August 25th, 2015, 10:20 pm
QuoteOriginally posted by: quartzQuoteOriginally posted by: hamandeggsAgree with Dominic. Most (decent) execution algos are built targeting specific benchmarks against which to measure market impact. Almgren and Chriss pushed this subject area forward in this paper. If you follow the maths in this you are going the in right direction. Other useful sources can be found by reference-searching for Almgren.These are definitely prerequisites, but are far from enough anymore to be competitive. They already stand to execution like Makowitz stands to portfolio management. There are vendors of impact models, benchmarks, TCA&c, just wade through their (grin) bullet points and prepare to recognize half of them, noticing how they're ahead of academia and remember how they're lagging the "serious" players.Exciting field but you won't find many mentors in the wild (there are 3-4 books trying to put order in the mess, but they fall short of being comprehensive and up to date). And everything is merging, you need to integrate a good transaction cost analysis, a risk management framework, merge investment & execution, joint portfolio execution, in short prepare to see the focus unfocus :-)Cant agree more.