Media attention drives stocks, at least at first

5 min read

Like celebrities who first are revered and then reviled, stocks get a boost from increasing media coverage but then underperform.

And high levels of coverage have a lasting positive effect.

That’s the upshot of a new study which looked at 89 years of company news coverage in the New York Times.

“Stocks with increases in media coverage outperform stocks with decreases in media coverage during the formation year and underperform during the subsequent two years,” Alexander Hillert and Michael Ungeheuer of the University of Mannheim write in the study, which studied coverage in the Times from 1924 to 2013. (Full disclosure: some of my columns appear in the Global New York Times)

Specifically those stocks which are in the top quintile in terms of increase of coverage beat those in the bottom one by 10.68 percentage points in the first year, only to underperform the same group by 5.04 points over the next two years.

It isn’t just relative changes in media coverage which can predict stock returns: absolute levels matter too. Top quintile stocks in terms of the number of stories outperform bottom quintile ones by 2.76 percentage points a year, a finding in contrast to earlier studies. That outperformance persists, with the top quintile beating the bottom by 2.76 percentage points in the second year and 2.28 in the third.

The phenomena of stocks doing well when media coverage increases makes a certain amount of sense, being consistent with the idea that stocks may become overvalued when enthusiastic new investors find out about them. They may then overshoot their fundamental value, giving some of it back in subsequent years. This is both the essential underlying dynamic of momentum strategies, which seek to buy that which is going up, but also shows the weakness.

There is also an agency angle here. Managers and insiders at firms which get a rapid increase in coverage may be able to turn it to good effect for themselves, using it as an opportunity to pursue corporate events like public offerings or acquisitions which are driven more by being “hot” than by fundamental strategy. That can enable “empire building” at firms, which destroys value as well as serving as an opportunity for insider sales.

A puzzle, or advertising pays?

The finding that more coverage leads to a durable outperformance is a bit of a puzzle, being hard to reconcile with many of our expectations about how capital markets work. As early as 1987 Robert Merton, in his asset-pricing model, maintained that a higher level of company recognition leads to lower expected returns. Merton’s thinking was that the kind of diversified investors who pick up on highly recognized stocks themselves require a lower future return because their overall portfolio returns are more stable.

Many investors have also argued in favor of small capitalization or thinly followed stocks on similar lines. Less well-known stocks, as a group, should in theory offer better returns in order to compensate investors for the work of finding them. This “diamond in the rough” idea does not accord with the findings of this particular study.

Interestingly the result of a separate analysis of tone of stories was mixed, with a lack of consistent relationships between positive or negative stories and stock returns.

While the study does not address this, it seems possible that much of what is being described here is driven by the behavioral tendencies of investors rather than by the fundamental strengths or weaknesses of the stocks. Perhaps what we are seeing is simply the result of exposure creating opportunity for stock purchases.

While professional analysts and fund managers read the media closely in order to find reasons to both buy and sell shares, this may be far less true for the average investor.

As a general rule people don’t read the paper in order to find out about movies not to watch or books not to read. They look for things to become attracted to and consume.

Perhaps it is the same with stocks, with all coverage being of net benefit, though more at first, because it serves as effective advertising, prompting more purchases of stocks than sales.

To be sure, how this data holds up in a broader Internet age is open to question, but it is a long data set and can’t be lightly ignored.

More news, and more frequent news, may simply be good news for investors.

James Saft