Equities
may have been the right decision in the circumstances .”
Customised algorithms The world of algorithms ( algos ) has been a key part of the electronic trading landscape for many years . Now though , the way the buy side uses algos is changing , once again driven partly by regulation . MiFID II places a greater obligation on the buy side than ever before to understand the algorithms it is using and how they work , and to be able to explain this to the regulator on demand . Arising partly out of incidents such as the Flash Crash of May 2010 , when $ 1 trillion was briefly wiped off the value of the US stock market by algorithms feeding off each other in a spiral to the bottom , MiFID II demands a new approach to algorithms . Many buy side firms report that they are cutting the number of algos they use , and paring down the amount of different customisations . The reason given by many is that the regulation suddenly makes having a large number of different algos a risky proposition , since any firm that finds itself unable to explain exactly how each one works faces punishment from the regulator . “ We see two competing forces ,” said Hilton . “ One is a focus on decreasing complexity from the buy side . They feel they have a greater regulatory obligation to understand what the tools do , so want to simplify the strategies they have access to . On the other hand , there ’ s a huge focus on performance and TCA . The buy side want to get into the guts of performance and understand what they can do to improve things . There ’ s more focus on tweaking things to improve outcomes by a small degree . For example , figuring out a more intelligent minimum when trading in the dark , or modifying the VWAP curve because the trader has momentum in the names they are trading , or maybe just a small tweak that makes better use of the closing auction . These are the kinds of changes that are popular .”
The other factor , said Cousens , is a better in-depth understanding on the buy side in terms of what drives alpha for their flow . This greater awareness feeds back into the trend already noted above , since it leads the buy side to consolidate its algo providers and the number of algos that it uses . Credit Suisse notes that most buy-side clients are using two customised algos , on top of a selection of headline strategies . As with many aspects of the market today , data lies at the heart of the issue – and there are a number of issues to tackle . Hilton notes that trying to compare performance across different data sets , whether that be two strategies , or two brokers , is not easy . The data needs to be as similar as possible for any conclusions to be meaningful , but this is not always a simple task . It ’ s necessary to consider whether the data has been impacted by events in the market ( such an election ), or that the data is simply not homogenous , which would make broad comparisons flawed . The best way to analyse data may be to test different trading scenarios side by side , for example by running an order to strategy A and then running it to strategy B in a randomised way over a period of time , to see which produces better results .
The shift to the close The profile of the trading day is changing . A number of market participants have observed in conversation with K & KGC that there is a shift of trading activity towards the close , with liquidity thinning out in the intra-day trading sessions . The scarcity of intra-day liquidity led one buy-side trader to ask , half-jokingly but not entirely without serious intent , whether traders should just start trading at 4.30pm and trade the close . Credit Suisse estimates that intra-day depth of order book has been at its lowest level in the last four years . However , one possible criticism of the close is that since it is still based on the national exchanges , it is quite monopolistic – a situation that plays into the hands of exchanges , which can then charge higher costs . “ There has been a growth of the closing auction ,” said Hilton . “ That goes hand in hand with investment in passive products . Active fund managers are struggling to beat bench marks . Passive fund managers have done well . They often have the close as their benchmark . On the FTSE 100 , 18 % of daily volume
30 www . buysideintel . com December 2016