By considering different approaches to defining a fundamentally weighted, global, small-company equity strategy, the author is able to isolate the impact of various definition criteria. Using these results, he observes that certain definitions perform better than others. He then expands on these findings to better understand what may be causing certain definitions to outperform others.

Current industry practice is to group investments with others of similar attributes. These attributes can be geographic location, size, industry, or any other common underlying characteristic. The author develops his own definition of a particular investment strategy and seeks to determine the best version of the definition. The criteria he uses to evaluate these definitions include Sharpe ratios, excess returns, tracking errors, and information ratios. The author builds on prior research that investigated various security weighting schemes.

The author finds that all strategy definitions outperform the MSCI All Country Small Cap Index benchmark. But certain strategies appear to perform better than others. For example, superior returns can be achieved by excluding potentially expensive large-market-cap companies. The author further contends that a portfolio’s expected return should be inversely ranked by the portfolio’s average market value.

To achieve these superior returns, the author uses various security selection restrictions, such as excluding large-market-cap stocks or excluding large fundamental stocks as determined by a four-factor measure of company size. He expands on the market-cap restrictions by applying market-cap classification bands. Lastly, he uses “rankdiff” adjustments to curtail investments with large fundamental weights relative to their market-cap weights. This approach reduces both total risk and tracking error.

These conclusions are significant because they underscore the idea that an investment strategy’s definition can have a significant impact on the strategy’s performance. Moreover, these findings reinforce prior research that emphasized smaller market capitalization as a key component for earning excess returns. Ultimately, the results suggest that one definition of small companies may be preferable to another.

To simulate portfolios of publicly listed securities, the author applies various constraints to the MSCI All Country World ex-Australia Index. For example, companies must have reported at least one full year of financial results, have a minimum market cap of A$50 million, and maintain a trailing one-month average daily volume of A$250,000. Simulated portfolios are reconstructed annually at the end of December for 1993–2010.

The author uses a four-factor measure of company size in the fundamental scoring and weighting methodology. The factors include annual sales, book value of equity, cash flow, and payout. He also uses five-year averages, or less if available, for the sales, cash flow, and payout factors, as well as the most recent book value of equity. Payout factors of zero are excluded from the averages. He then applies limitations to the investment strategy’s definition. These limitations include fundamental rank cutoffs, market-cap rank cutoffs, and universe counts.

Lastly, the author considers the use of market-cap classification bands to reduce turnover and improve performance, which is accomplished by specifying various midpoints and bandwidths to be included. He concludes that using these bands decreases turnover without hindering performance.

Defining an investment strategy represents a critical step toward implementing an investment philosophy. The author offers an interesting approach to discovering a superior investment strategy definition. Any definition that is able to produce superior returns could have important ramifications regarding asset allocation strategies. Ultimately, I hope future research expands on these findings, both in terms of the simulated time frame and the benchmark used.