Transaction Cost Analysis A-Z

Transaction Cost Analysis A-Z — November 2008

V. Trading Performance Measurement

widely used, it appears to be an improved metric that is more consistent and stable because it allows comparisons across securities and across days. Furthermore, this approach shares the principle of peer group review with the innovative and systematic framework for assessing the quality of execution that we propose and will develop further in the document. (1) Benchmark comparison This approach involves comparing the monetary value difference between the executed position and the position evaluated at some benchmark price. For a given transaction, the monetary value per share is given by the signed difference between the average price obtained for the trade and the benchmark price. A positive value indicates an execution that is less favourable than the benchmark price, while a negative value indicates execution more favourable than the benchmark. As we have mentioned above, the benchmarks usually used include pre-trade, intraday and post-trade reference prices. (a) Critical review of benchmarks The pre-trade benchmarks (T-1 close, T open, last trade price, last bid, last ask) provide transaction cost measures, not performance measures. They deliver good proxies for the costs of spread, market impact, operational and market timing delays, but they do not provide any insight into actual execution performance since they do not help gauge whether the costs incurred are acceptable given market conditions at the time of trading. The intraday benchmarks (VWAP, LHOC) compare the execution price and a kind of average market price of the day or of a

given time period. The belief behind these benchmarks is that you did a good job when you did better than the “average trader” of the day or period. In this sense, intraday benchmarks are the most consistent with a performance measure, assuming that the benchmark is a good indicator of the fair market price. However, although intraday benchmarks account for market conditions and trading activity, they do not provide a meaningful metric that is consistent across days and across stocks. We will look into this key problem later. The post-trade benchmarks (T close, next mid bid-ask) attempt to measure the market impact of trades. When used in trading performance analysis, they serve more as a measure of investor skill than as a measure of trader performance. For funds that track a benchmark index, post-trade benchmarks indicate the contribution of execution to total tracking error. (b) Major shortcomings As a whole, when used to assess trading performance, benchmark comparison has major shortcomings thatmake it inefficient and/or misleading. These shortcomings are listed and described below. Benchmark comparison may be biased The benchmark comparison may provide biased performance measures that essentially depend on the reason for the trade. According to Harris (2003), biases may arise when trading decisions depend on past price changes or when investors are well informed about future price changes. Some benchmark prices can deliver performance measures that will be systematically good or bad depending on whether the investor uses a momentum

59 An EDHEC Risk and Asset Management Research Centre Publication