Wednesday, 25th April 2018. 1455 – 1335.
Assessing how emerging technologies – once implemented – can help firms increase efficiency is top of the agenda for many, and this panel gave a greater insight into the marketplace, pitfalls and potential.
Panellists featured spokespeople from Dataminr, FlexTrade and FactSet, looking at how execution and market data can be combined to provide real benefits for the desk, alongside an assessment of automation, OMS and EMS platforms.
There are limits to the effects and potential of technologies like Machine Learning (ML) – they can only go so far and are only being applied in a handful of areas currently.
Automation and electronification are going to complement traders – not put them out of a job.
There are some hints and tips that firms should be aware of before hiring a vendor for outsourcing – ranging from the longevity of the company to the provision of on-going training services for the purchaser.
Key Questions and Quotes
Can you talk about some integration work?
“One of the things we saw through 2017 is that there is a lot of consolidation going on among asset managers. Multiple managers are coming to us with OMSs looking to aggregate data from lots of different sources. They are looking to create one market/execution view so that they can consistently do their best execution across their book(s) of assets.”
“It’s the future, where we should all be going – meshing user solutions together in a tight integration. In the past, we’ve relied upon different solutions. API driven, direct connectivity allows traders to look at one homogenised screen and one holistic offering.”
Workflow automation – talk through what you see as state of the art from automation.
“We’re seeing a lot of discretionary asset managers looking for APIs. They want the information, but don’t want to tap in something new.”
“There’s a couple of things that we should clarify. First, automation and electronification does not replace knowledge. Second, there’s a misconception that automation is going to put traders out of a job – it isn’t! Like every other industry, the trader is becoming more efficient.”
“One of our buy-side desks has a high-touch and a low-touch team. It’s quite an interesting trend which we’re starting to see.”
“Nobody worries about the low-touch. What you can deal with in the low-touch can be automated – dealers want to deal with the high-touch. High touch provides a higher service to the front office.”
Do you have an example where you’re combining ML with data? Is anyone feeding this into an algo?
“It hasn’t been fed into an algo just yet. Our clients have eyes on the screen and are touching things constantly. Given the fact that there’s so much data, we provide real-time services.”
“The data and ML is evolving. We’re finding few applications of ML and AI currently because of the technology’s current limits.”
“Execution data can inform the choice of the brokers, using the data to make decisions. While there is a limit in how much you can automate, there are some things we can put in place – including hard stops and warnings.”
If you’re talking to a buy-side trader, what should they be asking their vendor to make a good choice?
“What post-transaction service you can provide? After we sign the deal, what kind of on-going training can you do? Also, what’s coming up in the future – how does your product evolve?
“The implementation process – how many IT people do I need to evolve this process. How scalable? What disaster recovery do you have? Do you have data receptors in multiple regions?”
“The data is the buy-side’s own data. You should be able to access and interact with it at any point – vendors cannot shut the buy-side off from the data.”
“Now we’re over the ‘hump’ and we’ve got the data/reports, so we can look at the benefits. There’s a tonne of benefit.”