Passive and active portfolio managers may be taking on more risk than they realize.
Wilshire Associates Inc., Santa Monica, Calif., calculates that an equity manager could take on as much as 5% value at risk. For a $10 million portfolio, that could mean a $500,000 unexpected risk exposure during the course of a year.
Biased risk estimation by portfolio optimizers is at fault. On average, overestimated risks tend to cancel out underestimated risks. But managers don't favor a mix of risky and less risky stocks — they seek stocks that provide the most bang for the buck.
The problem is that the optimizer tries to minimize risk, and thus "tends to favor portfolios for which risk is underestimated," said Peter Matheos, a Wilshire managing director, in a news release.
Even active managers that don't use an optimizer in selecting stocks may be vulnerable to these biases, said Robert Kuberek, a senior managing director who oversees quantitative research and software development for analytic products offered by Wilshire, in an interview.
Some active managers buy securities based on their risk ratings, as determined by the optimizer's covariance matrix, he said. For example, in choosing between two stocks, a portfolio manager may pick the one that offers the higher reward/risk ratio, based on faulty data.
Wilshire is now correcting for these estimation biases in its SHaPTSE estimator by reducing correlations between factors not related to geographic proximity or industry in the sample covariance matrix. Those changes have been incorporated into the latest versions of its analytics products, he said.
