Active Management Really Works
June 23, 2009
The study confirms the popular belief that small funds are more actively managed than large funds. They also found that a significant percentage of large funds were closet indexers after their size increased over $1 billion in assets. In fact, the active share of “active” all-equity mutual funds in the U.S. ranges from 30% to 100%, with an average of 66% for large-cap funds. This means that the average large-cap fund essentially indexes one third of its assets, while the worst closet indexers index two thirds of their assets. The study found that there has been a significant shift from active to passive management in the 1990’s. Prior to the 1990’s most mutual funds were truly active, while in recent years the number of actively managed funds has dropped to 20% - 30% of all funds. This is partly due to index funds, but “an even larger part is due to closet indexers and a general tendency of funds to mimic the holding of benchmark indexes more closely,” say the study’s authors. They also found that half of the active positions at the fund level cancel out within the mutual fund sector, so that the entire mutual fund industry is less actively managed.
But here’s the truly earth-shattering conclusion from this study. They find that not only does Active Share predict fund returns, but that funds with the highest Active Shares significantly outperform their benchmarks both before and after expenses, while funds with the lowest Active Share underperform after expenses. On average, funds with the highest Active Share exhibit some skill and pick portfolios which outperform their benchmarks by 2.00 – 2.71% per year. After fees and transaction costs, this outperformance decreases to 1.49 – 1.59% per year. However, the highest performing group of funds was the group with the highest Active Share, smallest assets, and the best prior one-year performance. This group outperformed their benchmarks by 6% per year, even after deducting fees and transaction costs. In addition, the study finds that active management, as measured by Active Share, is persistent, where tracking error is not. In other words, ranking funds by their relative performance versus a benchmark alone is not likely to predict winning funds in the future, but adding active management to the analysis does.
While Cremers and Petajisto excluded pure index funds from their study, they did find that the funds with the lowest Active Share (closet indexers) underperformed. These funds actually had lower returns than actual index funds because of their higher fees. Importantly, however, funds that were the most actively managed showed persistently higher returns, even after considering fees and expenses. In fact, they found no correlation between fund returns and fees and expenses when studying Active Share. This is an important discovery since one of the biggest objections to active management is that the added transactions involved create additional fees that are presumed to be an insurmountable headwind to beating index performance. The 6% alpha or outperformance of the best performing group, consisting of small funds that had the best performance in the prior year and the highest active share, after fees and expenses, is staggering. It gives credence to those investors who have said that while fund managers on average may not be able to outperform, the particular mutual fund manager that they own in their portfolio is simply superior to the average.
It appears that the reason that active managers couldn’t outperform passive benchmarks is because they weren’t really actively managing the portfolios in the first place. With the publication of Fama and French’s Three Factor Model and the subsequent overwhelming concern about manager style, it is now clear that 70% of funds became either indexed or became closet indexers in their zeal to stay within their Morningstar Style Box. In such a world, the passive funds with the lowest fees would relatively outperform. It is only recently, using the methodology of Cremers and Petajisto’s 2007 Yale study that we can now focus on the 30% of funds that are truly actively managed. The results show that active management does, in fact, add significant value.
Enlightened readers will consider that the question of whether active or passive management should be used to invest the individual asset classes that make up a diversified portfolio, where each active manager is compared to a single benchmark or market portfolio, has little to do with the questions of whether or not asset classes themselves are always efficiently priced on an absolute or relative basis, and whether or not it is prudent to rely on past performance rather than judgment and experience to forecast asset class returns. However, if answering that sticky question about why active fund managers can’t outperform their style-specific benchmarks is a critical issue for those who are considering tactical asset allocation, they can now rest a little easier.
Moving on
Given the tremendous global palette of asset classes now available to today’s investors, it is pure nonsense to assume that all asset classes are fairly valued all of the time, either absolutely, based on their past metrics for value, or based on their relative value to other asset classes. Now that Cremers and Petajisto have shown us that active managers can outperform their benchmarks, and Stanford professor Mordecai Kurz and H. “Woody” Brock, founder of Strategic Economic Decisions, have given us the academic and theoretical reasons why we should spend time and money to attempt to earn excess returns, it is time to begin the practical discussion of how we can find value at the asset class level. It is one thing to say that we should be active, tactical asset allocators, but it is entirely another thing to understand how to best execute such a strategy. Part Two of this book takes on the practical challenges of becoming a successful tactical asset allocator in the real world.
For more information about active share, read our earlier article on this topic.
Display article as PDF for printing.
Would you like to send this article to a friend?
Remember, if you have a question or comment, send it to .
