David Hume, Causation, and the Markets

As I wrote here last week, TV finance pundits regularly tells us that price moves are the result of "momentum," or "bargain buying," or "profit taking." Together, these labels are so widely applicable as to drain themselves of any real significance. So let us put themn aside and try some other route forward.

On Tuesday, after the market closed, the wire services carried the following, “Morgan Stanley downgrades XYZ.” On Wednesday, nobody could be found to buy XYZ at the previous day’s price. Is that an answer to the question “why”?
 

This is complicated. One line of argument would contend that the words “Morgan Stanley downgrades XYZ” didn’t tell the market anything the market didn’t already know.

The headline allusion is actually to what some researcher paid by Morgan Stanley has said in a 30-page report. That report itself is most likely the compilation of publicly available data: John Smith (as we’ll call the analyst) with admirable industry dove through a lot of SEC filings and such and has added the benefit of his own analysis. Why should that move the market?
 

Look at it this way: If Joe Smith wanted to do business as “Joe’s stock analysis shop” and peddle his 30-page reports … well, one anticipates start-up difficulty. Who would especially care what Joe thinks?

We might suspect, then, that if anything moved the market it was the invocation of the “Morgan Stanley” brand name. Is this rational, or does this suggest that markets sometimes move irrationally? 

Suppose we decide to study the question empirically. We build a database of instances between 2000 and 2010 in which the wire services have linked the name of a stock to (a) the name of a well-known investment bank with analysts on the payroll and (b) the word “upgrade” or “downgrade.” We then examine cases in which the first such report appeared between the close of one trading day and the start of the next. Does the word “downgrade” foretell a fall in the price of that stock when the market opens the following day?

If it does, then, on what one might fairly call the David Hume theory of market causation, downgrades cause price falls. After all, causation is nothing but the regular sequence of A and then B. what else could it be?

But I find this unsatisfactory.