Stock or Industry Selection

A paper published by Jeffery A Busse and Qing Tong, of the Goizueta Business School at Emory University, explains that for consistent performance in a mutual fund its better to have a manager who can pick industry sectors than one who can pick stocks. The paper explains 'that industry selection contributes substantially to fund performance, accounting for roughly half of a fund's abnormal performance.' The paper states that this importance of industry selection is 'stable across time, with little year-to-year variation in the mean contribution across funds.'

The paper is based on analyzing almost 4,000 actively managed domestic mutual funds from 1980 to 2006. The researchers analyzed a manager's industry selection ability by setting up a hypothetical portfolio which replaced each stock in the fund with an index for that stock's industry. For example if a manager bought Johnson and Johnson, the researchers added an equal dollar amount of an index representing major drug manufacturers. If the hypothetical portfolio beat the market, the manager was showed to be good at industry selection. If the manager's stock picks outperformed the hypothetical portfolio, the manager was judged to be good at stock selection.

Based on the analysis, the researchers found that a manager's margin for beating the market was split almost evenly between stock selection and industry selection. However, over a longer period of time, industry selection was a skill that would most likely persist. The paper also found 'a negative relation between fund portfolio size and stock-selection skill'. This 'negative relation' was not present when comparing the size of a fund and industry selection. This fact probably explains why some large mutual funds preform quite well, while others suffer.

Below is a chart comparing Fidelity's Magellan Fund (market cap $38.40 billion) to American Funds Growth Fund of America ($165 billion). Between 1977 and 1990, the Magellan Fund, under the management of Peter Lynch, had an annualized returned of 29%. The fund ran into problems in the late 90's when the funds asset base swell to $100 billion dollars and ended up a 'closet' index fund. As the fund grew, the manager was unable to put that money to work effectively. Stock selection which had been the strong suit of the fund's new manager broke down under the weight of its asset base.

Source: bigcharts.com

On the other hand the Growth Fund of America, has an annualized return of 13% compared to 8.2% for the S&P 500 over the past 10 years. This fund has been successful due to its management structure and sector selection. Rather than relying on one manager, or even a team of in-house managers, it is managed by multiple management teams. This structure has allowed it to grow, while still preforming better than the market. Granted that at $165 billion, this fund will probably have a difficult time maintaining its returns, and will probably end up mimicking the returns of the market. (How large is too large?)

Below is a chart of Van Wagoner Emerging Growth Fund (market cap $15 million) compared to the Russell 2000. In the mid to late 1990s, the Emerging Growth Fund was returning 50% plus returns each quarter by investing in small-cap technology stocks. However, since 1995 the fund has an annualized return of -7.8%. You would have made more money leaving it under your mattress.

Source: bigcharts.com

The funds dismal returns are due to stock picking, and industry selection. The fund focused on small-cap stocks in the technology sector, not the best place to be since the Internet bubble popped in 2000.

Source:
'Mutual Fund Industry Selection and Persistence', By Jeffery A Busse and Qing Tong
'Picking the Forest or the Trees', by Mark Hulbert, NY Times