When Top Managers Underperform

When assessing underperforming fund managers, it’s only in hindsight that one can judge whether patience was a virtue or a costly mistake. If a manager in the midst of a tough stretch ultimately goes on to turn in stellar long-term performance, investors who stuck around are credited with foresight and discipline. But if the short-term woes turn to long-term underperformance, investors who go down with the ship are judged as complacent, stubborn or simply naïve. In either case, experiencing an ongoing period of underperformance brings to mind writer Ambrose Bierce’s description of patience as “a minor form of despair, disguised as a virtue.” That despair is forgotten if a fund going through a bad patch eventually rewards investors for their patience. But how often is that the case?

Certainly, there is evidence to suggest that even top managers will test their investors’ loyalty. In a famous 1984 speech and subsequent article for Columbia Business School, Warren Buffett presented returns from nine “Superinvestors of Graham-and-Doddsville” who had outperformed the S&P 500 over various long-term periods. A brief perusal of calendar-year returns achieved by these investors shows cumulative long-term outperformance but not consistent annual outperformance. In fact, three of the nine had periods of underperformance that lasted at least three years. It is worth remembering that while a stock holding may initially decline in value, it is the price at which the stock is eventually sold that determines whether or not it was a successful investment. We believe that common traits shared by many great managers include a willingness to own unpopular names, making decisions based on long-term analysis and having the discipline to ignore sometimes painful shorter-term swings. On our own list of admired investors and Masters’ Select sub-advisors, Bill Miller has beaten the S&P 500 in each of the last 15 calendar years , but that 100% rate of outperformance falls to approximately 80% when his returns over the same time period are viewed on a rolling 12-month basis. That is still a great record but it underscores that perfection is an unrealistic expectation.

Recently we decided it was worth looking at some historical data to re-assess the value of patience. This look has some timely relevance for Masters’ Select since our funds and several of our sub-advisors have been lagging their benchmarks over the shorter term. Our study was based on the analysis of equity mutual fund returns and sought to answer a couple of questions.

• First, we were interested in determining the extent to which funds that outperform over the long-term (10 years in our study) experience shorter-term periods of underperformance within the same time span. While we don’t generally worry about how long a manager or fund may lag a benchmark, we were curious to discover how often this underperformance stretched over at least three years—a period that would sorely test the patience of both the manager and the fund’s investors.
• Second, when these long-term outperformers did experience three-year periods of underperformance, we wanted to measure the magnitude of the underperformance. Certainly sticking with a fund (or manager) that slightly trails its index over three years is easier than sticking with a fund or manager that trails its benchmark by a wide margin.

By starting with a group of funds that outperformed their benchmark over the long term, we could go back and test how low managers could go (in terms of length and magnitude of underperformance) and yet still reward investors with superior long-term results. We could also determine how common such periods of underperformance are among managers with top long-term records. Finally, our research team as a whole discussed the results of the study to look for insights and data that might assist us in future manager due-diligence efforts.

We began by screening Morningstar’s Principia database for actively managed funds across six domestic equity categories (large- and small-cap value, blend and growth). After eliminating index funds and redundant share classes of the same fund, we looked for funds with a 10-year track record through December 31, 2005. This produced six groups ranging from 33 funds in the small-cap value category to over 200 funds in the large-cap blend category. We then identified the top performers in each group, defined as those funds that outperformed their relevant benchmark by at least one percentage point annualized over our 10-year time period. Six groups of 12 to 90 funds made up our final “overachievers” list.

Next, we examined the percentage of funds in each group that underperformed the benchmark by at least two, five, 10 or 15 percentage points annualized during any rolling three-year period. The results are shown in the graph and summarized in the table.

The data show that among funds with the best long-term records, most experienced a prolonged period of underperformance relative to their benchmark. In fact, every single value fund (and almost all of the funds in the remaining four categories) underperformed the benchmark by two percentage points or more annualized during at least one rolling three-year period. If nothing else, we find this useful for setting expectations. We believe that applying a superior investment approach consistently should lead to long-term outperformance but it simply won’t work for all the people all the time. In other words, the results of our study tell us that on the road to long-term outperformance, not only should we expect underperformance, we can expect it to last years. (It is worth noting that a fund that underperforms for three years typically does not consistently underperform for the entire three-year period. More typical is a shorter period of underperformance that is not offset by prior or subsequent outperformance during the three years.)

