Trend strategies based on changes in stock price or earnings are widely used by investors. In this report, we examine the performance of a trend strategy derived from gross profitability (“GP”). Gross profitability trend (“GPtrend”), was proposed by Akbas et al. who argued that the trajectory of a firm’s profitability is just as important as the level (GP). We define GPtrend as the year-on-year difference in either quarterly or trailing twelve month GP, where GP is calculated as revenue minus cost of goods sold, divided by total assets. Our back-tests confirm that GPtrend has historically been an effective stock selection signal globally, with the added benefit of low to moderate correlation with commonly used investment strategies. Our findings include:
• GPtrend generated statistically significant average annualized long-only and long-short excess returns in five of the six regions we tested the signal. Performance was strongest (long-short basis) in Asia ex-Japan (6.68%), Europe ex-U.K. (6.66%) and the U.S. (6.50%), and weakest in Japan (1.15%).
• Gross profitability trend was effective across multiple investment style categories, indicating that the factor can be beneficial to a value, growth or core and large/small cap investment process.
• GPtrend’s performance is not subsumed by gross profitability, earnings revision or price momentum: GPtrend retains its ability to separate winner stocks from loser stocks, after controlling for GP. The average annualized return of the most attractive GP/GPtrend interaction portfolio minus the least attractive interaction portfolio is 12.19%. The factor’s excess return is also still significant after controlling for both earnings and price momentum.
• Performance was robust to several methodologies of determining trend: We computed gross profitability trend using six different methods and all six trend metrics generated statistically significant average annualized long-short excess returns in the Russell 3000 universe.