Wednesday, December 5, 2012

Used Car Auctions, Disclosure and Investor Relations


Economists love auctions. Nowhere else can they as clearly observe the interplay of various causes upon supply and demand. It is capitalism at its most naked and allows economists to indulge their insatiable desire to measure the effect of different inputs upon prices. And sometimes they even come up with some useful stuff. 
The finance section of the November 24th edition of The Economist magazine featured an article on the work of several microeconomists that might actually have some useful application to real world markets. The one that caught my eye was a study by a couple of economists on the effect of information disclosure on used car auctions. (Information Disclosure as a Matching Mechanism: Theory and Evidence from a Field Experiment by Steven Tadelis and Florian Zettelmeyer, Electronic copy available at: http://ssrn.com/abstract=1872465).

Now you may roll your eyes and mutter, “What do car auctions have to do with investor relations?” but indulge me for a moment. What we are interested with here is the information being disclosed and its affect upon pricing, not the item being sold. To quote the authors of the article:

“A market’s efficiency critically depends on whether its participants have sufficient information about the nature of the goods and services being traded. The potential hazard a buyer faces when trading in markets with information asymmetries often leads to market imperfections and stifles efficient trade. Indeed, in resale, housing, labor, health care, and corporate securities markets, sellers may have better information than buyers about the good or service being traded. Furthermore, sellers may have control over how much information to disclose, and buyers may choose how much information to acquire.”
In a survey of over 8,000 used car auctions, what the authors found was that increased information disclosure regarding the quality of the cars increased expected revenues. This is in line with current academic theory. But there was an interesting twist to their findings. Cars in the middle of quality rankings saw only modest gains while the biggest gains in revenue resulting from increased disclosure came for the best and worst quality cars. You might expect this with higher quality cars, as people will bid up the price if they know they are getting higher quality, but if the disclosures are of bad quality, logic would lead you to the conclusion that prices would go down further. Here’s what the authors say about their findings:
“When disclosed information coincides with expectations given observables, then it does not affect the composition of bidders who bid on the vehicle, and as a consequence, the outcomes are the same as they would be without information disclosure. However, when the information disclosed is either a positive or negative surprise relative to expectations, it will attract bidders who are relatively strong given the disclosed information. This benefits the seller regardless of whether information is good or bad news.”
In other words, the additional disclosures, even if they are bad news, attracts new bidders who are interested in that class of goods and who may have a better understanding of the merchandise and its value to them. The result is that they pay more than bidders with a more general outlook. 
Now think of the implications for the stock market. Typically, weak performing companies tend to disclose as little as possible about their problems. As a result, there is usually a fair amount of uncertainty about their future performance and that helps keep their stock price depressed. According to this new research however, they would be better off to more fully describe their problems. By eliminating some of the information uncertainty, even though the news is bad, they will attract more value and deep value investors who are better able to assess the risks of the investment. The result would be a better alignment of the firm’s intrinsic value with its market value. (Of course, it will still be a lousy stock price because the outlook is bad, but it will be a better lousy stock price than without the additional disclosures.)
I would love to be able to test this academic theory in a real world setting where a company has hit a bad patch, but something tells me that I am more likely to find an honest used car salesman than I am to find a CEO willing to bare all about his company in bad times.