The Full FX Conversation: With Alex Knight, Baton Systems

With settlement risk a hot topic in the FX industry, Colin Lambert talks with Alex Knight, head of EMEA at Baton Systems, to find out what the industry is doing to reduce it, and how far down the road to reform it is

 Colin Lambert: The BIS flagged the size of the settlement risk challenge in the 2019 Triennial Survey; the Global Foreign Exchange Committee then took up the baton – no pun intended – with the FX Global Code; and more recently the Global Foreign Exchange Division chipped in with a white paper. What are your thoughts on the GFXD paper? 

 Alex Knight: We fully agree with the need to reduce settlement risk and we’re supportive of the efforts being made by the GFXD. We completely concur that settlement netting is preferable to gross settlement, however through discussions with industry participants we’ve been repeatedly surprised by just how much flow fails to be netted. This is often because netting presents such a massive challenge to operations groups relying on legacy, siloed technology without the flexibility to net on a continuous or configurable basis.

If netting is to become the norm we believe the adoption of automated and configurable netting practices that flow seamlessly into a PvP settlement process will be key. The report cites CLS as the example, but there are other alternative offerings available that empower firms to achieve automated, configurable bilateral netting and then settle on a PvP basis on demand.

Although the paper references the use of manual settlement netting processes we would advocate this should be seen as a last resort. Technology is available today to automate these processes. Manual netting is massively resource intensive and introduces the potential for human error and operational risk both in the calculation of the netted values and through the agreement process.

 I think most people are now aware there is an issue with settlement risk, but less perhaps understand the real challenges the industry is trying to solve for – can you explain that? 

 The pre-trade space has enjoyed enormous investment in technology, volumes have rocketed accordingly and transactions times shortened, but the post-trade technology space has, in reality, barely moved in the last 20 years. This reliance on dated, often siloed, technology increases not only settlement risk but operational risk more generally.

With our existing clients and a very visible high-quality pipeline, the network amongst top-tier banks is going to be material by Q4 of this year. We are getting a lot of traction.

Manual processes tend to be rushed, which is when people make errors. You can have as many eye checks as you like, two, four, six, eight, but people are inevitably going to make errors if they are operating manually. We talk to senior people at the banks, who are terrified of the potential implications (cost and otherwise) of a payments error. As an industry we need to automate these processes.

I mentioned netting earlier and clearly risks increase when there isn’t as much netting going on as there could be . Not only that, but what does go on is heavily siloed. Under ISDA you can net multiple cashflows, but most institutions net cash FX trades only, then they settle their NDF and options trades separately, not to mention collateral and cross-currency swaps.

Market access becomes an issue at some point as well, because credit officers are unwilling to give out huge settlement lines, which has an impact on FX – where we know that more than 50% of transactions are settled outside PvP – as a product that requires large settlement lines and allocation of capital. These lines are often shared across a range of products as well and then you end up with the situation where FX is consuming a large chunk of that settlement line, but it’s for a relatively low yield, which means inefficient use of finite resources. Settling trades via PvP eliminates these challenges.

A third issue is liquidity, which is a growing challenge that will only become bigger in the current interest rate environment. Big banks have to put aside enormous amounts of funding and resources in order to meet their intraday liquidity requirements and managing those nostros is hard work. The fact is with processes the way they are now, banks can’t predict when they are going to get paid. Under the regulators’ eyes, it’s not sufficient to break payments because you haven’t been paid, you have to be funded, so generally you over-fund your account, which is, again, inefficient, but more to the point you don’t have visibility or transparency over your liquidity position during the day.

 So, this sounds very much like a technology challenge, is that what the Bank for International Settlements was getting at in 2019?

 From a technology perspective, the cost of maintaining these post-trade systems is huge. You have legacy solutions that are really hard to take out, being supported by technologists using legacy programmes they may have rarely, if ever, worked with. It needs modernising if the industry is to have effective payment rails at the wholesale level.

Currently, we have an environment where post-trade systems are heavily siloed, they’re server-based, rely upon daily-batched processing, which is inefficient, and the technology is expensive to maintain.

As for the BIS raising the issue, I think the data was showing the problem – the data really just spoke for itself and needed highlighting. Some say that the data might be overstated, nevertheless, the number is still huge and banks are talking about 60% of their volume being not safe-settled – this needs solving for.

 There is, however, pressure on the system, as highlighted by the BIS, what is Baton doing to alleviate that pressure?  

 What we are doing here at Baton is to deliver market infrastructure rails to drive high speed and riskless settlement, for both cash and securities, in a manner that’s accessible to all wholesale market participants, globally.

We see this as filling a gap in the market. CLS is a massively important component of our industry, but we also know that there is, as the BIS indicated, a massive part of the market that is not, for whatever reason, covered by that mechanism.

We look to complement this through a different approach, we do bi-lateral netting rather than multi-lateral, and we don’t rely on central bank accounts. This means people can have flexibility about how and when they settle – especially as we enable firms to mutually configure how they net and to do so on an automated basis.

Because we’re not using central bank accounts, we don’t need to limit participants to those institutions who have central bank access and we don’t need to be beholden to changes in law and central bank processes to provide PvP settlement for a new currency. We’re relying on well-established electronic funds transfer legislation for the settlement process we operate. As a result, firms can settle on a PvP basis with any counterparty, in any currency they choose with legal settlement finality.

You used the word ‘complement’ there, how would I, if I were sitting in a bank, blend Baton and CLS for example?

