Practical Applications from the Experts - May 2010

categories: Product Documentation, general

Equity Performance Attribution

Performance attribution refers to various methods that explain why the return of a portfolio differs from its benchmark. In particular, performance attribution focuses on a portfolio’s active return, or alpha, and decomposes it in a way that gives insight into the portfolio’s return characteristics. The alpha is decomposed into the following effects:

• Selection – Refers to the manager’s ability to pick securities within a particular segment.
• Allocation – Refers to the manager’s ability to effectively allocate the portfolio’s assets to various segments.
• Currency – Refers to the manager’s return that stems from currency movements.
• Interaction – Refers to the combined effect of Selection and Allocation. Here, overweighting a good selection results in a positive interaction, while underweighting a good selection results in a negative interaction.

This type of analysis allows asset managers to determine whether or not a portfolio’s alpha is being generated in-line with what the manager claims are his or her strengths. For example, if a fund manager claims to excel in equity analysis, the selection effect should reflect this fact by being positive.

RiskMetrics has built out a Performance Attribution product, PerformanceManager, which uses the Brinson method for arithmetic attribution. The application is flexible in that it supports custom drill down structures and can analyze attribution from different perspectives, such as top-down or bottom-up. Two options are available for computing returns, which are the key inputs into computing attribution effects:

• Holdings-based – Adequate for a buy and hold strategy or a portfolio with very little turnover. Return approximates a true time weighted return. The advantage of this approach is that it is straightforward and requires very little data.
• Transaction-based – Appropriate for more dynamic portfolios and considers the effect of various cash flows, such as redemptions and dividends.

Below is the result of an attribution analysis run on a long only portfolio and broken down by sector. In this vein, we are interested in the manager’s ability to overweight sectors that outperform the benchmark.

Click for larger view

A few points come to mind immediately when looking at this breakdown. Looking at the total, one can see that this fund manager did not allocate his assets well, but was very effective in selecting securities. In particular, the manager has largely over-weighted an under-performing sector in Information Technology. Additionally, the manager under-weighted an over-performing sector in Financials. These two decisions proved to be the most costly in terms of the Allocation effect.

These bad decisions, however, were balanced out by above-average stock selection in the Information Technology sector. In fact, nearly his entire alpha of 3.2% has been generated here. The Currency and Interaction effects were negligible. The same analysis could also have been made at a more granular level, such as industry or position.

Download file