We depend on the stats. We view numbers as cold, hard facts and hope they're a good predictor of the future. But they're not infallible. Let's look at the stats on free agent Oliver Perez, who is signing a three-year deal with the Mets for $36 million.Perez, a left-handed pitcher, had an earned run average of 4.22 last year -- not his personal best, but better than some years. Over the last five seasons, Perez's ERA doesn't establish clear expectations. He went from a 2.99 in 2004 with Pittsburgh, to a 6.55 or so in 2006, then back down again to a 3.56 in 2007. Just from the stats, we'd find it hard to predict how Perez will perform this year. But that's true of any pitcher. Stocks have a measure that can predict their behavior during market swings, but only to a point. Beta indicates a stock's likely behavior, but it isn't foolproof. And different analysts can come up with different betas for the same stock. So investors shouldn't expect beta to be the whole story.
Let's look at examples. From Jan. 8 through Jan. 14, the New York Stock Exchange fell 8.7% and the Nasdaq dropped 7.9%. During the same period, tech stock VMware ( VMW - Get Report), which has a beta of 2.5, fell 17.4%, while IBM fell 4.6%. Both stocks' betas were good, though imperfect, measures of market tracking. Beta is often misunderstood in two key ways. First, investors sometimes assume beta is a standalone measure of a stock's volatility. It's not. Beta is a comparative risk measure that cannot be calculated without a benchmark, usually an index of stocks -- the Russell 3000, for example. In this sense, investors sometimes confuse beta with standard deviation, the technical term for volatility. Unlike beta, which is a relative figure, volatility - a measure of how much a stock price fluctuates - can be calculated just based on its historical return numbers, without comparison to a benchmark. We could probably calculate the volatility of Perez's ERA from year to year, based on his history. But we can't calculate his beta from that. Because the benchmarks used to calculate beta vary, a stock's beta listed by different data sources may also vary widely. Some data sources provide a beta figure that is more historical, while others may have more of a forecasting orientation. And data firms will look at a stock's history at different frequencies over different lengths of time. To determine beta, one data source may record a stock price weekly, while another may take monthly samples.
The second common misperception of beta is that it is just a measurement of risk. It measures both a stock's expected risk and expected return against the benchmark. While investors usually want low risk, a low beta also implies a stock will have a lower return relative to the market. Beta was devised as an element of the capital asset pricing model, which is a formula that won William Sharpe a Nobel prize in 1990. CAPM can be used to determine a company's cost of equity. Knowing the cost of capital is useful in determining a stock's future equity value and for assessing how profitably a company has invested its capital. Now let's look at the beta of some of my recent picks and wins: Monday's pick Arcelor Mittal ( MT - Get Report) has a high beta ranging from 2.0 to 2.3. Sources give GlaxoSmithKline ( GSK - Get Report), my Jan. 28 pick, a beta of 0.6. Cisco ( CSCO - Get Report), my Oct. 10 win, rates a beta of 1.0 to 1.3, while Tesoro ( TSO), my win from Sept. 12, has a beta ranging from 1.6 to 1.7. Microsoft ( MSFT - Get Report), which has paid off for me many times, has a beta ranging from 0.8 to 1.1 As you see from the variation, beta doesn't play a big role in my DITM strategy. It's just another tool in the box. Lenny Dykstra manages Nails on the Numbers, a subscription service sold by TheStreet.com. Dykstra is 94-1 in his options picks. Click here for a free trial to Nails on the Numbers. Mr. Dykstra writes regularly about options trades for TheStreet.com.