Richard Turner, Senior Trader at Insight Investment in London, spoke with Markets Media Group Senior Writer Julie Ros about FX markets and trading.
You were working at currency management firm Pareto Investment Management when it merged with Insight Investment. Tell me about that history.
I had spent about seven years with Pareto prior to the 2013 merger with Insight Investment. Our team combined with the existing Insight FX capability to become an FX specific investment team within Insight. We moved bringing our existing FX mandates.
As an asset manager with £629.8Bn under management, we have a lot of different areas in terms of strategy. Historically Liability Driven Investment (LDI) drove our business, but we now offer an integrated approach to pension scheme risk management. We help our clients by offering them help in areas focused on risk management, fixed income, responsible horizons, multi-asset, currency, cash management, and specialist equities.
Since our initial merger, there’s been quite a lot of structural change in the marketplace with regards to FX. In particular, there’s been a client focus towards more passive management. We still have discretionary mandates and model-based Alpha strategies, but the largest growth in AUM has come from passive hedging mandates. That said, we have seen a pickup in demand for currency alpha in the past couple of years. We believe our Currency Solutions Team is one of the industry’s most experienced, and we offer dynamic currency management and absolute return currency alongside our passive offering. This places us in a very competitive spot to help our clients with all of their currency needs.
What has driven the shift to passive management?
FX is a very cyclical market where we see cycles of interest in both passive and active management depending upon client needs. Our challenge is to help clients find solutions to their specific needs.Historically, the movement to passive mandates was driven by UK legislation regarding collateralization of FX swaps. Another factor was providing an effective hedge for our LDI clients, particularly those coming to maturity who looked to de-risk to passive management.
What are your thoughts on the practice of pre-hedging?
It’s a debate that’s ongoing. The question is really whether pre-hedging is appropriate for your execution. We speak to our banks on a day-to-day basis about the behaviors of their execution and ask that they publish their disclosures on pre-hedging for transparency purposes. That ties in with the work we do around the microstructure of how the algos trade and what footprints are left in the marketplace. I think it’s very important that you do your due diligence to work out appropriate behaviors for your execution. We try to align our execution with best practices consistent with the FX Global Code.
Tell me a bit about how your desk operates.
We are very much a flow desk. We have a lot of different stakeholders, so we get a lot of flows from different parts of the business. We have model-based, quantitative strategies as well as discretionary and fixed income teams to name a subset. A lot of what we do is managing the flow when it comes. As a team of three, we look to execute flow as efficiently as possible. Our biggest challenge is minimizing market impact. A lot of our time is spent looking at how we trade and how we interact with the marketplace such that we leave as small a footprint as possible.
As a side, and in order to collaborate, we also interact with portfolio managers. This can be the discretionary team who are very FX focused or it can be portfolio managers, managing the various Insight funds. We spend some time trying to find out the pertinent information people want in terms of market color and what’s driving market moves. In essence we are trying to add alpha at all levels of the investment process. We are also looking at limits to liquidity – how much we can get through the market at any particular time of the day.
We constantly assess the microstructure of individual markets and try to minimise any costs where we can. Whether it’s challenging vendors or reducing fees, we like to know what’s in a price. This includes understanding the costs banks are being charged for access to liquidity, as all prices we trade on will include explicit charges. We collaborate with the banks to try and make the costs of liquidity as transparent and disclosed as possible, aligning with the FX Global Code.
Which currencies do you trade and how important is electronic trading to your business?
We trade everything from G10 to frontier markets. We predominantly trade electronically and do so via bank algos or on a “request for quote” (RFQ) basis. We trade spot and then roll it forward to a specific date for each individual client or group of clients, typically rolling swaps in the one- to three-month tenors.
We do trade NDFs electronically, but bank algos are typically not aligned in construct to how we would like to trade. We are hopeful of improvements in this area in the future such that our NDF electronic volume will grow in percentage terms.
As early adopters of electronic trading, where’s your focus at the moment?
We’re looking at a lot more automation to enhance our processes. It’s akin to an airline pilot using autopilot. That’s how we view FX trading going forward. With automation, we’ll use a rules-based approach to automated execution. Certain acceptable parameters will be set and if the automated validation is not reached, the order will be kicked back to the trader for manual intervention. The trickier trades will take more of our focus, while the smaller trades that satisfy our criteria can be auto-RFQ’d or algo’d.
As with all our trading, there is a post-trade exceptions report that uses transaction cost analysis (TCA) to parameterise outcomes and check them against acceptable benchmarks. Any execution that is above the exception level will be investigated and signed off post trade.
How do you incorporate working with your tech team?
I lead an internal technology group (Trade Tech) that looks at cross-asset trading issues. We often find that two or three desks have common problems, and we use this group to highlight these in the hope to get the solution prioritised from a tech perspective. The trading tech group meets every two months to discuss ideas and focus on any distinct problems to get them prioritised in the tech queue.
