Presence of Luck Vs Skill
In 2010, Fama and French develop on the debate of the presence of skill and luck within actively managed funds. Their findings were reviewed by the cross-reference of other articles dating from 1995 to 2016. The review is broken into the main argument of whether luck or skill is present, following on with how these funds perform compared to benchmarks and the question as to why they still exist.
Presence of Luck and Skill
Grinblatt et al (1995) concluded through the usage of momentum measures, time-series regressions and cross-sectional regressions of fund performance that mutual funds are inclined to purchase past winners discrediting the stock picking skill as they stick to familiarity.
Similarly, Carhart (1997) analysed funds between 1962 to 1993, grouping them into deciles. In testing the ability for fund managers to consistently earn abnormal returns, Carhart came to the conclusion that 46% of this abnormal return was credited to the 1-year momentum factor, which is the continuation of returns due to the previous year. This resulted in the slight discrediting of abnormal returns being due to skill as he believes, similarly to Grinblatt et al (1995), that fund manager’s performance is due to identifying past winners rather than actual stock picking skill.
Building on Carhart’s model, Bollen and Busse (2004) used an extended period of measurement. Results show no evidence of superior ability after accounting for the momentum factor which is proven through their inability to generate abnormal returns for long periods. However, doing the same for short and medium terms, it can be seen funds were able to generate these abnormal returns, as Carhart showed.
Kosowski et al (2006) took this further by applying the four-factor regression model to bootstrap simulations, involving randomly simulated runs of U.S. equity fund returns. The bootstrap simulations allow for more reliable results due to the ability to test larger sample sizes whilst randomly assigning data to tests. By doing this, they concluded only the top 5% managers were able to outperform the benchmark of passive funds once costs were deducted on a consistent basis. Therefore, disagreeing with the previous papers by concluding top managers outperform due to skill rather than other factors such as random chance or the momentum factor. Using the same data as Kosowski et al (2006), Cuthbertson et al (2008) are in agreeance as they find the top 5%-10% of managers have skill as they are able to do it consistently.
Using the same empirical methods applied in Kosowski et al (2006), Fama and French (2010) applied bootstrap simulations and time-series regressions for U.S. equity funds from 1984 to 2006. By comparing the estimated alphas of time-series regressions to actual alphas from bootstrap simulations, they conclude that very few active funds are skilled enough to create returns above the benchmark after fees. However, the ability of funds to significantly outperform is due to a combination of luck and skill.