The Truth About Performance Records by Andrew Hunt, author, Better Value Investing: A Simple Guide to Improving Your Results as a Value Investor, full bio below
Global value investing has been having a yough time of late. Indeed, the FT[i] and the Brandes Institute[ii] have reported that, since the financial crisis, value investing has had its weakest streak of relative performance ever, even compared to the early 70s or the TMT bubble.
Given this background, it seems a good time for many value investors to consider what historic performance records can actually tell us. Reading through the extensive literature on performance records, a number of salient and important points emerge.
Great investment records are very rare
While different studies quote slightly different numbers, less than five percent of funds have beaten the market by 1% or more over 20 years, and that falls to less than one percent by the time you extend to 30 years.
It is also worth emphasizing what counts as great long-term performance. The best mutual fund since 1970 (Franklin Templeton Mutual Share) has managed c.3% annual outperformance. Only three funds (out of a starting sample of 350) made over 2%.[iii] Given the enormous resources and intelligence behind most strategies, that shows how difficult it is to outperform over the long haul. One or two percent long run outperformance should be viewed as amazing.
Performance is not even ? in fact, it invariably includes some very long, very bad stretches.
Secondly, those stunning performance records are not achieved by a steady course of beating the market year after year. Indeed nothing could be further from the truth: even the greatest investors underperform about 40% of the time. But what is really startling is quite how uneven those performance records are; and how prolonged the tough times can be.
In a long running series of papers by The Brandes Institute, titled ?Death, Taxes and Short Term Underperformance?,[iv] the researchers looked at the best performing funds over ten years in various asset classes (Global, US, EM, International Equities and Credit). They found that all the top performers dropped into the bottom deciles over rolling one and three year periods during those ten years. In many instances these included periods of falling over 30 or 40% behind!
It is also quite eye-opening to look up a few of your investment heroes and dig into their records. To give five examples of investors I admire:
David Einhorn ? the well-known hedge fund manager at Greenlight Capital ? has beaten the market by 7% p.a. since 2000. However this includes a streak of 27% p.a. average outperformance from 2000 to 2006, and consecutive losing years averaging -4% p.a. since 2007. Same investor, same approach, very different results.
Anthony Bolton ? the UK?s most successful and best known fund manager, who beat the market by 6% p.a. over 25 years ? fell 30% behind over two years during the middle of his record-breaking streak.
Rick Guerin?s incredible run ? which was the best of all Buffett?s superinvestors, outperforming by 23% p.a. over nineteen years ? included a 6-year streak of consecutive underperformance in which he fell behind by a cumulative 70%.
Walter Schloss ? 8% outperformance over fifty years ? fell behind for a whole decade from 1989 to 1999.
Bill Ruane ? founder of the Sequoia fund and another of the top superinvestors ? fell 40% behind over his first three years.
These are truly horrendous results over prolonged periods. And yet we know, ex post, that these are some of the greatest investors ever!
In probably the best paper on the subject,[v] Eugene Shahan examined the stunning records of the Superinvestors from Buffett?s famous paper, ?The Superinvestors of Graham-and-Doddsville.? Shahan emphasises that their key advantage was independent mindedness:
?Independent decision-making is the common factor?the ultimate decision making must be the result of disciplined individual judgement.?
In other words, the superinvestors? performances diverged so wildly from benchmarks because they were thinking and acting independently. This was what made them great long term investors. Very tough periods are the inevitable consequence of genuinely active and genuinely independent investing over many years.
Performance Records can change very rapidly
A natural consequence of volatility in performance is that performance records can change very quickly.
Take this graph from one of the aforementioned Brandes studies on top-performing US equity funds. Note how 3-year rolling performance frequently swings from top to bottom or back again in a couple of quarters:
Performance Records
It goes to show that an investor?s record is never really locked in. You can never just sit back. If a couple of strong quarters drop off the tail and are replaced by a couple of weak quarters it can completely change round. The converse is also true: bottom decile can become top decile just as quickly.
