Attribution Analysis

Attribution analysis is a sophisticated method for evaluating the performance of a portfolio or fund manager. The method focuses on three factors: the manager’s investment style, their specific stock picks and the market timing of those decisions. It attempts to provide a quantitative analysis of the aspects of a fund manager’s investment selections and philosophy that lead to that fund’s performance.

Attribution analysis begins by identifying the asset class in which a fund manager chooses to invest. This will provide a general benchmark for comparison of performance. An asset class generally describes the type of securities that a manager chooses and the marketplace in which they originate. European fixed income or U.S. technology equities could both be examples.

The next step in attribution analysis is to determine the manager’s investment style. Like the class identification discussed above, a style will provide a benchmark against which to gauge the manager’s performance. The first method of style analysis concentrates on the manager’s holdings. Are they large-cap or small-cap? Value or growth? Bill Sharpe introduced a second type of style analysis 1988. Returns-based style analysis (RBSA) charts a fund’s returns and seeks an index with a comparable performance history. Sharpe refined this method with a technique that he called quadratic optimization, which allowed him to assign a blend of indices that correlated most closely to a manager’s returns.

Once an attribution analyst identifies that blend, they can formulate a customized benchmark of returns against which they can evaluate the manager’s performance. Such an analysis should shine a light on the excess returns, or alpha, that the manager enjoys over those benchmarks.The next step in attribution analysis attempts to explain that alpha. Is it due to the manager’s stock picks, selection of sectors, or market timing? To determine the alpha generated by their stock picks, an analyst must identify and subtract the portion of the alpha attributable to sector and timing. Again, this can be done by developing customize benchmarks based on the manager’s selected blend of sectors and the timing of their trades. If the alpha of the fund is 13 percent, it is possible to assign a certain slice of that 13 percent to sector selection and timing of entry and exit from those sectors. The remainder will be stock selection alpha.

Academic research on the significance of market timing in manager evaluation has led to a wide range of conclusions on the importance of timing. In general, most analysis of the subject agrees that stock selection and investment style result in a greater share of a manager’s performance than timing. Some scholars point out that a significant portion of a manager’s performance with respect to timing is random, or luck. To the extent that market timing can be measured, scholars point out the importance of gauging a manager’s returns against benchmarks reflective of upturns and downturns. Ideally, the fund will go up in bullish times, and will decline less than the market in bearish periods. Because this is the most difficult part of attribute analysis to put into quantitative terms, most analysts attribute less significance to market timing than stock selection and investment style.