Tue, 24/04/2012 - 14:16
Trading algorithms are nothing new. But the way they’re being used is changing, with managers today requiring more and more customisation. From a technology perspective, this is putting sell-side institutions under greater pressure.
Algorithms have to behave in the right way for a particular trading account in a particular market. Buy-side traders know exactly what they’re getting when they initiate an order.
The only downside to customisation is that transparency can become something of an issue; quite simply, running multiple algos with multiple brokers can quickly become an operational headache. “It’s starting to become a case where some clients are reluctant to do too much customisation because it makes it more murky and confusing in terms of what they have set up with different banks. They will find their preferred brokers for certain market sectors and particular styles of trading and rely on those relationships as a key part of their automation process,” explains Owain Self, Global Co-head of Direct Execution, UBS.
As the trading environment gets more fragmented and regulated it is putting added pressure on brokerages to meet buy-side needs. Customisation is helping achieve this, and with sources of alpha becoming ever harder to find, trader want algos to find not necessarily the fastest liquidity, but the right kind of liquidity. Brian Gallagher, head of European electronic trading at Morgan Stanley recently said that the issue was not about better algorithms “but better usage of algorithms” and knowing when to use the right kind of algos in different periods of volatility.
“When there’s a lot of liquidity around people will tend to trade in non-displayed dark venues but as liquidity dries up what you tend to find is they have to resort to trading in lit venues (trading facilities offering trading via a visible order book) that will often dictate their algo choice. Traders are using algos that manage that balance for them and ultimately execute according to their liquidity objective – find the liquidity, and if you can’t, resort to the lit venue,” says Self.
Another area where brokers are helping the buy-side community is the use of algos not just for execution but automating small orders. This is helping managers focus more on larger orders where they can make a real difference in improving overall fund performance. Brokerages can effectively analyse trade flow and evolve the algos so as to improve execution on those smaller parts of the business. Not only does the trader benefit by adding value to larger trades but their broker(s) is able to optimise the smaller trade flow.
“That’s actually been pretty popular over the last 12 months. Every time I speak with a client they’re looking to move into that space and automate some of their small order flow,” confirms Self.
Even if traders don’t request customisation, brokerages still have to adapt quickly to ever-changing market conditions and evolve their algos accordingly to ensure clients get a consistent product. And in terms of future technology developments, Self thinks they will continue to follow the same key themes as last year – developments in customisation, in multi-asset trading (fixed income/futures/FX), and multi-stock trading (using portfolio algos).
“People get frustrated that you refer to the same themes but these things don’t happen overnight. The algos used in 2001 are nothing like those used in 2004, 2007 and 2011. Brokerages have started to deliver different asset class algos, portfolio algos and customisation and things are evolving at a significant pace.
“I also think execution consultancy – analysis and evolution of the buy-side trading process – will be an important theme where technology will play a big part. In the past it’s been more human-based but one of the newer themes we’re going to see this year is that process becoming more technology-advanced, more scientific,” adds Self.
Wed 23/12/2015 - 08:00
Thu 25/06/2015 - 10:40
Thu 15/01/2015 - 08:19
Tue 22/07/2014 - 13:01
Wed 23/12/2015 - 08:00
Wed, 31/Aug/2016 - 14:17
Wed, 31/Aug/2016 - 09:53
Wed, 31/Aug/2016 - 09:37
Wed, 31/Aug/2016 - 07:51
Wed, 31/Aug/2016 - 07:48
Wed, 31/Aug/2016 - 07:44