I am not enamored of using rolling 3 year returns to assess persistence.
A 3-year time period will often be all up or all down. If a fund manager has an investing personality or philosophy then I would expect strong relative performance in a rising market to be negatively correlated with poor relative performance in a falling market, etc.
It seems to me that the best way to measure persistence is over 1 (or better yet more) market cycles.
So, the question becomes:
Did those funds which outperformed in their style group (preferably on a risk-adjusted basis) continue to do so in the next complete market cycle?
One may formulate this more precisely in one of several ways, such as:
Does alpha over a complete market cycle correlate with alpha over the next cycle?
Does percentile rank within a style group over a market cycle correlate with percentile rank within the group over the next cycle? (Might get a different answer for less efficient market segments, such as small value, etc.)
What proportion of funds in a style group with the top (10%, or 25%) Sharpe ratio over a complete market cycle maintain that position?
I'd love to see more on this if anyone knows of such work.
There was a paper out of the U Mass Amherst research center on securities which looked at the alpha question, and the answer was supportive of persistence. I did a study a few years ago looking at no-load large cap value funds which had existed for at least 8 years and asked if the ranks in style group for the first 7 years could predict the rank in style group for the 8th year. The correlation between the first 7 years' rank (7 predictors) and the 8th was approximately .50, suggesting that there is some persistence.
I will say that buying and selling at the bottom or the top of a "cycle" will provide a positive or negative momentum that impacts the way we ride the wave.
First, it is not well defined. There are many peaks and troughs depending on the time scale you look at - monthly, yearly, decade, etc. This is the same problem TA that looks at patterns like head and shoulder or M or W suffers from. A pattern at one scale looks like an irrelevant squiggle at a larger scale. Seems like it needs to be additionally defined with some minimum duration or minimum swing or something like that to make the measurement of performance meaningful. Perhaps, there is such a definition?
Second, it doesn't remove the rising or falling market bias that rolling year based performance suffers from. Aggressive funds do well in rising markets, conservative funds do well in falling markets relatively speaking to indices. So, a rising market - peak to higher peak with a small trough will make the aggressive fund look good. But that performance is unlikely to be repeated if the next market cycle is a peak to lower peak or has a deeper trough. So, it will look as if there is no persistence based on how the market went and over time, it will look like there is no persistence at all. You can imagine this will be the case, without even having to conduct a study although critics of active management may very well do this as many of them seem to do studies for which they already know the answer will be supporting their thesis. You need some way to normalize for those base slopes and depth of swings to compare for persistence between two market cycles.
Alternatively, you need some kind of a rolling definition of market cycles starting from any point in time, not just peaks or troughs. And then study the average and variations over all such rolling periods and see if there is a persistence signature. Don't know what that might be.
Without coming up with a valid measuring metric, persistence may be a garbage in garbage out result. But it is worth finding a suitable metric.
The folks at Steele Mutual Fund Expert actually include "full cycle" returns in the database we use to compute our ratings metrics.
Here's how they define it:
A full cycle return includes a consecutive bull and bear market return cycle.
Up-Market Return (Bull market)
A Bull market in stocks is defined as a 20% rise in the S&P 500 Index from its previous trough, ending when the index reaches its peak and subsequently declines by 20%.
Down-Market Return (Bear market)
A Bear market in stocks is defined as a 20% decline in the S&P 500 Index from its previous peak, and ends when the index reaches its trough and subsequently rises by 20%.
Looks like there have been 9 such cycles since 1962:
1 Full Cycle 9/08-12/13
2 Full Cycle 4/00-8/08
3 Full Cycle 8/90-3/00
4 Full Cycle 9/87-7/90
5 Full Cycle 12/80-8/87
6 Full Cycle 1/73-11/80
7 Full Cycle 12/68-12/72
8 Full Cycle 2/66-11/68
9 Full Cycle 12/61-1/66
Will dig a bit deeper. (I believe in the recent cycles above, trough is more like 3/09 not 9/08 and I've put inquiry into Steele.)
Working on an.idea, is there a easy (non manual) way to output top performing funds between any two dates?
I am interested in finding out the top 20-30 large cap funds in the following two ranges and how many funds remain included in both.
1. 10/09/2007 and 08/17/2012
2. 3/22/2000 and 10/11/2006
If anyone has pointers to an online tool or have access to such a tool, would appreciate it.
