Reporting on Stock Moves

Many thanks to Colin Barr, who joined in the conversation after I called him out for his reporting on Merrill yesterday. My problem was with Barr saying that "shares of Merrill Lynch rose 8% after CEO John Thain said he won't entertain talk of a breakup of the brokerage firm". Barr replied:

While I'm as happy as anyone to mock the ridiculous conventions of journalism, I'm not sure that saying the stock rose "after" Thain said something really amounts to "looking for causality." The alternative as I see it is to admit what everyone knows to be true, that rarely do any of us ever have a real lock on what causes a stock to move. Among other things, this coud result in riveting constructions like "Merrill shares rose 8%. Earlier, in a development that may well have been totally unrelated, CEO John Thain said he wouldn't entertain a breakup of the brokerage firm."

As a general rule, if there's something which everyone knows to be true, then it's not a bad idea to admit it. And if the "after" construction doesn't involve "looking for causality", then it would be OK to write that "shares of Merrill Lynch rose 8% after Hugo Lord, of Penge, in south-east London, decided that his herbaceous borders could use a bit of pruning".

The problem here is the uncritical internalization of three assumptions:

1. When shares move, they move for a reason;
2. If there is a reason for a move in share price then it is possible to know the reason for a move in share price;
3. The reason for a move in share price will be related to news about the company in question.

None of these assumptions turns out to be very true in the real world, even if they hold in the world of stock-market theory. And it does a disservice to readers of the financial press when journalists pretend that they're true when they're not.

The way I see it, there are two different kinds of news that financial journalists can report. The first is real-world news: something happened, like a Merrill Lynch annual meeting, or the fact (weirdly ignored by most of the financial press) that Merrill Lynch has just raised $2.55 billion of new capital at an interest rate of 8.625%.

Buried inside such a story, it's perfectly fine to insert a sentence like this one:

Merrill shares rose 7.1%, or $3.18, to $48.09 as of 4 p.m. in New York Stock Exchange composite trading.

When many people read about the company, they're often interested in where the share price is, and what it did that day: there's no reason not to tell them. But the story is very much about the news, and the share-price information is simply added in as a public service.

There's a different kind of story, which is generally only of use to dedicated stock-market watchers, and that's a market report about stocks which went up and stocks which went down. On any given day there are bound to be a few big movers on the stock market, and some people are interested in which stocks those might be. Once again, it's fine to report that fact; it's often a good idea to do so in the context of the stock market as a whole. The index did this, financials did that, MER went up 7.1%.

In such a story there's no harm in mentioning any news about the company: it's germane that the rise in Merrill shares happened on the same day as the annual meeting. But it's clear that the focus of the story is not the real-world news but rather the market movement.

The thing which bothers me is when the two different types of story get conflated, and a market-news story is presented as though it's a real-news story. It often starts with a journalist or editor looking at a big stock-market move and asking why it happened. There's then a search for news, and whatever news about the company happened to come out that day automatically becomes the presumed reason for the change in the stock price. It's lazy and unhelpful journalism, but it's also endemic, so I'll admit it was a bit unfair of me to pick on Colin Barr specifically. Even if his purported reason for the stock move was more unlikely than most.