By Stanislaw Zarzycki with Ken Marshall
We have all been there: Holding on to the active manager that keeps underperforming his/her benchmarks or, worst, just keeps losing money. We may justify holding on to the actively managed mutual fund by telling ourselves " he/she has to bounce back, just based on the law of averages…" Or perhaps we rationalize buying the underperforming managers more aggressively, thinking that we will benefit from dollar-cost averaging.
When should an investor dump the underperforming managers? Are there clear fact-based guidelines that the investor can take advantage of? To find answers, we decided to look at what the numbers tell us. The conclusions may serve to guide investors' decisions not only with the laggards but when looking at all equity-based mutual fund managers.
Following on from our previous study of active managers vs. index funds, we looked at the performance of underperforming managers. We first looked at the in-sample period 1999-2005 of 160 funds. Over that initial period, 65% or 103 of managers underperformed the index. We measured underperformance by comparing the compounded returns over these seven years with the compounded return of the S&P 500. The underperformance ranged from -0.4% to -14%. We then ranked these managers into ten groups (1 being the worst underperformers, 10 being the least or "best" underperformers based on their initial compounded return for 1999-2005) in order to compare and rank their performances against the S&P going forward.
How did these under-achievers perform over the following seven year period? How many of them bounced back and rewarded their loyal investor? If in fact there is a mean reversion effect, we would expect to see better performance in the ensuing period. Alternatively, if poor performance is a kind of "destiny", we would expect to see the same poor skill reflected in the same lousy performance.
We analyzed the average compounded performance for each decile of the manager group over the out-of-sample period (2006-2012). The findings were astounding:
The managers that most (1st decile) underperformed the index in the initial period (1999-2005) continued their underperformance in the out of sample period. Almost 100% of the managers in the original first three deciles (1st-3rd) continued to significantly underperform going forward. Even more astounding was the autocorrelation of underperformance and the fact that the level of underperformance was almost linear. Clearly underperformance may be due to fee drag, rotation costs, expense ratios or friction, but it may also be attributed to mediocre skill.
The data we used indicate that, contrary to popular rationale, the investor typically is not rewarded for holding on to a poorly performing fund. In fact, they appear to be punished by enduring lower returns than what they would earn in a passive ETF (like SPY). Investors may avoid the whole dilemma before allocating to a new actively managed mutual fund by using this simple metric: compare the compounded return of the target manager to the compounded return of the S&P over the past 3-7 years. If in the past the manager has underperformed the index, the overwhelming probability is that the manager will continue to underperform. Elsewhere ("ETFs vs. Active Managers") we have shown that it is extremely unlikely that one can pick an active manager who will outperform the index. Obversely, a manager who has historically underperformed is expected to continue to underperform. The investor is likely to be better off dumping his underperformers. To get index-like returns it is far easier allocating to the benchmark index (eg. SPY).