 There are opportunities across three obvious horizons to start with – those currencies that are non-CLS eligible; those counterparties who cannot access CLS; and non-CLS products.

The BIS highlighted how much of the FX world is not able to access PvP mechanisms, so there is a huge gap that we can fill there. Secondly, for many participants it is not economically viable to become a CLS member or access the mechanism through a third-party settlement provider.

These are not exclusive challenges, they overlap. For a major bank, being able to promote counterparties onto the settlement system as a node on the shared ledger is a powerful proposition. If you’re going to make the investment to use the solution you need to use it as much as possible. As well as reducing settlement risk, this also has the impact of improving market access.

And, as noted, there are trades that are difficult to settle using CLS mechanisms, like short-dated swaps, the out-leg of the in-out swap and products like intra-day swaps – we’re working to help banks better manage their risk and liquidity in these situations.

 Recently, HSBC and Wells Fargo announced they’d be using a blockchain-based solution for FX settlement, it was the world’s first PvP settlement outside of CLS, that solution is built upon Baton technology isn’t it?

 It is and it’s important to note that the solution is in full production. People often suppose it is a proof of concept because so much of what you read about in this space is PoC, but that is not the case here. This is DLT delivering real value to institutional finance today. This is validated and fully-operational and with the focus on settlement risk, we think this is significant. It’s gone really well, its being used every day, and meaningful volumes are going through, across a range of currencies. We are looking forward to extending the scope of that relationship in the coming weeks and months.

 Going back to the GFXD paper, it stated that consistent settlement practice is preferable to ad hoc arrangements, which to me seems obvious, but it is really saying ‘get this automated’, which is where DLT comes in isn’t it?

 Not all transactions are going to be netted and some have to be pulled from a netting set for various reasons, but this does not necessarily mean those trades then need to be settled manually or without the safety of PvP. With distributed ledger technology, because the parties are viewing shared data and using shared workflows, they simply need to agree that those transactions are to be removed from the netting set and submit them to an additional automated process for isolated settlement on a PvP basis.

It’s important to note that DLT is the foundation of our solution rather than blockchain. The latter has many valuable attributes and we try to take out the best of those attributes, but one of the challenges with blockchain is that everybody’s process has to be the same. This means you get a quite brittle relationship because if one party wants to do one part of the workflow in a different sequence, or they want an additional step, it becomes very difficult on blockchain.

Distributed workflows promote a collaborative process, whereby two legal entities going through the process of matching and netting and settling a set of transactions, can agree to do so in a way they both understand, with a common set of data and mutually agreed workflows. Crucially, however, they can also have their own private steps in the workflow to meet their own specific requirements.

 You mentioned earlier siloed technology – and that obviously does not talk to standardised practices – is one of your challenges dealing with these siloes and legacy technology?

 Certainly one of the challenges around the amount of siloed and legacy technology is that people don’t want to, or cannot, replace it quickly – doing so can be really difficult. This means they need solutions that are designed to work with these legacy technologies, using all of the standard communication protocols to access and deliver real time data and instructions, in a production environment.

Our solution isn’t a box that sits in isolation, it needs to and does talk to a myriad of systems. By consuming data as far up the pipe as possible, rather than from a database further down the pipe, we allow the customer to think about decommissioning elements of its technology stack over time because we can take them out of the loop in a controlled manner.

The retirement of legacy systems is one of the end goals for the industry, we allow that by creating a ring road around the legacy piece. It’s not a big bang, it’s too important to move too quickly, but it will happen.

 How onerous is the integration? Obviously technology teams at a lot of banks are over-stretched and will not want another big project.

  Very light, there’s a Swift integration and one to the payments gateways, as well as to the deal capture systems so we can run a continuous reconciliation, but beyond that there’s not a huge amount that needs to be done.

The technology work with our existing customers was completed in a matter of weeks, so once the decision is made, it’s a fairly easy process.

 Looking ahead, how important is scale going to be? Howe much of it is about getting more banks onboard?

 With our existing clients and a very visible high-quality pipeline, the network amongst top-tier banks is going to be material by Q4 of this year. We are getting a lot of traction, which can become self-fulfilling. The validation with HSBC and Wells Fargo was an important part of that. Across our solutions we’re now handling over $15 billion in transactions daily.

There is still the question of how we sponsor access, because there is still a gap for most firms where they would like better control and access to a set of counterparties they don’t currently have – regional specialists perhaps. With a solution called Core-Payments, we are providing firms with the ability to intelligently sequence and manage those outbound payments with counterparties that don’t need to be nodes on the Baton ledger.

That sits very nicely alongside Core-FX, where a customer is using CLS for a segment of its flow, and using Core-FX from Baton for another portion of its flow. They can settle PvP with counterparties that are a node on the ledger, and then use Core-Payments workflows for those counterparties that are neither.

Pull that together then you have a really nice, strong, robust, harmonised operational process that is efficient and involves much lower risk and is operating with real time data.

Going back to the GFXD paper, it cites CLS as an example of an automated PvP system and the need to submit and match transactions in a timely manner. The reality is, however, that banks often conduct transactions that fall outside of CLS cut-off times such as value today transactions. Settling these transactions via CLS within the given time constraints may not be possible and participants are obliged to use other settlement processes.

To meet the need for a more flexible approach whilst retaining uniformity, Baton’s technology enables netting calculations to run continuously right up to the point the parties choose they want they settle – which they can then do on-demand, multiple times a day and then complete the entire process in a matter of minutes.

As a result, banks now have the opportunity to settle as large a population of their transactions as possible through an automated and inclusive PvP mechanism.