Housing Bubbles and Insider Sales (from WSJ)

Talk is always cheap. That's why I like prediction markets like Tradesport.com and the Iowa Electornic Markets. the prices of these contracts give you an idea of the predictions by people who are willing to put their money where their mouth is. So, you could say that the purchase of a prediction-market contract is a costly signal.

Insider trades are also often viewed as a signal. If we believe that corporate insiders have better information than outsiders about their companies' prospects, their trading behavoir cold have predictive power.

Tuesday's Wall Street Journal (online subscription required) has an article titled Insider Sales Hit Record at Builders that indicates insiders at home-building firms are engaging in insider selling, and that it could signal the end of the housing boom. It says:
Merrill Lynch quantitative analyst Richard Bernstein said he has seen record net insider selling during the past 10 months at eight of the 12 home-building companies he has tracked.

"This record level of selling comes despite the general perception that the stocks are undervalued, and despite that Wall Street analysts have few 'sell' recommendations on the stocks," Mr. Bernstein said in a note to clients.

He sees parallels between the housing sector and the technology industry before the bursting of the tech bubble. He said that within a year of the tech bubble's peak, about 14 of the 20 tech companies he tracked at the time reported record net insider sales and 16 of 20 within 18 months of the peak.

He could be right, but there are a few reasons to be skeptical. First off, "net insider selling" is typically defined as "sales-purchases". Most academic studies of insider trading indicate that insider selling is a lot less informative than insider buying. Insiders purchase their company's stock because they believe it's undervalued. OTOH, they sell for a variaty of reasons - the company is overvalued, they want to diversify (and they're recently gotten a lot of stock because of stock or option grants), or they simply need liquidity (maybe their teenage son just wrapped the family Porshe around a telephone pole). So, aggragating sales and purchases (like the net selling measure does) might not be a very informative measure.

In addition, the analyst quoted in the story has a pretty small sample to work with - 12 firms. Still, it's food for thought.