Beyond The Stars: Improving Active Fund Selection Based On Manager Skill

Others believe that conviction is an indicator of future performance, described alternatively as active share, manager belief, best ideas and concentration. The proposition is this: truly active managers make bets (i.e., position weights differ from benchmark weights); evaluating the performance of those bets separately from the portfolio uncovers skill; those with the greatest skill and largest bets are likely to outperform.

Supporting this general assertion, Yale professors Martijn Cremers and Antti Petajisto state “Among the highest Active Share quintile [top conviction funds], there is significant persistence in benchmark-adjusted fund performance even after controlling for momentum.”  Skill measured in this way may lead to persistent performance but not necessarily to superior alpha.

Discussing why managers with high conviction may not provide top returns, Randy Cohen of Harvard University finds “powerful evidence that the typical mutual fund managers can, indeed, pick stocks. The poor overall performance of mutual fund managers in the past is not due to a lack of stock-picking ability,” he argues in his paper, “but rather to institutional factors that encourage them to over-diversify, i.e., pick more stocks than their best alpha-generating ideas.” Skill thus viewed through the lens of conviction may identify managers capable of making effective decisions, without the desired benefit of pointing to funds likely to be tomorrow’s outperformers.

Advisors are, nonetheless, in the business of helping clients make fund allocation decisions. Many turn to ratings services for help.

Stargazing

Let’s say you’ve narrowed the search down to two competing funds: both are large-cap, with a tilt toward value and exactly the same returns for each of the past one-, three- and five-year periods. Which fund do you pick? You could split your allocation between them or toss a coin; historical returns alone provide no ability to differentiate the qualities of these funds.

Alternatively, you might turn to Morningstar’s ratings system, which augments return data with their notion of risk: reflecting the average annual amount each fund underperformed its benchmark in recent years. If one of the funds experienced significantly more downside volatility it receives a lower star rating. Knowing one fund is a five star and the other is a four star certainly can break the tie. But will it guide you to the fund that is more likely to do better going forward?

Although the debate over the predictive power of the star ratings continues, the mounting evidence is not encouraging. A recent study by Robert Huebscher concluded that future relative performance cannot be ascertained from Morningstar ratings, especially during periods of volatility. “ Morningstar’s ratings lost virtually all of their predictive ability when measured over a full market cycle,” he wrote. Others analyzing fund ratings have drawn similar conclusions.

New insights for better choices

In addition to the information above, let’s say you suddenly learn the following: Fund A derived all of its excess return from six stocks that have been in the portfolio for many years; it purchased no winners in the last thirty-six months. Fund B, in contrast, has been picking up slivers of excess return from two-thirds of its new buys, year after year. The outperformance of fund A is concentrated in a few old names, while that of fund B is broad based and seems to be repeatable. In which fund are you inclined to invest your client’s capital now? 
 
Most professionals see greater value in Fund B. Understanding skill, even modest insights like these add a whole new dimension to fund comparisons. It provides a greater level of transparency with which you can begin to see patterns of decisions that account for the results. You learn how performance is constructed, which skills are sharpest and how persistently they result in excess return. Combining insights about manager skill with information describing returns, alpha, conviction and ratings, advances your understanding of the products available and what you can expect from them. You are more confident in your recommendations.  

New slant on skill

Skill is the measure of how well you perform the tasks that comprise a bigger event.

In golf, for instance, skill is evident in how well you hit drives, fairway shots and putts. Skill in portfolio management concerns effectiveness in three basic investment decisions – buying, selling and position sizing. The score after a round of golf describes how a particular game went, not the skill of the player. Similarly, historical return only tells you how well the fund did over a time period, but it says nothing directly about the manager’s skills.

Previous attempts to evaluate skill have looked at the performance of high-conviction positions. This is a little like counting only those swings where the golfer appeared to be really trying. The results might be interesting, but they are not likely to identify tomorrow’s PGA winner.

A new framework for considering manager skill now exists that is based on analyzing decisions rather than holdings. It begins by identifying every buy, sell and sizing decision within a fund’s history. Each group is then further analyzed to measure both level of skill and its consistency.

Buying. It’s all about name selection. Skilled buyers choose names that outperform – more often than not. Their picks tend to lift up portfolio performance. Skilled buyers usually demonstrate a consistency between attributes of stocks they buy (e.g. price/value, earnings growth, momentum, etc.) and their stated strategy and process.

Selling. Skill here involves capturing the alpha from great buys and limiting losses. Selling winners involves pushing them out as they cease outperforming, not afterwards. Tired stocks dampen performance as precious capital is tied up in positions unable to continue lifting returns. Effectively managing losers requires correctly assessing the likelihood that underperforming stocks will either bounce back in a reasonable time or continue to drag down performance and it requires acting accordingly. Eliminating winners and losers successfully enables their capital to be redeployed into names with the greatest likelihood of future outperformance.

Sizing.  It’s evaluated separately from buying or selling, and reflects ability in managing to a thesis. Honing this skill requires knowing if your typical winner tends to be a slow starter or perform fast out of the gate. Having the right amount of capital invested in each position can turn good buys into great positions.

Developed by Cabot Research, this framework is being used by a rapidly growing number of equity managers to help them hone their skills and calibrate investment processes. This same set of analyses can support those interested in sharpening their ability to select funds for their clients. Advisors would not only know more about the skills of the managers they picked, but could use this information to develop their fund short-list and to conduct more effective manager interviews. Knowing that a manager is, on average, highly skilled and very consistent in delivering excess return might be that extra insight you need to help clients stay with a great fund that turned in a bad year.

Conclusion

Investors continue to want their capital invested in actively managed funds. The choice of which fund they use is often a decision that their investment advisors makes or at least heavily influences. Yet advisors are at a disadvantage in recommending funds likely to deliver outperformance, since they rely primarily on return and stars with limited application of alpha and conviction. The predictive ability of existing metrics for identifying funds likely to outperform tomorrow is, at best, uncertain. Widespread confidence in fund selection, therefore, does not seem possible based on these tools.

An improved method for considering a fund’s ability to deliver future outperformance is within sight, based on a more rigorous understanding of manager skill. This method of analysis has been applied to more than $500 billion of professionally managed assets. It is being used today by many fund managers to better understand their skills and to work toward more consistent and stronger performance. Elements of this framework can readily be adapted to support picking funds. The incremental value is finding products today more likely to deliver desired performance tomorrow.

Michael A. Ervolini is the CEO of Cabot Research LLC, a Boston-based company that supports equity portfolio managers.

Read more articles by Michael Ervolini