TY - JOUR T1 - Sell-Side Algorithmic Offerings JF - The Journal of Trading SP - 43 LP - 45 DO - 10.3905/jot.2006.609175 VL - 1 IS - 1 AU - Richard Rosenblatt AU - Joseph Gawronski Y1 - 2005/12/31 UR - https://pm-research.com/content/1/1/43.abstract N2 - This study examines anecdotal and the empirical evidence available to date regarding the relative performance of algorithms. While there are clear efficiency gains from the use of algorithms, the performance we have experienced to date in our use has been adequate, but not impressive. Beyond the anecdotal, more reliable and comprehensive empirical data has unfortunately been very limited. The most ambitious study so far was completed by ITG in the spring of 2005. While the study is a good first step, unfortunately it raises more questions than it answers due in part to the limitations of the empirical data, but also to some unjustifiable leaps in logic. For instance, the benchmark being aimed for by the trades in the control group data set is unknown, yet comparisons are unfairly made to executions where the strategies employed are known. Similarly, when the absolute performance of VWAP algorithmic trades is cited, a two basis point average miss is lauded whereas we would conclude instead that the equivalent? of a penny miss per share is underwhelming as well as a confirmation that active strategies are on average beating passive, algorithmic VWAP strategies because it's a zero sum game with all trades together creating the VWAP on the day. Clearly, a lot more digging needs to be done before any real conclusions can be drawn about the relative performance of algorithms.TOPICS: Statistical methods, equity portfolio management ER -