THE VALUE OF TRANSACTION COST FORECASTS: ANOTHER SOURCE OF ALPHA
Mark Coppejans and Ananth Madhavan
This paper examines the impact of transaction costs on portfolio performance. Previous research on this topic has focused largely on post-trade considerations, that is, the impact of realized transaction costs on investment performance. By contrast, we focus on pre-trade considerations, namely the impact of transaction costs on portfolio breadth, turnover, and expected returns. Although costs reduce the information ratio, improvements in transaction cost modeling can mitigate these effects by increasing skill and breadth. Transactions cost considerations also affect the choice of portfolio turnover. Greater turnover allows for more active bets, increasing breadth, but also magnifies the impact of trading costs. Balancing these considerations yields an optimal turnover level. The analysis provides insights into the determinants of optimal fund capacity. We show that capacity problems are manifested gradually in the form of higher expected costs, reduced breadth, and lower turnover. Capacity is an elastic concept that is surprisingly responsive to even relatively modest gains in transaction cost control or forecasting ability. This suggests that fund managers can influence their capacity through investments in better execution research and technology.