Do you interact with bank algos as a means of execution?
Yes, we interact with lots of different bank algos and have quite a detailed policy around how we do that. When a new algo is built, we don’t just trade it straightaway. We’ve got an integrated signoff process that involves compliance. We have to make sure that we understand the algos we’re using, and that they interact in the ways we want them to. Algo selection is aligned with the FX Global Code to ensure that if an algo says it’s going trade like A, it doesn’t then trade like B. We do a lot of due diligence around algos before we employ them.
We’re very much focused on the relationship model with the banks and do everything in our control to avoid interacting with counterparties that HAVEN’T signed the FX Global Code. We extend this to the underlying LPs in the algos we trade where possible. We see the banks as partners, so we collaborate with them as much as possible. Quite often that means helping them build out their algos so they’re designed in a way that is suitable to the marketplace. We have very good relationships with both “e” teams, and voice teams at banks. It’s very much driven on a relationship basis.
What’s on your radar in terms of market structure issues?
We’re always on the lookout for what could cause us disruption in the future. We think the move to T+1 settlement in the US has the potential to make a big impact, so that’s on the radar.
Another key topic for us is the curation of liquidity and how people behave in the marketplace. When we are interacting with the market, we want to know that we are trading in a manner that satisfies our execution outcomes and the guidance given in the FX Global Code. It’s important to us to make sure the behaviors of the people we’re interacting with are consistent with our view of the FX Global Code.
Do you prefer that counterparties and trading venues sign onto the Global Code?
Yes, it’s definitely our preference. We encourage everybody who we interact with – whether it’s banks, liquidity providers, vendors, etc, to adhere to the principles of the FX Global Code. It’s important that in this world we operate in, that everyone is aligned with the nuances of the Code.
One of the challenges for us is to make sure that the people who’ve signed the Code are behaving in a way that is consistent with its principles. It’s important that the Code is not just something you sign up to and then forget about – it must be embedded within your day-to-day processes. It’s also important to challenge your counterparties to be consistent with the Code.
Insight Investment recently joined the Foreign Exchange Professionals Association (FXPA) and you are chairing the Buy Side Working Group. Can you tell me about some of the priorities?
Advancing awareness of the FX Global Code amongst the buy side is among our top priorities. Additionally, we’re all striving to get liquidity in the marketplace, so looking at ways to improve the way we source liquidity is also a high priority. Where we identify common problems, taking a more joined up approach with our peers will provide us with a better chance of solving them, and in a more timely manner than trying to do so as individuals.
In terms of TCA, do you do your own work or use an off-the-shelf system or systems?
All of the above. We use BestX as our off-the-shelf TCA provider incorporating day-to-day exception reporting as well as best in class analysis of our trading protocols. BestX is heavily embedded within our algo execution to help us achieve “best-in-class” outcomes. We also have a relationship with Tradefeedr to do a lot of the forensic, deep dive work. I also challenge bank TCA in a very unautomated way. I’ll download information we get in the TCA reports from banks and get into the weeds of what we see in those. In my experience there’s really three angles to our analysis.
How do the banks react when you approach them about some of the things that you find?
Very well. The banks that we partner with get our flow, so ordinarily whenever we have queries, they’re more than happy to help answer our questions. More often than not, we’ll find something that is mutually beneficial. We really see our banks as partners, not just liquidity providers. That basis is very important for us to be able to continue to strive and improve the way we execute in the marketplace.
How do you feel about peer-to-peer trading?
It’s something that’s very interesting, and there’s quite a few different vendors out there that are offering different solutions to this. They offer another source of liquidity that we can interact and execute with.
We interact with peer-to-peer via the algos. Peer-to-peer liquidity providers appear as another venue with which we interact. As with other liquidity providers, we analyse peer-to-peer fills for information leakage and fill behaviors, to ensure that this liquidity is suitable for us to trade against.
Does Insight trade digital currencies?
We are constantly keeping an eye on these. Our company wrote a paper on Central Bank Digital Currencies (CBDCs) last year, but so far haven’t traded digital currencies. There are a lot of digital currencies available at present and Insight will be monitoring the market for opportunities moving forward. Rationale for CBDCs will focus on ensuring financial stability and the promotion of financial inclusion, but the central banks will have other factors that will drive the adoption of CBDCs.
Any thoughts on AI or machine learning?
We deal with a lot of data on a day-to-day basis. The problem for traders like myself is deciding what data points are relevant and useful. The pieces of the jigsaw are not that complicated. If you’ve got a trade on your blotter such as a buy of $100 million versus euro, you (or an algo) may break it up into 100 trades of $1 million. For each child order there is only so much pertinent data. There’s direction, price, timestamp, underlying market (or neutral market feed) and size. The analysis is then quite straightforward as to how you ascertain whether that’s a good execution at child order level.
Outside of financial markets, AI and machine learning will be increasingly important in areas such as medicine and engineering, but I’m not so sure about the benefits for our sector yet.