In the investment industry, few debates are waged more intensely than that between “active” and  passive” investing. As investors read media coverage about the futility of trying to pick stocks and the advantages of investing via ETFs instead, more and more are questioning the fees they’re paying for investment advice.
That’s why an award winning research paper by two Yale academics, looking at active managers in a different light is so important. This research, which is the focus of my column in today’s Globe and Mail, identifies a new measure called “active share” – and demonstrates that managers with high active share can in fact beat their index over time.

The roots of the active passive debate
The passive argument really began with Princeton professor Burton Malkiel’s classic 1973 text “A random walk down Wall Street.”  The book argues that the stock market is fundamentally efficient, reflecting everything that is known at a given point in time. As a result, it is futile to try to outperform the stock market – and the best way to invest in stocks is through low-cost index funds that essentially match the market’s overall performance.
In the United States, the growth of index investing was spurred by Vanguard Funds and its founder and Chairman, John Bogle – today Vanguard manages over $1 trillion dollars (almost twice the size of the entire Canadian fund industry) and is among the top three U.S. fund complexes.
Over the past decade, the launch of exchange traded funds has given passive investing a boost, as these offer investors the opportunity to participate in an ever more focused group of indexes.
The majority of academics subscribe to the efficient market theory and support passive investing – but the events of the past couple of years have called into question just how efficient the stock market actually is.

A different definition of active management
In 2006, Martijn Cremers and Antti Petajisto of Yale University published a paper titled “How active is your fund manager? A new measure that predicts performance.”  Winner of best paper at the Finanical Research Association 2006 annual meeting, the authors examined over 2500 U.S. mutual funds from 1980 to 2003, looking in particular at what the authors called their “active share”.
The active share of a fund is the extent to which its holdings don’t overlap with the benchmark index that the fund tracks – a fund with active share of 100% would have no overlap with the index, while an active share of 0% would exactly track the index. You’d expect index funds to have a 0% active share, but what about so-called actively managed funds?

Some of the key conclusions from the authors:

• In 2003, only half of self-described active funds were truly active, with an active share of 80% or higher. This compares to eight out of ten funds with a similar active share in 1980, at the beginning of the study.
• And larger funds tended to have lower active share. When looked at by assets under management, the percentage of assets with active share under 60% grew from under 2% in 1980 to over 40% in 2003. By contrast, the percentage of assets with active share over 80% fell from 58% in 1980 to under 30% in 2003.
• Part of the decline in active share is due to the rise of index funds. Beyond this, however, the authors identified significant growth in what the industry calls “closet indexers” – funds that charge management fees for active management but that closely track their index. 
• There is a direct correlation between true active management and performance. Funds with the highest active share outperformed their index after expenses by 1.5%; the least active funds underperformed by 1.5% - so there is a gap between high active and low active share funds of almost 3% annually.
• Size matters – smaller actively managed funds outperformed large actively managed funds by about 2% per year. 
• Fees were not an indication of active share – closet indexers charged just as much as funds that were truly actively managed.
• High active share doesn’t mean greater volatility – the study showed that funds run by strong stock pickers had similar volatility to their benchmarks, even though their portfolios were quite different and their returns tended to be higher.

 
Extrapolating performance
While attending the annual meeting of the American Economics Association in Atlanta at the beginning of January, I had a chance to chat with Martijn Cremers, one of the study’s co-authors.
Cremers discussed a key finding on the persistence of above average performance – as long as they have a high active share, managers who have done well in the past three years tend to do well in the next three. 
He also talked about the dramatic growth in closet index funds, supposedly active funds that didn’t appear to be active funds at all.
He attributed this to three factors: First, as funds became successful and grew in size, some managers became more risk averse and cautious about making big bets. Second, the greater difficulty of deviating from the index as a fund grows in size. And finally, he suggested that in the U.S. some managers with demonstrated skill had left the mutual fund industry for the greater financial rewards in hedge funds.
Finally, he differentiated between two types of funds with high active share.  The first were run by traditional stock pickers, whose focus was on taking concentrated positions in a relatively small number of stocks, making decisions on the merits of each individual stock.
The second category were run by fund managers who made what the industry calls “factor bets” – essentially making macro judgements to rotate from one industry sector to another or in some cases go to cash entirely.
Cremers said their research showed that as a general rule, funds focused on stock selection had better returns than those that made market timing calls, although as long as they were truly active, both tended to outperform their index after expenses.

Today, Canadians go to great lengths to ensure they get good value on major purchases. This research suggests that when it comes to recommending actively managed mutual funds, advisors need to go beyond just performance numbers to dig deeper into how a manager delivered those numbers – and in particular the extent to which they deviated from their underlying index and had a high “active share” as a result.

To read today’s column in the Globe and Mail on active share, click here: http://www.theglobeandmail.com/globe-investor/investment-ideas/only-the-truly-active-fund-managers-lead-the-pack/article1476655/

To watch the interview with Yale University’s Martijn Cremers, click here: http://clientinsights.ca/video/martijn-cremers-new-research-on-outperformance/type:investor

And to read the award winning study by Cremers and Petajisto, click here: http://rfs.oxfordjournals.org/cgi/reprint/hhp057?ijkey=M0noS3O1M6QvzdG&keytype=ref