Carolina, Scandinavia has a large and diverse community of buy-side firms. Why are increasing numbers of them starting to look at algorithmic FX trading?
The buy side firms in Scandinavia are looking at algorithmic execution for a variety of reasons. It is a mix of need for transparency, interest of being in control of their execution, accessing directly new pools of liquidity and having information on how much of their execution is done on the various external or internal venues of the banks. The reporting in form of TCA (Trading Cost Analysis) is also key to some buy side firms and simplifies dialogues with their clients or their different control departments or management.
SEB has one of the industry’s most advanced platforms for electronic FX. Why did the bank decide to start offering FX algos and how do they complement your existing market-making services?
Client demand is the main reason why we decided to start offering FX algos as a complement to our market making services. Some of our clients have been using algos for a long time and they were looking for superior algos in Scandi currencies from their Nordic provider. As a leading Scandi bank, it felt natural for us to offer algos using our vast experience in Scandies and taking advantage of our large Scandi franchise.
What do you see as the key value propositions of FX algo execution and in what ways are you helping clients to become more comfortable with it?
The value proposition varies depending on the clients and their own challenges and interests. One of the key value propositions is about accessing venues that you would not have direct access to without an algo. It can also be getting a transparent reporting of how you executed providing easy documentation to show best execution practices.
Getting our clients comfortable with the whole algo execution also varies a lot depending on how experienced the clients are with algos. Sometimes the clients just need to know how our own offering differentiates from other algo providers, sometimes it is about explaining algos in general to clients that have never used one before.
Giving the appropriate level of education on our product is key and SEB takes this as a serious responsibility.
Some leading banks have decided to offer a wide variety of different FX execution algos whilst others are focusing on providing just a few and making their workings more transparent and easier to understand. What route has your team decided to take and why?
We focus on a few algos that are easy to understand. Most clients already have a wide variety of algos from several banks at their disposal and we believe that less is more in our specific case. We might add a couple of algos as things progress but we will not offer 15 algos.
We prefer focusing on fine tuning and improving our existing algo offering, and putting significant consideration to netting effects and clients benefitting from our flow franchise.
What types of FX algo strategy are proving most popular with your clients and what direct benefits are they getting from their use?
The two strategies that are most used by our clients are our Peg and our TWAP. The benefits they are getting from those are saving on cost of spread and reducing the market impact on very large orders.
This is particularly the case in Scandies where the possibility to use our internal pool that is made from our franchise has had a very meaningful positive outcome for our clients. On top of this they get valuable information on the execution from the TCA.
The two strategies that are most used by our clients are our Peg and our TWAP
What steps has SEB taken to enable clients to get easy access to your FX algos and place orders directly themselves?
The integration towards Bloomberg and FXall was key to meet the clients where they want to place their algo orders. Some of our clients like to place algo orders on our single bank platform Trading Station or direct APIs, but the majority access them via multibank portals.
Does SEB develop its algos in-house and, if so, how important has feedback from clients been in helping you design, deploy and fine-tune your strategies?
We do develop our algos in-house and client feedback has been key to our journey. Clients are very good at giving us ideas and the challenge resides more in prioritizing that feedback rather than getting it.
The feeback we got has helped us decide from what types of algos to start with to tiny details on the TCA report. It definitely has been a strength to be able to quickly adjust our offering based on it.
What solutions to evaluate trade execution performance does SEB provide for clients who are using your FX algos?
We provide our own in house TCA report and we also provide child fills on demand that we can either send to independent TCA providers as we do for some clients today or that we send directly to the client when they want to use that data directly themselves.
What can we expect to see from SEB in terms of new FX algo products and services over the coming year? Do you have anything particularly interesting in the pipeline?
We do have one or two extra algos that we are looking closely at and getting client feedback on. The TCA is also a part where we could do a lot more as our offering is now only a post trade TCA, so some things could happen in that space too.
As a pioneer in electronic trading do you expect SEB will be looking at ways to shape the future landscape and evolution of algorithmic FX trading, perhaps by influencing industry standards, finding new ways to engage with the buyside and promoting best practice principles?
Absolutely, the use of algorithms in FX comes with benefits for the clients but also with many questions and reflections. SEB is keen on fostering those reflections and we are as we have been historically always keen on adopting and spreading best practice principles.
Artificial intelligence is a very interesting field in the context of algo execution. Our quant team is collaborating with the Royal Institute of Technology (KTH), and WASP, Sweden largest individual research program ever.
We are doing this in several ways including in supervising PHD students on AI applied to FX execution. We believe it is important to support basic research as well as experimenting with what emerging technologies can do in the market.