As mentioned above, we also wanted to understand the magnitude of underperformance that might occur among our elite group of outperformers. Most funds underperformed by two percentage points annualized over some three-year period but how many lagged by at least five, 10 or 15 percentage points annualized during a three-year time span? As shown in the tables below, the percentage of funds underperforming by at least five percentage points annualized over a three-year period was also high, ranging from 50% to 72%. Only when we looked for annual underperformance of at least 10 percentage points over three years did the number of funds drop into the minority (but just barely for the small-cap blend category, with 49% of all funds showing up here.) Four of our six categories also had at least some funds (ranging from 7% to 16% of the total outperformers) that also experienced annual underperformance of at least 15 percentage points over a three-year time period and amazingly still managed to beat their benchmark by at least 1% annualized over the full 10-year time period.

Certainly, on the question of magnitude, a three-year period with annual underperformance of at least five percentage points was common among the long-term outperformers in all six of the domestic equity categories we studied. And while the value categories appear to have produced outperforming funds with a narrower range of underperformance, they are also the only categories with 100% of all funds lagging the benchmark by two percentage points or more annualized over at least one three-year period. In addition to patience, top-performing funds appear to require tolerance of sometimes significant degrees of underperformance (though this would likely be a function of the types of market environments experienced).

We started our study by screening for funds that had outperformed over a 10-year time frame, so in hindsight, of course, we know that patience was rewarded. The fact is, however, there’s no guaranteed pay-off for patience and it’s easy for fund families to argue that “timing” investments doesn’t work and the best strategy is simply to “sit tight.” In a speech earlier this year, Bill Nygren a Masters’ Select Value sub-advisor (and no stranger himself to underperformance of late) highlighted this problem, noting that when investors question performance, the “answer doesn’t seem to change based on how long we’ve underperformed. If it’s a quarter, investors are told that they need to look at years, not months. If it’s a year, we tell them we need to be measured over a market cycle. If it’s three or five years, we talk about really needing a decade to show a statistically meaningful performance record. I think investors deserve advice beyond just being more patient.”

We agree. The most important question in our minds isn’t how long investors should be patient. That, we believe, is largely unpredictable and the data we have just summarized doesn’t change that. The key is to know when patience is warranted. Our due-diligence process demands ongoing monitoring of all our sub-advisors, regardless of whether they are beating or lagging the benchmark. In either case, one goal is to be sure we understand what is driving performance. We spend a large amount of time getting to know a manager’s investment approach and defining the “edge” that we believe will enable them to outperform over the long term. The managers and their supporting analysts have to make their process transparent for us to do this. The advantage (both for them and us) is that we should be able to distinguish between times when the team’s long-term “edge” remains intact while performance is temporarily suffering and times when that “edge” has disappeared due to inconsistency or a lack of discipline or various other reasons.

In the case of a market environment that doesn’t favor a manager’s edge or even when they have slumped due to mistakes that we don’t believe are indicative of a loss of edge, we believe patience is warranted. We know of a number of managers who have maintained their disciplined approach and suffered for it but were eventually rewarded. Included in this group are a number of our sub-advisors. We wrote about some of the prior performance stumbles in this year’s semi-annual report. Examples like those lend support to our belief that managers who maintain a disciplined approach through periods of underperformance ultimately position their portfolios for later outperformance. It also underscores that long-term outperformers that have extended periods where they lag, which, based on our study, is the majority of outperforming funds, must, by definition, have very significant outperformance during the rest of the period. To be successful, investors need to be there for that outperformance. This suggests that investors should be clear on their reasons for dumping a poor performer and those reasons should go beyond the performance. Otherwise they risk missing out on the strongly positive periods that can lead to long-term outperformance. Using this same line of thinking to address our responsibility to hire and fire managers, our success ultimately rests on our ability to make assessments that are based on much more than recent performance. Rather than just reacting to underperformance, we must understand why a manager underperformed and we must determine whether or not their competitive edge is intact. We believe that acting based on our assessment of their competitive edge is much more likely to result in long-term success, and that reacting to short-term performance is highly likely to hurt our ability to outperform in the long run.

¹Morningstar, Inc. is an independent mutual fund research and rating service. Each Morningstar category represents a universe of funds with similar investment objectives. Fund total returns include dividends and distributions reinvested and do not reflect sales charges.

Mutual fund investing involves risk. Principal loss is possible. Please refer to the prospectus for special risks associated with investing in the Masters’ Select Funds, including, but not limited to, risks involved with non-diversification and investments in foreign and debt securities and smaller companies.