FTF News got time with Jonathan Ehrenfeld, head of securities strategy at Swift, to discuss the impacts of T+1 shorter settlement upon securities transaction lifecycles.
Jonathan Ehrenfeld
(The North American push for securities settlement within the trading day plus another day, a.k.a. T+1, is not going leave much time for fixing failed settlement efforts, says Jonathan Ehrenfeld, who is head of securities strategy for the Swift financial message cooperative. Ehrenfeld is responsible for Swift’s securities, foreign exchange, and digital assets strategies and for engaging with securities communities across the globe. During the discussion with FTF News, Ehrenfeld referenced the Securities View offering from Swift, which tracks a securities transaction throughout its lifecycle and across multiple intermediaries. The conversation took place at the SIBOS Toronto conference last month.)
Q: How does T+1 actually affect what you’re doing? And how does it affect what Swift does for securities?
A: I wouldn’t say it’s affected — it actually reinforces what we’re doing because most of the products that we are developing for T+1, or most of the products that we’re developing for post-trade at large, are trying to solve some of the inefficiencies that are still in the market. … Those inefficiencies become more apparent or more important when it comes to T+1 because the shorter time you have to settle, then the less time you have to repair things.
We’ve been working a lot in post-fact reparations and exception management, and we see that there is a tendency [among firms] to move into pre-validation of things because you don’t have time to repair things once it happens. You have to make sure that whatever goes into our network needs to be as clean as possible so that it doesn’t fail, right? In case something goes wrong, you don’t have the time that you had before.
At a macro level — at the Swift level — we see that even though it’s the U.S. that is going to T+1 — and Canada and Mexico — and then other countries, most of the flows that we carry in our network are cross border, and are not local to the U.S.
So, it means that if you now have a Japanese investor that needs to go through Europe to then settle an asset in the U.S., we’ll still have to deal with it. But the problem is that with time zones, for them, you’re reducing their time to 80 percent [of what they had before].
That’s how it affects us, in the sense that people look at their trades and say, ‘All right, you need to come up with a plan because most of these flows go through your network, and you need to have a certain number of tools so that we’re sure that it’s going to settle on time.’
Q: For a lot of the SIBOS participants, it’s really a move to T+0 — many aspects of settlement have to be done intraday. And there’s been talk that the rate of trade failures is going to increase.
A: We’ve seen what has been happening in India, for example, when the move from T+2 to T+1 happened. It’s a smaller market, but it’s a good sample.
I was there when we moved from T+3 to T+2 in Europe, and it was the same concern, ‘There’s going to be an increase of failure…’ And the reality is it’s different.
What we saw in India is that, yes, there are a few days or maybe, perhaps, one week after the move where you see a spike in failures. But then after a few weeks, it actually goes down because people are pre-validating more … So automation is the only answer there.
I would argue that when we moved from T+3 to T+2, everybody said, ‘Yeah well, when we moved from T+5 to T+3, it was not the same. Now it’s very short — it’s only two days,’ and it was like a Y2K moment, right? Everybody panicked for a few years, and then when it happened, there was zero impact.
You need to get other opinions, but I wouldn’t say it’s going to be like Y2K and nothing will happen. There are going to be challenges, but I don’t think the challenges are going to be as important as people think they are. I think it’s going to go fine.
Q: Do you help clients with software issues, the nuts and bolts issues? Or do you just give them guidance?
A: Our main initiative [with Securities View] is that we have a way to track everything that happens in our network. … We know exactly when something goes wrong because we pre-validate at every stage. We compare instructions on the two sides because you always have a sale instruction and a buy instruction.
We also compare discrepancies … If there are exceptions, we capture those in real time and distribute them to all the chains. If there is a problem, we know exactly where the problem happens. So, if there is any responsibility that needs to be assigned, we can see it in our network.
So it’s like an Amazon tracking tool where you see where your package is, you see your instructions, and if something goes wrong, you see why it’s blocked and you can resolve it immediately.
Q: What do you do when there’s a problem?
A: Either you can use a very simple Swift Case Resolution solution or exception management tool, or you can actually send that exception to any exception management tool that you have. Most firms already have exception management tools.
Q: So, do firms learn from their mistakes?
A: It’s a good question because they learn. … We are making a lot of noise about what we’re doing in A.I. because we have all the data from all our network participants and all the historical data from every single customer. We apply some of these A.I. capabilities to that, so that it’s not just the customer that is learning, it’s the system that’s learning. So then it can prevent a mistake before it starts — sort of a predictive type of analytics.
Q: What are the top causes of trade failures?
A: The number one reason why things fail is liquidity problems. The second problem is data discrepancies, meaning that the buyer and the seller have inputted wrong data on the messages they sent, different places of settlement, different SSIs [standard settlement instructions}, different quantities, and different amounts.
Today, exceptions that are linked to these discrepancies are only captured at the CSD [central securities depository] a couple of hours before instruction actually settles. So you don’t have time to repair anymore in a T+1 environment.
Q: Which one is the easier problem to fix — the liquidity shortage or the data discrepancies?
A: Well, I think the easiest one to spot or the easiest one to capture in a data tool like Securities View would be discrepancies.
With liquidity, it means you will have to have access to all the accounts that these firms have. So, it’s more complicated to solve.
Q: So, what you just described is what happens now — it doesn’t really get caught until the custodian sees it?
A: Correct. Sometimes it isn’t caught until the CSD at the matching site actually sees it, which already is too late.
Q: So, it would be too late for T+1?
A: It’s really too late for T+2. … For T+1, it’s going to be impossible to repair, especially with these time zone differences — once you want to do a repair, your counterpart is already sleeping.
Need a Reprint?