not necessarily entering bubble territory. It is entering high-risk territory. That’s what matters. I think it would be fair to say that stocks are selling at high-risk levels today. Those prices don’t guaranty an imminent crash. But I don’t think that those going with high stock allocations can derive much comfort from that.
There are cases in which drinking three beers will get you in an accident and there are cases in which it will not. What conclusions can we draw from that? I worry that, in the stock investing realm, the conclusion that is drawn when high prices do not produce a crash is that high prices are not so terribly dangerous. I see that as an unfortunate conclusion. The results that follow from being involved in either a car crash or a stock crash can be devastating. The aim should be to avoid these experiences. The fact that we cannot say with precision what sorts of inputs (cans of beer or levels of P/E10) will bring them on does not justify daredevil behavior (driving after consuming several cans of beer or investing heavily in stocks when prices are very high). Unfortunately, it seems to be human nature to conclude that, if doing dangerous things does not produce a wipeout on one or two occasions, it thereby becomes safe to take the chance a second or third time.
Everyone wants to minimize risk when investing in stocks. The trouble seems to lie in the manner in which the human mind assesses risk. Engaging in risky behavior that does not produce bad results causes us to let our guard down. In the drunk driving realm, there is nothing like hearing about a friend who was killed or seriously injured after drinking a few beers to put fear back in our hearts. Unfortunately, there is no way to experience close calls in the investing realm. When crashes arrive, they hurt everyone at the same time. It’s too late to protect ourselves from danger by the time that we hear about others experiencing it.
We can know when stocks are dangerous. The P/E10 level reveals the level of danger. There is a little bit of risk when stocks are modestly overvalued and then a lot when they are very much overvalued. But there is no particular P/E10 level that we can use to distinguish times when stocks are at “bubble” levels and when they are not. A P/E10 level that in one circumstance will produce a wipe-out crash will in another circumstance produce no downturn in prices or will even bring an upturn in prices. Crashes are the product of irrationality. So we cannot expect them to turn up in predictable ways.
I cannot say that stock prices have entered bubble territory today. I can say that they are so high that the risk of owning stocks is much greater than it would be if prices were more reasonable.
Rob’s bio is here.