(MoneyWatch) When Richard Roll made his presidential address before members of the American Finance Association in 1988, he shared with those in attendance a surprising conclusion he had reached: Only about one-third of the variation in market indexes can be attributed to economic influences. Here we are, 25 years later, and updated research shows that we may know even less about what causes markets to move than we did at the time.
Roll's address inspired the study "What Moves Stock Prices?," published in the Spring 1989 edition of the Journal of Portfolio Management. The authors examined the 50 largest one-day returns on the S&P 500 index over the period 1946 to 1987 and found that many of the index's largest movements couldn't be linked to any convincing account of why future profits, or discount rates, might have changed: "Our inability to identify fundamental shocks that accounted for these significant market moves is difficult to reconcile with the view that such shocks account for most of the variation in stock returns."
Perhaps motivated by the massive increase in media attention given to financial news, Bradford Cornell of the California Institute of Technology updated the prior research in a paper published in the Spring 2013 edition of the Journal of Portfolio Management. His study covered the 25-year period 1988 to 2012 and examined the 50 largest moves in the market to see how well they could be explained by the day's events.
Cornell found that "with a few exceptions, such as such as the terrorist attacks of September 11, 2001, and the decision by governments throughout the world to support troubled banks on October 13, 2008, it is hard to conceive how the news arriving on the day in question led to revaluation of the entire market value of U.S. equity on the order of 5 to 10 percent."
Thus he came to the same conclusion as his predecessors: "Despite the passage of time and the massive improvement in information technology, it is, if anything, more difficult to tie major stock price movements to fundamental economic news sufficient to rationalize the size of the observed move. The bottom line is that the source of large movements in overall market prices remains as much of a mystery today as when Cutler, Poterba, and Summers published their paper in 1989." He concluded: "If anything, the mystery has deepened because the size of the unexplained market movements has grown."
In his paper, Cornell makes an important point we've repeatedly emphasized: Whether the news is good or bad doesn't really matter. What matters is whether it's a surprise. Here are two examples.
The government seizure of Fannie Mae and Freddie Mac on Sept. 6, 2008, had all the earmarks of a big negative event, but the market return on that day doesn't rank among the top 50 -- presumably because the bankruptcy was largely anticipated by the time it occurred. On the other hand, market movements on Aug. 8, 2011, after the U.S. government's debt was downgraded, did make the list. This was an event that should have been anticipated because it followed weeks of speculation that such a downgrade was imminent. There's really no logic to why one day made the list and the other didn't. Yet the media always look for explanations, presumably because they think the public wants them, even if the media know they're likely to be meaningless. Cornell notes that the explanations are most often about fears and doubts about some particular issue -- what I would call psychobabble. Yet there are plenty of advisors at Wall Street's marketing machines ready to provide those explanations, along with advice on what you should do to protect yourself.
Of further interest is that Cornell found that the major moves tend to "bunch." Of the 50 largest movements, 14 occurred on pairs of consecutive days. If a gap of one day is allowed, 21 of the 50 largest movements were bunched in strings of "consecutive" days. He also found that all but one of the big returns on consecutive days involved a reversal! Cornell notes: "Such reversals are odd because a very large stock price movement, if rational, requires long-run changes in investor expectations regarding future cash flows and discount rates. It is peculiar that long-run views on the part of investors would be altered on one day only to be reversed on the next."
Cornell's paper caught my attention because almost any time there's a major downward move in the stock market, I receive a call from some member of the financial media who asks me to explain what happened and to provide a forecast and recommendation on what investors should do. While sometimes it's easy to explain why markets crashed (as when the markets opened after the events of Sept. 11, 2001), many times I've responded by stating that there's no logical explanation for such a large move. As for the forecast, my typical response is along the lines of: "The market will go up and down but not necessarily in that order," or "My crystal ball is always cloudy, how's yours?" And when asked what investors should do, my response is always the same: "If you have a written investment plan, stick with it, and if you don't have one, sit down immediately and write one."
Image courtesy of Flickr user 401(K) 2013