Is Small Better When It Comes to Investment Management?

I wrote an article for the Spring 2005 edition of the Canadian Investment Review that analyzed about 300 long-only equity managers to study the effects of a number of organizational and process factors on their performance over a six year period ending in December 2003. The study was used to compare the performance of those managers with ‘entrepreneurial’ qualities of high employee ownership and smaller size with managers that are owned by an institution and larger in size. The analysis was a self-described ‘Moneyball’ type approach to manager selection. Data came from the Brockhouse & Cooper (now Pavilion Advisory Group) database and covered Canadian, U.S. and EAFE managers.

The original study looked at eight organizational factors (employee ownership, firm age, firm size, product age, product size, size of investment team, experience of investment team and personnel turnover) and four factors related to the firm’s investment process (top-down vs. bottom-up, number of company visits, number of securities in the portfolio and portfolio turnover). Two factors were used as control factors: value vs. growth and market cap bias.

Regression analysis was used to analyze performance relative to the appropriate benchmark as well as on a risk-adjusted basis by information ratio and by Sharpe ratio. Performance for each factor was grouped into quintiles and results were reviewed for statistical significance.

The results showed that four factors were statistically significant: a high level of employee ownership, a low level of personnel turnover, low portfolio size (i.e. portfolios that were concentrated or had a fewer number of securities) and firms with a top-down process. The first three factors were all positively correlated with each other and with another ‘entrepreneurial’ factor – small firm size. Top-down was not correlated with either entrepreneurial or institutional factors.

Updated Study

The study was re-run to cover the January 2003 – December 2012 time period. The performance and characteristics of over 400 long-only equity managers were analyzed, again using information from Pavilion’s Canadian, U.S. and EAFE manager databases. Three additional organizational factors were added: institutional share (the percentage of a manager’s business that comes from institutional clients), firm size in 2008, and product size in 2008. Firm and product size at the end of 2008 were added to get a sense of the ex-ante size of the firm/product in addition to the firm/product size at the end of the period (2012). A true ex-ante date for firm/product size would have been at the end of 2002, but unfortunately this data was not readily available.

The data from the factors were again regressed against excess returns over the benchmark to test for their statistical significance. In addition, each factor was grouped into quartiles to measure their performance by excess return (alpha), information ratio and Sharpe ratio.

The results from the analysis indicate that there were seven factors that had some degree of statistical significance:

• High Employee Ownership* – Higher levels of employee ownership resulted in higher performance.
• Low Portfolio Size** – Portfolios that were more concentrated (i.e. fewer securities) outperformed.
• Low Personnel Turnover** – Low levels of personnel turnover resulted in better performance
• Low Firm Size (2008** and 2012*) – smaller firms as measured at the end of the period (2012) and four years from the end of the period (2008) performed better than larger firms.
• High Product Size 2012*** – High product size at the end of the period was associated with better performance. However, low product size in 2008 was also associated with better performance, although this was not statistically significant. This suggests that the best performing products attracted a large amount of assets by the end of the period under study.
• Experience* – Higher levels of average experience resulted in inferior performance. However, this is driven by underperformance in the second highest quartile (firms with investment professionals with 20-23 years average experience). Since the other quartiles were roughly equal, this might be a spurious result.

Summary & Conclusions

The magnitude of the results from this second study were somewhat less robust than that of the first study, when there was more dispersion of active management due to factors such as the technology boom and bust of the late 90s and early 2000s. As an example, the top-to-bottom quintile differential for employee ownership was approximately 2% in the first study, compared to just 0.5% between the top and bottom quartile in the second study.

However, the results from this study are directionally similar to the original study. High employee ownership, low portfolio size (concentrated portfolios) and low personnel turnover were statistically significant drivers of better performance in both studies. These factors were also positively correlated with each other (except for portfolio size and personnel turnover) as they were with another other entrepreneurial-related factor: low firm size (in both 2008 and 2012). The significance of high product size in 2012 is an outlier, but it is tempered by the results of low product size in 2008 and the unavailability of product/firm size data for earlier periods. The outperformance of firms with below average experience appears to be spurious due to a lack of monotonicity.

Overall, the evidence again suggests that entrepreneurial investment management firms of limited size with high employee ownership have a leg up on their larger, institutionally owned counterparts.

*statistically significant at 90%

**statistically significant at 95%

***statistically significant at 99%

David Finstad is director, Hedge Fund Management at OMERS Capital Markets.