Thanks Numbergal for initiating this thread.
It is good to hear from an investor who is specifically dedicated to financial numbers. The markets are awash with a tsunami of statistical data sets. I am not surprised by your question or interest. It is both probing and challenging. The empirical evidence is that women are better long-term investors than men. Good for you all.
I completely agree with your judgment that an examination of persistent performance demands a period that is longer than 3 years. I ventured down that same road in the mid-1980s with attempts to discover fund managers with superior long-term stock picking skills. It was relatively easy to find managers who outperformed in upside markets, but far more difficult to find those who maintained that advantage to dampen downside risk. In general, I failed.
That very specific information was not readily available in that timeframe. It’s easier these days with outfits like Morningstar recording mutual fund Alphas for extended periods (like 15 years). From the Alphas, an Information Ratio can be computed that will permit some initial sorting. but it is still a daunting task, especially given the level of detail that you seek.
In a qualitative sense, a much longer data accumulation period is needed beyond 3 years. If specific quantification is truly required, a 10-year data period seems like a good departure point.
However, if more precision is required, it is a doable task with some considerable research effort and elbow grease on your part.
I would not invest too much time reviewing deep history. The applicable timescale for current fund managers likely goes back only 20 years or so. There is some good news in that regard. Jeremy Siegel’s post WW II stock data suggests four major trending periods since that date: two upward pricing sub-periods and two more or less neutral return groupings. His latest segment (2000-2013) is representative of a neutral period with significant both upward and downward price movements.
A fund manager’s stock picking acumen can be tested against this recent period. One attractive candidate method would be an initial sort to identify plus Alpha managers. Next, the manager’s downward resistant skills can be challenged by examining his performance in minor to major downward movements. Identify more numerous small downward return periods (like 5 % losses), less numerous 10 % negative periods, and finally tumultuous and rather rare 15 % or more panics. Examine the manager’s performance in these selected downward directed spirals as a test of his flexibility.
That’s a ton of work. I have been down that same road with little success. Many pitfalls and traps exist. Manager’s change funds within an organization, and also change firms. The research staff that supports any manager influences his buy/sell decisions, and that is in constant flux. Even successful football quarterbacks of the Payton Manning and Tom Brady caliber depend greatly on their supporting teammates, staff and organization’s policies.
Past performance is surely no guarantee of future success. In fact, since the strongest pull in the investment world is a reversion-to-the-mean, the likelihood of a reverse outcome is more probable.
The Index Fund Advisors (IFA) have some excellent review articles on this matter. Here is a Link to their Section 5 which documents some of the studies on this topic from guys of David Swensen’s quality:
Of course, IFA is not an unbiased organization. But the assembled articles summarize some serious academic research. Be sure to read “The Fired Beat the Hired” and the “Pension-Gate” reports.
Even with almost limitless resources, institutional agencies can not often identify managers who will outdistance Index benchmarks with similar risk characteristics. The record is dismal for active managers even under professional selection resources that an individual investor can never hope to equal or emulate.
As noted in the referenced work and credited to Bob Dylan: "the first ones now will later be last, for the times they are a changin." Persistence seems like an absent commodity in the investment universe.
As I mentioned earlier, I have traveled this rocky road. In the 1950s I traded individual stocks. In the 1980s I discovered mutual funds and still own one of my originals. In the 1990s, I became acquainted with John Bogle’s “Common Sense on Mutual Funds”. During this extended period I morphed from a 100 % active investor to a current 50/50 % mix. I am a slow learner. I anticipate that I will end with the conventional 20/80 commitment, heavy in Index products. For me, the evidence is overwhelming and compelling.
I recommend you consider aborting any detailed and time-sink project to identify that rare superior fund manager under all market circumstances. Many have and continue to try, and about the same number have failed. The odds of finding such persistent mangers are extremely low, and the potential excess returns are pitiful, especially when contrasted against the more likely outcome of choosing unwisely and potentially suffering more painful and portfolio damaging below benchmark performance.
I always hesitate to throw cold water on proposed research. We only learn when we explore, even when we fail to satisfy our original objectives. But the pathway that you propose to explore has been well traveled without many persistent management discoveries. All these Gods have feet of clay; superheroes are a myth.
Whatever your choice, I fully respect it, and will forever wish you well.