Algorithmic Trading and Transaction Costs

Abstract

“Algos” are algorithmic trading strategies that are meant to optimize the execution quality of the trades in terms of transaction costs and market-timing. This chapter presents the transaction costs taxonomy and popular algorithmic execution strategies. Authors empirically examine a dataset of hedge fund transactions. Our results suggest that implicit transaction costs are characterized by a significant buy-sell asymmetry. To get some insight about the possible determinants of Implicit Transaction Costs, authors investigate the algo type and stock characteristics such as market capitalization, relative volume, inverse prior close, price momentum, buy indicator and trade duration. Both in-sample and out-of-sample tests show that a significant portion of transaction costs can be anticipated before the trade execution. Results show that high-level execution strategies can be constructed to optimize the algo choice.