CEOs, Corporate Performance, and Houses

Yesterday's Wall Street Journal featured, as one of its three front page articles, a piece by Mark Maremont describing academic research into the relationship between various events and actions in the private lives of CEOs, and the stock price performances of the companies they headed.

The fact that this article was given such prominence tells you that the editors of the Journal feel it is important and reputable. Perhaps the most interesting and methodologically valid piece of research presented in the article was one by Yermack (Penn State) and Crocker (Arizona State) relating CEO home building and subsequent stock price performance of their companies. While it's worth reading the Journal article in its entirety, the short version of this particular research effort's result is to find that those CEOs who built/purchased large homes ( >10,000 sq ft) saw their companies' stock prices, on average, decline 25% in three years, while those with 'other' homes saw, on average, increased stock prices of roughly 20% by year three.

No measures of statistical significance were provided in the Journal article, although one surmises that one is available in Yermack's and Crocker's paper. Their approach was novel and inventive. They chose a single point in time, the end of 2004, and secured housing data on 488 of the S&P500 CEOs.

The basic premise being described, of course, is that of a CEO engaged in building or buying a large home is taking his mind and attention off of shareholders' welfare, and concentrating, instead, on spending his compensation on luxurious trappings of power. Among the CEOs mentioned in this manner, by name were: Leslie Wexner (The Limited), Stephen Bollenback (Hilton), and Trevor Fetter (Tenet Healthcare).

So far, so good...and simple-minded.

You see, there is another possible, and reasonable explanation for this phenomenon. But you'd have to be open-minded about observing and measuring corporate performance in order to see it.

My proprietary research has indicated that once companies have consistently outperformed the S&P500 Index on a total return basis for five or more years, the chances of this performance continuing falls precipitously.

Thus, a CEO building a large house with the fruits of her/his successful efforts at outperforming the S&P in the prior years would make sense. And the probabilities of continued outperformance would be slim, anyway.

Thus, Yermack and Crocker may have simply validated that succesful CEOs reap the rewards of their efforts for shareholders, and the game is over for a few years.

As I look at the nearby, Yahoo-sourced stock price chart of The Limited (Brands), Hilton Hotels, and Tenet Healthcare, all mentioned by name in the research piece, I draw a different conclusion than the study's authors.

To me, up through 2004, while the Limited struggled from 2000 onward, staying roughly even with the S&P, Hilton outperformed the index, and Tenet outperformed for about 4 years, then fell sharply.

Thus, I find Hilton and Limited to not be examples of the behaviors suggested in the Yermack and Crocker research. Tenet may be. It's declined has persisted since 2003.

The frame of reference you use, and what you know about potential corporate total return performance, matters in how you evaluate subsequent CEO behaviors after a period of good or great performance versus the index. I don't find the study that the Journal article spotlighted to be so bullet proof now as I did when I initially read of it.

Now, I would be very interested to know how their results would differ if they applied my knowledge of the probabilities of corporate consistently superior performance over time, versus the S&P, to their existing work.