In a [link|http://papers.ssrn.com/sol3/papers.cfm?abstract_id=970413|working paper] titled "Where are the Shareholders' Mansions?" David Yermack of New York University and Crocker Liu of Arizona State wonder whether there is a relationship between CEO home-buying behavior and stock performance.

[...]

Using property records, public databases, and search engines, Yermack and Liu were able to identify the primary residences of 488 of the 500 CEOs of the S&P 500. These guys\ufffdand they're almost all guys\ufffdare living large. The mean residence of a CEO was anything but mean: 6,145 square feet, 12 rooms, 5.37 acres of land, and a market value of $3.1 million. For the 164 in the sample who bought new houses after being named CEO, the mean house was even less mean: 6,635 square feet, 13.1 rooms, 6.13 acres, and a market value of $3.9 million. "Aerial photographs indicate that outdoor swimming pools, tennis courts, boathouses, formal gardens, and detached guest houses or servants' quarters are common features of CEOs' homesteads," Yermak and Liu write. CEOs are also engaging in the same sort of financing that the home-buying masses do.

Then the professors examined the returns of the CEOs' stocks, and discovered that the bigger the home, the worse the stock performed. In 2005, the stocks of companies whose CEOs lived in larger homes (i.e., above the average for all CEOs) returned, on average, 3.35 percent less than companies whose CEOs lived in below-average homes. And the CEOs who lived in the biggest homes (at least 10,000 square feet or over 10 acres) underperformed their peers who inhabited more modest homes by 6.9 percent, on average.

Next, they looked at stock returns for 164 companies whose CEOs bought new homes after becoming CEO. Here, again, they found trouble, especially for CEOs who bought mega-homes on mega-plots. They found "a significantly negative stock performance following the acquisition of very large homes by company CEOs," on the order of 1.25 percent performance lag per month. The conclusion: "We interpret the stock return evidence as consistent with large CEO home purchase indicating entrenchment and foreshadowing poor future stock performance." What's more, when CEOs sold stock to finance the purchase of a home\ufffdas was the case in 32 percent of the instances\ufffdthose stocks performed worse than companies where CEOs held on to their stocks. Yermack says this is the most interesting finding. The data seem to indicate that a good number of CEOs are selling shares\ufffdostensibly to raise cash to buy a house\ufffdjust before their stock goes south. "I think it could be that these guys know something and want to get out of the stock," he said. "Ordinarily, you can't do that as an insider unless you have an alibi, and buying a house is about the best alibi you can come up with."