Past performance as an indicator of future performance
Inevitably, there has been much ink spilt on this. And while there is an abundance of research, the findings have been remarkably consistent[vi]:
There is a momentum effect. Funds with strong recent (1 year) performance tend to carry on winning over the short term (1 year).
However, over three to five years, fund performance tends to mean revert. Such that prior winners are more likely to trail losers and prior losers more likely to outperform.
There are various explanations for this. The most obvious is style cycles and fund flows. When certain managers are successful, they see greater inflows which drive up their securities to overvalued levels ? hence momentum and then mean reversion. Out of favour managers and portfolios undergo the opposite. The damaging effects of hubris among top performers (in contrast to the self-help and humility among laggards) may also have something to do with it.
Inevitably, this is probably one of the most abused pieces of data in existence. Managers with weak performance inevitably wheel out the mean reversion data. When the same funds enjoy good performance, they swear blind that it doesn?t apply to them!
What about longer records? One would expect that longer records might provide a truer picture. There is surprisingly little data on this, and one has to be very careful to avoid data-mining or survivorship bias. However, what research there is reaches a surprising conclusion: there is absolutely no relationship at all between a lengthy historic record of good or bad performance and subsequent performance! For example, a recent study from the Financial Analysts Journal[vii] looked at ten year performance records among mutual fund managers. It found that, In terms of predicting future success, a ten year record of outperformance is meaningless. However, what it does tell you is that those with good ten year records were more likely to hang on to their jobs (but no more likely than anyone else to outperform subsequently)!
Overall, it is hard to argue that historic performance provides anything helpful, in spite of the emphasis investors, consultants, the media and managers themselves put on it. This all seems rather difficult, even frustrating. In almost every other sphere (think sports, exams or consumer products), past records give a Global value investing has been having a yough time of late. Indeed, the FT[i] and the Brandes Institute[ii] have reported that, since the financial crisis, value investing has had its weakest streak of relative performance ever, even compared to the early 70s or the TMT bubble.
Given this background, it seems a good time for many value investors to consider what historic performance records can actually tell us. Reading through the extensive literature on performance records, a number of salient and important points emerge.
Great investment records are very rare
While different studies quote slightly different numbers, less than five percent of funds have beaten the market by 1% or more over 20 years, and that falls to less than one percent by the time you extend to 30 years.
It is also worth emphasizing what counts as great long-term performance. The best mutual fund since 1970 (Franklin Templeton Mutual Share) has managed c.3% annual outperformance. Only three funds (out of a starting sample of 350) made over 2%.[iii] Given the enormous resources and intelligence behind most strategies, that shows how difficult it is to outperform over the long haul. One or two percent long run outperformance should be viewed as amazing.
Performance is not even ? in fact, it invariably includes some very long, very bad stretches.
Secondly, those stunning performance records are not achieved by a steady course of beating the market year after year. Indeed nothing could be further from the truth: even the greatest investors underperform about 40% of the time. But what is really startling is quite how uneven those performance records are; and how prolonged the tough times can be.
In a long running series of papers by The Brandes Institute, titled ?Death, Taxes and Short Term Underperformance?,[iv] the researchers looked at the best performing funds over ten years in various asset classes (Global, US, EM, International Equities and Credit). They found that all the top performers dropped into the bottom deciles over rolling one and three year periods during those ten years. In many instances these included periods of falling over 30 or 40% behind!
It is also quite eye-opening to look up a few of your investment heroes and dig into their records. To give five examples of investors I admire:
David Einhorn ? the well-known hedge fund manager at Greenlight Capital ? has beaten the market by 7% p.a. since 2000. However this includes a streak of 27% p.a. average outperformance from 2000 to 2006, and consecutive losing years averaging -4% p.a. since 2007. Same investor, same approach, very different results.
Anthony Bolton ? the UK?s most successful and best known fund manager, who beat the market by 6% p.a. over 25 years ? fell 30% behind over two years during the middle of his record-breaking streak.
