Document category: Mutual Fund (Unit Trust and OEICs) Persistence
[Return to category list]
[Return to category list]
"False discoveries in mutual fund performance: Measuring luck in estimated alphas"
Laurent Barras, O. Scaillet, Russ Wermers, University of Maryland and Swiss Finance Institute, May 2008
Laurent Barras, O. Scaillet, Russ Wermers, University of Maryland and Swiss Finance Institute, May 2008
This report is quite complex but the initial summary and final conclusion will give readers a good understanding of the research.
Alpha is generally used to measure the skill of a fund manager. Beta is the return an investor could get from the market (effectively an investment free lunch). Therefore, an active manager must achieve sufficient returns through luck or skill to provide higher than market returns AFTER CHARGES to provide real Alpha to clients. The study reviews fund manager returns after charges between 1990 an 2006 and splits the managers in three groups.
The first group is totally unskilled managers and accounts for 24% if all funds. Almost 1 in four managers are unable to pick stocks well enough to even recover their trading costs and expenses. Surprisingly, unskilled funds have a relatively long average life expectancy - 12.7 years. Such funds exist by attracting unsophisticated investors and charging them high fees.
The second group "Zero Alpha" includes all managers who exhibit sufficient stock picking skills to cover their trading costs and other expenses including fees but deliver no benefits to investors. In this case all of the gains from their efforts go to the manager and the client gets only the return they could have received with a simple index tracker fund. This group accounts for 75.4% of all funds.
The third group actually performed well enough to outperform the market after fees but it is impossible to say whether this was a result of skill or simply luck. Unfortunately, this group comprised only 0.6% of all managers.
True skill in managers seems to be negatively correlated to portfolio turnover. Unskilled managers apparently trade frequently to appear skilled which ultimately hurts their performance.
The proportion of skilled managers has diminished rapidly over the past 20 years. In 1990 14.4% of funds were skilled but this had fallen to only 0.6% by 2006. If charges are excluded the number of skilled funds at 2006 rises to 4.5% - still not a great result but a significant improvement which shows the effect of high charges on investors real returns.
There was a significant rise in the number of funds in existence between 1996 and 2005 and the survey found that "Either the growth of the fund industry has coincided with greater levels of efficiency, making stock picking a more difficult and costly endeavour, or the large number of new managers simply have inadequate skills."
Interestingly, 2.4% of funds exhibited true skill over short periods (5 years) showing that, although rare, short term skill does exist. However, there is no evidence of long term manager skill. It is almost impossible for investors to capitalise on the short term skill of these rare funds as the report shows "outperforming funds only exist until investors are able to locate them."
Conclusion - The report confirms our belief that it is simply not worth the risk of trying to select outstanding fund active managers to pick stocks when 99.4% of all managers either under perform, or at best equal, the market over longer terms.
Alpha is generally used to measure the skill of a fund manager. Beta is the return an investor could get from the market (effectively an investment free lunch). Therefore, an active manager must achieve sufficient returns through luck or skill to provide higher than market returns AFTER CHARGES to provide real Alpha to clients. The study reviews fund manager returns after charges between 1990 an 2006 and splits the managers in three groups.
The first group is totally unskilled managers and accounts for 24% if all funds. Almost 1 in four managers are unable to pick stocks well enough to even recover their trading costs and expenses. Surprisingly, unskilled funds have a relatively long average life expectancy - 12.7 years. Such funds exist by attracting unsophisticated investors and charging them high fees.
The second group "Zero Alpha" includes all managers who exhibit sufficient stock picking skills to cover their trading costs and other expenses including fees but deliver no benefits to investors. In this case all of the gains from their efforts go to the manager and the client gets only the return they could have received with a simple index tracker fund. This group accounts for 75.4% of all funds.
The third group actually performed well enough to outperform the market after fees but it is impossible to say whether this was a result of skill or simply luck. Unfortunately, this group comprised only 0.6% of all managers.
True skill in managers seems to be negatively correlated to portfolio turnover. Unskilled managers apparently trade frequently to appear skilled which ultimately hurts their performance.
The proportion of skilled managers has diminished rapidly over the past 20 years. In 1990 14.4% of funds were skilled but this had fallen to only 0.6% by 2006. If charges are excluded the number of skilled funds at 2006 rises to 4.5% - still not a great result but a significant improvement which shows the effect of high charges on investors real returns.
There was a significant rise in the number of funds in existence between 1996 and 2005 and the survey found that "Either the growth of the fund industry has coincided with greater levels of efficiency, making stock picking a more difficult and costly endeavour, or the large number of new managers simply have inadequate skills."
Interestingly, 2.4% of funds exhibited true skill over short periods (5 years) showing that, although rare, short term skill does exist. However, there is no evidence of long term manager skill. It is almost impossible for investors to capitalise on the short term skill of these rare funds as the report shows "outperforming funds only exist until investors are able to locate them."
Conclusion - The report confirms our belief that it is simply not worth the risk of trying to select outstanding fund active managers to pick stocks when 99.4% of all managers either under perform, or at best equal, the market over longer terms.