No use for ‘active share’
Hard on the heels of my post pointing out the irony of the FCA’s conclusion of its thematic review, that managers of retail funds need to do more to help investors build realistic expectations of the fund’s relative performance profile, I happened to read new academic research debunking ‘active share’, currently the most popular measure of a manager’s willingness to differentiate its strategy from a benchmark, distance it from a ‘closet index’ approach and inform expectations about relative returns.
The new paper in the latest Financial Analysts’ Journal by Andrea Frazzini, Jacques Friedman and Lukasz Pomorski, Deactivating Active Share, contradicts earlier research papers by Cremers and Petajisto (2009) and Petajisto (2013) but uses the same data for US equity funds as those earlier papers. They find that the original conclusions that higher active share (bigger bets off the index, less closet indexing) is associated with outperformance (by a compelling 2% pa) was explained by the differences in sector performance, i.e. style, not active share. There were large differences in risk-adjusted returns relative to the S&P500 between small and large cap sectors in the period analysed. Controlling for differences in benchmark returns, the authors of the new paper found no statistically-signficant association between active share and relative performance.
Even without following the data-driven arguments closely, this conclusion was always the most likely logically, as long as the universe of mutual fund managers holds ‘the market portfolio’ – i.e. there are no significant differences in the stocks owned by mutual finds compared with all other portfolios. There is good evidence in the US they do hold the market portfolio. Since closet index funds on average earn the index return less fees, the rest must also on average earn index returns less their fees. As the authors say: “Among the high active share investors will be winners and losers, but as a group they cannot systematically outperform the closet indexers.”
While the authors reject the use of active share when it comes to expected relative performance outcomes, they do believe it is useful information when it comes to fees. “In general, fees should be commensurate with the active risk the fund takes”. That too we agree with. But whilst it is clear that a closet index approach could at some level of fees be self-defeating, the idea that the option of risk-adjusted outperformance implicit in selecting active managers is worth more because the bets are bigger, though intuitive, needs some refinement. What is clear is that, for any given willingness on the part of an investor to to add alpha uncertainty to systematic market or beta returns, which you can think of as their tolerance of active-manager risk, a better value deal is available when combining a high-conviction manager with a pure passive fund than by paying a full active fee for a fund that has that same combination embedded in its total portfolio.