Today’s New York Times has an article reporting that financial markets are much more volatile these days than they have been historically [http://www.nytimes.com/2011/09/12/business/economy/stock-markets-sharp-swings-grow-more-frequent.html?_r=1&hp]. Three percent or greater swings in market indices—both within a trading day and between days—used to be rare, but now are commonplace. The facts are clear, but the cause is not. The article discusses several possibilities, all of them internal to the current world economic situation and/or the velocity with which trading occurs in our computerized age. There may be truth to some or all of the possibilities, but I want to suggest another possible causal factor that has nothing to do with market fundamentals, economic events, or trading practices. I want to implicate the media.
I get almost all my news from NPR, the Times, or the PBS Newshour. These sources are all models of probity; they all wear sensible shoes. Nonetheless, what I hear and read virtually every day from these sources are locutions like “financial markets soared (or plummeted) on, or after, or with, reports that Greece is closer to default, or Bank of America’s earnings were below expectations, or high-tech sector sales are down, or [fill in the blanks.] I want you to focus on those three little words—“on,” “ after,” and “with.” None of those words says outright that the financial news event in question caused the market rise or fall. Financial reporters are much too savvy to make clear causal statements outright. But each of these words implies causality. They imply a connection between the financial event and the market’s behavior. If not, why would you even use those words. You would report on hi-tech sales, and you would report on market behavior, just as you might report on the September 11 anniversary observances and the winner of the women’s singles at the U.S. Open Tennis Championship.
It is a part of virtually every introductory psychology course to make the point that “correlation does not imply causation,” and you might think that this somewhat careless financial reporting is merely an example of the commonly-made correlation/causation mistake. Sadly, it’s even worse than that. What we might say here is that “coincidence does not imply correlation.” With a single event in the antecedent category (eg., low earnings for Bank of America) and a single event in the consequent category (market prices), no correlation can meaningfully be computed. One-time events do not “correlate” with anything. At the very least, reporters or analysts would have to put these antecedents into classes or categories, so that they might be able to say something like “past reports of low earnings from financial institutions have been associated with drops in the market.” Obviously, reporters never do that. But even if they did, the placing of isolated events into classes is by no means uncontroversial. Can we really treat a low earnings report in 2011 as equivalent to a low earnings report in 1975? Doesn’t context matter? And what industry from the 1960’s is analogous to the hi-tech industry today?
So, assuming that we can agree that even the most respectable news sources can be sloppy, what does that have to do with increased volatility? Here’s my speculation. When we hear, again and again, that the market dropped on, after, or with some financial event, we “learn” two things—both of which may be false. The first thing we learn is that financial events of that type cause changes in the market. The second thing we learn is that a causal analysis of market behavior is actually possible. Now, when we read about significant financial events, we feel compelled to act on them, because of the implied causal role they play in market behavior. And since everyone hears the same news, you get herds of investors acting on the news and the causal relation that it implies. The financial events themselves may not be causal (a model of market behavior that remains plausible decades after it was first proposed treats the behavior of stocks as largely random), but news reports of those events may be causal.
More than three-quarters of a century ago, noted economist John Maynard Keynes famously wrote of “animal spirits” (read “psychology”) and the enormous effect they had on the economy. About this, as much else, Keynes was surely correct. But he did not have much to say about what fed the animals. And I’m proposing that one significant source of animal nourishment is thinly disguised, but completely unjustified, causal implications that have become a part of our daily diet of financial news.
What can we do about this? As individuals, I suppose we can vow not to act on “with,” “after,” and “on.” But (a) as individuals we will affect little, and (b) worse yet, if everyone else is acting on them, we may take a financial bath. More likely to be helpful is if we campaign vigorously to stop this carelessness. Every time you see an “after” in a Times story, write a letter to the editor. Every time you hear an “after” on NPR, send an email. Maybe, eventually, they’ll get the message. There probably isn’t much we can do to stop Jim Kramer’s hysterical rants about what the financial future holds, but we may be able to stop the very influential feeders or our animal spirits from continuing the feast. They need to help us go on a news diet.