One of the stories in today’s Wall Street Journal describes how a number of U.S. cities are coming to terms with their inability to pay their obligations. Earlier articles in the WSJ and elsewhere gave some details—specifically, in years when investments did better than expected, many cities took the excess returns to add to payments, making the cookie jar smaller when the inevitable underperforming years came. They had confused the short term with the long term.
I get the same kind of question, in a different form, when I speak with investors. Should we invest in the stock market? Well, I say, what is your time frame? Over the long term, you absolutely have to invest in the market. Over the short term, you might be best off not doing so. Is this a one-time investment or a continuing stream of investments? How old are you? And so on.
The point here is that the right thing to do depends on your situation; there is no absolute right answer. Detroit apparently defaulted to the short-term option that was politically easiest, and it’s now paying the price. Many other cities seem to have done the same thing. The federal government is grappling with a similar issue, although the ability to issue debt and run deficits, which states and municipalities (except for Vermont) don’t have, has cushioned the blow thus far.
The short-term focus of cities shows its downsides in the pension obligations, but it also has its positives. Because they can’t run a deficit, local governments were obliged to cut early and hard, which has been a contributory factor to weak growth over the past couple of years. Now that growth has started to pick up again, and many of the cuts mentioned in the WSJ article are starting to take effect, the future should be better. Short-term weakness may actually lead to better future performance, as revenues improve and spending remains lower than it would have been.
This is a general principle, called mean reversion—that is, things tend to revert to average levels over time. One of the most successful institutional investors, Jeremy Grantham, bases his firm’s approach on this idea. Good years will be followed by bad years, but, over time, things will tend to converge on the average. Detroit has been bitten by this, and we will certainly see more examples.
How can we incorporate this concept in our investing? First, by being very skeptical of claims that this time is different. It may be, but extraordinary claims require extraordinary evidence, which usually is lacking.
One example is current corporate profit margins, which are close to all-time highs. One of the reasons for this is that wage growth has been below average—significantly so. Let’s think about this for a minute. Continued growth in corporate earnings requires higher sales growth, which may come, at least in the U.S., primarily from the consumer market. So we’re basing our expectations for continued corporate growth on starving the primary consumers. This doesn’t seem to be consistent and suggests that current success may not be sustained.
Another factor that suggests margins may decline over time is capitalism itself. One of the things that makes capitalism work is competition, so other companies will presumably arise to provide the same services at a lower price—and lower margins. If capitalism works, margins should be driven down again. If not, we may have deeper, systemic problems.
A second example of mean reversion might be stock valuations. If you review the past 20 years, for example, current valuations look very reasonable, or even low. If you extend the data window past that, though, you find that we’re at levels above any time except for the past 20 years, the mid-1960s, and 1929. Perhaps the conditions of the past two decades—declining interest rates, the prime earning and producing years of the baby boomers, among others—will continue for the next two. If not, perhaps the longer data set is more relevant.
I plan to use mean reversion as an organizing principle going forward, as it frames many of the issues I think are important. I will hit more on these specific issues as well, but the broader point is that the time frame for many analyses out there is more limited than I think it should be. We’ll look at multiple examples of that going forward.