“Active share” is the percentage of a portfolio that differs from a benchmark index. This measurement is often used to determine the degree of active management. Unfortunately, there have been a few poorly thought out research studies suggesting that active share is predictive(positively correlated) with excess return. As usually the media quickly picked those up, misleading many investors to believe such a relationship exists. Let’s explore why this is not the case.
The formula for active share:
Wf = Weight of the security in the portfolio
Wi = Weight of the security in the index
Very simple formula: take the weight of each security in the portfolio and subtract from it the weight of the same security in the benchmark. Add those up and divide by two. Right away you can see where the active share percentage comes from. There is nothing magical about it.
Sources of portfolio active share are:
- Including stocks that are not in the benchmark
- Excluding stocks that are in the benchmark
- Holding benchmark stocks in different weights than the benchmark
So if the claims for excess return were accurate, all one needs to do is hold securities outside the benchmark. If I want to outperform the US equity market, all I would have to do is hold non-US securities (for example hold Japanese equities). Holding stocks outside the benchmark will instantly give me an active share of 100%. Do you guys see how ridiculous the notion of correlation with excess returns is?
A high active share does not predict outperformance. It just shows how different a portfolio is compared to the measurement benchmark. It many cases a high active share indicates that the portfolio is not benchmarked correctly. It may reveal out-of-mandate holdings, style drift, and/or a size bias, any of which may expose investors to unexpected risk. Investors should always question the sources of active share in a portfolio in order to evaluate it properly.
Don’t get me wrong, active share could be a useful statistic. Yes, it helps tell if a manager is indeed active but it has to incorporate whether the securities are at all in the benchmark. If I want a more diversified exposure to US equities than an S&P 500 index ETF, I can find some very high active share US Large Cap managers with concentrated positions whose performance will frequently deviate from the index. But then again, this does not mean those portfolios will outperform. Only that my exposure will be different compared to the index.