“While Orca-Direct is itself an algo, it is also the basis for our smart order router for all of the UBS algorithms”, Goodman explains. Since its launch in May 2018, UBS has not added any additional venues, choosing instead to focus on curating the liquidity, the probability of execution and working to improve clients’ trading experience. “Our view has always been don’t add another venue unless it has differentiated liquidity, so liquidity we can’t currently get, otherwise all we’re really doing is adding to information leakage, which is a key concern for our clients,” explains Goodman.
The number of FX algo users on UBS has also increased in that time as well as the percentage of flow, with the bank having onboarded 56 new clients since the start of the year alone. Goodman adds that from all the long only buy-sides who spoke to the bank a year ago about just getting into algos, nearly all of those clients are now actively using algos to some extent. “It is now rare to come across a client who doesn’t use algorithms for at least a portion of their flow if they are asset management long-only or a hedge fund with sizeable orders,” he adds.
In terms of performance, UBS has over the past year been putting a lot of work into the algo behaviour as well as the liquidity access. The Lattice platform, which is where the algos actually run, was built with the concept of big data and machine learning in mind. “We’ve been able to make a step change with Lattice in terms of the inputs into the algorithm which describe the individual pair, characteristics such as spread, volatility, volume, etc. By taking 10 years’ worth of data, then by applying statistical analysis and machine learning techniques, we’ve been able to get the precision of these values down to the actual second of a particular day of the week and account for seasonality or known events,” Goodman says. “We deployed that last October and a recent review indicated a material improvement in average performance for our clients combined with a reduced standard deviation.”
On the analytics side, Goodman adds that while the post-trade market is consolidating around the third-party providers, there are still questions about pre-trade TCA provision. Clients are not actively interested in these tools but adoption is not yet industry-wide. UBS is focusing on developing a consultancy-type relationship with clients and providing them with the data and analytics to be able to make better informed product choices.
Looking ahead, there now needs to be some effort from the industry to properly define the concept of internalisation in order to ensure that clients are able to make fair comparisons between different offerings, adds Goodman. “Then in terms of FX algo products beyond spot, NDFs seems like the natural next choice. As we start to see the liquidity become more reliable, then you start to think that you can build an algo offering on top of that. In the next few months will be a key focus for us and our competitors.”