Rick Guerin?s incredible run ? which was the best of all Buffett?s superinvestors, outperforming by 23% p.a. over nineteen years ? included a 6-year streak of consecutive underperformance in which he fell behind by a cumulative 70%.
Walter Schloss ? 8% outperformance over fifty years ? fell behind for a whole decade from 1989 to 1999.
Bill Ruane ? founder of the Sequoia fund and another of the top superinvestors ? fell 40% behind over his first three years.
These are truly horrendous results over prolonged periods. And yet we know, ex post, that these are some of the greatest investors ever!
In probably the best paper on the subject,[v] Eugene Shahan examined the stunning records of the Superinvestors from Buffett?s famous paper, ?The Superinvestors of Graham-and-Doddsville.? Shahan emphasises that their key advantage was independent mindedness:
?Independent decision-making is the common factor?the ultimate decision making must be the result of disciplined individual judgement.?
In other words, the superinvestors? performances diverged so wildly from benchmarks because they were thinking and acting independently. This was what made them great long term investors. Very tough periods are the inevitable consequence of genuinely active and genuinely independent investing over many years.
Performance Records can change very rapidly
A natural consequence of volatility in performance is that performance records can change very quickly.
Take this graph from one of the aforementioned Brandes studies on top-performing US equity funds. Note how 3-year rolling performance frequently swings from top to bottom or back again in a couple of quarters:
Performance Records
It goes to show that an investor?s record is never really locked in. You can never just sit back. If a couple of strong quarters drop off the tail and are replaced by a couple of weak quarters it can completely change round. The converse is also true: bottom decile can become top decile just as quickly.
Past performance as an indicator of future performance
Inevitably, there has been much ink spilt on this. And while there is an abundance of research, the findings have been remarkably consistent[vi]:
There is a momentum effect. Funds with strong recent (1 year) performance tend to carry on winning over the short term (1 year).
However, over three to five years, fund performance tends to mean revert. Such that prior winners are more likely to trail losers and prior losers more likely to outperform.
There are various explanations for this. The most obvious is style cycles and fund flows. When certain managers are successful, they see greater inflows which drive up their securities to overvalued levels ? hence momentum and then mean reversion. Out of favour managers and portfolios undergo the opposite. The damaging effects of hubris among top performers (in contrast to the self-help and humility among laggards) may also have something to do with it.
Inevitably, this is probably one of the most abused pieces of data in existence. Managers with weak performance inevitably wheel out the mean reversion data. When the same funds enjoy good performance, they swear blind that it doesn?t apply to them!
What about longer records? One would expect that longer records might provide a truer picture. There is surprisingly little data on this, and one has to be very careful to avoid data-mining or survivorship bias. However, what research there is reaches a surprising conclusion: there is absolutely no relationship at all between a lengthy historic record of good or bad performance and subsequent performance! For example, a recent study from the Financial Analysts Journal[vii] looked at ten year performance records among mutual fund managers. It found that, In terms of predicting future success, a ten year record of outperformance is meaningless. However, what it does tell you is that those with good ten year records were more likely to hang on to their jobs (but no more likely than anyone else to outperform subsequently)!
Overall, it is hard to argue that historic performance provides anything helpful, in spite of the emphasis investors, consultants, the media and managers themselves put on it. This all seems rather difficult, even frustrating. In almost every other sphere (think sports, exams or consumer products), past records give a good indication of the sort of performance to expect. Just not here.
Conclusion
Great performance is very rare. And great records inevitably include prolonged periods of poor performance. Volatile results are the inevitable consequence of running a truly differentiated and active portfolio.
Further, past performance ? even over periods as long as five to ten years ? gives no indication of future performance.
For investors who have struggled, this should provide some encouragement to persevere. Conversely, it should also serve as a humbling reminder to those with strong records. In this game, you can never really say you?ve won.
The upshot is that investors can never sit back. They must keep questioning their holdings, seeking out challenge and dissent, learning from mistakes and making incremental improvements to their investment processes.
