New paper offers solution to problem of beta calculations

BVWireIssue #157-3
October 21, 2015

A new research paper puts forth a beta estimation model designed to eliminate the idiosyncratic noise and statistical chance with other calculations of beta. The paper has received enough interest to have it listed on a top 10 download list on the Social Science Research Network (SSRN).

Noise trouble: The abstract to the paper, Idiosyncratic Noise Filtered (INF) Beta, points out that making a reliable estimate of the beta of an individual stock has been elusive because price movements are a function of market/systematic movements and company-specific/idiosyncratic movements. “Since traditional beta estimation calculations make no attempt to minimize the impact of the idiosyncratic movements, the beta regression coefficient will be fooled to best fit both the desired market-based movements as well as the undesired idiosyncratic movements. If the idiosyncratic movements are not perfectly uncorrelated with market movements, the regression beta will be inaccurate.”

BVWire asked Mark Mitchell, director of valuation services at Peterson Sullivan LLP, to comment after reading the paper. “To my knowledge, the use of trading volume in the precise manner presented in the paper is a new concept that hasn’t been presented in other papers,” says Mitchell. “I think the paper advances the ball quite a bit in addressing some of the problems (which are well documented) with beta calculations. More importantly, a solution to the problem is offered and the results appear to be promising.”

Simple elegance: “The solution has a certain simple elegance to it and is intuitive inasmuch as even casual observation of the market (particularly as it relates to smaller stocks) suggests that price swings often occur around volume spikes due to company-specific news events,” adds Mitchell. “I believe the paper’s approach to include that data rather than simply eliminate it in estimating beta is also a unique take on the problem. In addition, the paper makes an effort to test the model against growth-adjusted P/E ratios, an easy-to-understand concept that supports the notions considered.”

The paper’s authors are Bob Dohmeyer (Dohmeyer Valuation Corp.); Igor Gorshunov, a business valuation and private equity professional; and Peter Butler (Valtrend LLC).

In Mitchell’s opinion, the paper addresses a topic that isn’t discussed as much as it should be in the business valuation literature. “CAPM is still a widely taught construct for estimating the cost of equity despite all of the literature devoted to pointing out its shortcomings,” he says. “At the very least, the paper’s ideas are a valuable basis for further discussion.”

Please let us know if you have any comments about this article or enhancements you would like to see.