Settling scores: industry pushes back on new penalties in settlement efficiency drive

In order to go fast, you first need to perfect the art of going slowly. Legend has it that in preparation for his 1955 recording of Bach’s “Goldberg Variations”, Glenn Gould spent 32 hours playing through every single note at an agonizingly slow tempo. But when his time in the studio came around, Gould took the “Variations” at a blistering pace—his recording lasts less than half the time of most conventional performances of the work.

In Europe, discussions of how and when to reduce settlement cycles to T+1 have given extra impetus to the question of how to improve settlement efficiency on T+2. There is no point playing the “Goldberg Variations” in 39 minutes before you are able to play them in 32 hours, after all. 

But proposals for cutting down settlement fails have raised concerns from market participants and market infrastructures, who say that they could entail costly technical changes to the way they operate.

In 2014, the EU implemented the Central Securities Depositories Regulation (CSDR), a piece of legislation designed to harmonize the post-trade process by giving central securities depositories across the jurisdiction a common set of requirements. But some of the ideas relating to settlement discipline envisaged by CSDR were either postponed—like the system of penalties for failed transactions, which only entered into force in 2022—or shelved altogether, like the obligation for a seller who is unable to deliver a security to pay the buyer to acquire it elsewhere.

Data from the European Securities and Markets Authority (Esma) shows that the penalties introduced in 2022 have slowly brought down the rate of settlement failures. But Pardeep Cassells, head of buy-side customer experience at data and insights provider AccessFintech, warns that the battle with fail rates has not yet been won. 

“In my opinion, they’re still not at a comfortable level. According to Esma data, we’re sitting on a fixed income fail rate of about 11% across the CSDR trades, an equity fail rate of about 10%, and an ETF fail rate of 20%,” Cassells says. By comparison, she adds, “North America has a fail rate of between 5% and 7% across all instrument classes through the DTC/Fed.”

In a bid to bring Europe’s rates in line with other developed markets, Esma has issued a consultation paper calling for industry feedback on changes to the CSDR penalty mechanism. The tweaks suggested by Esma would either see progressive penalties introduced—increasing the fine for a settlement failure the longer it goes unresolved—or change the amount that is charged for failure to deliver certain asset classes depending on their liquidity profile.

Both options would make the penalties significantly more expensive. Although public data on penalties is incomplete—even Esma does not have a full breakdown of penalties by asset type and duration of the settlement fail—Peter Tomlinson, director of post-trade at the Association for Financial Markets in Europe (Afme), says it is possible to work out roughly how much the penalties would increase under each option. 

“We estimate, based on T2S data—which covers a number of the CSDs, but not all of them—that the current annual penalty amount that is debited is €1.7 billion ($1.84 billion) in total. It would be around €10.2 billion ($11.1 billion) under Proposal One, and around €23 billion ($24.9 billion) under Proposal Two,” he says.

Critics of the proposals warn that they could increase bid/offer spreads by forcing market-makers to factor the cost of higher potential penalties into their prices. But they would also have a more immediate impact for market participants and CSDs alike, by requiring them to adopt a new penalties-calculation methodology. 

Having adjusted to the penalties introduced in 2022, the new changes would bring in another layer of complexity, obliging firms to reprogram systems and increase the cash buffers they keep on hand to fund higher penalties.

“Everyone—from the CSD as the calculating entity through to the market participants who have to validate those numbers and try to predict the costs—would have to rebuild their technology to incorporate that increased complexity,” Tomlinson says.

Penalty shoot-out

Brokers are likely to be among the most affected market participants if the proposed changes to CSDR go through. As the conduit between market participants and the CSD, a significant proportion of fines go to the broker in the first instance—although the broker can choose to pass them down the custody chain to the client that caused the fail.

Tom Springbett, head of compliance advisory at TP Icap, points out that inter-dealer brokers are heavily involved in the settlement process. Inter-dealer brokers like TP Icap have two settlement obligations on every transaction: to deliver the securities to the buyer and the cash to the seller. In order to meet both of those obligations, Springbett explains, the inter-dealer broker relies on both clients first delivering to them. 

“We will often be the party at the sharp end of a settlement fail, because we’re the one with the obligation to deliver in the CSD. But it’s like if you’re buying and selling a house: If something fails in the chain four links away from you, then you still wear the costs. If something fails in the settlement chain four links away from us, but we have the obligation at the CSD, then the fine will attach to us in the first instance,” Springbett says.

This also means that inter-dealer brokers have a big job planning for, validating, and passing on penalties. Mark Vaughan, head of ops business partners at TP Icap, says the proposal of a penalty rate that changes with the duration of the fail would make this job considerably harder. 

“It’s fairly easy at the moment, because it’s a fixed penalty, whether it be a cash-generated penalty on failed purchases or a security penalty on failed deliveries. But if you add in progressive rates as well, it’s going to be something that we’ll have to seriously consider, and it will obviously be a cost to us,” Vaughan says.

There is also a question around resource management. The higher penalties would require all market participants to put aside larger cash reserves, impacting business models. 

In the case of inter-dealer brokers, there is one particular scenario that could leave them facing a hefty fine even if the settlement fail was not their fault. If they have enabled automatic partial settlement (receiving the available portion of the purchased securities rather than the full amount), but the buyer does not accept a partial, then the broker is penalized for failing to deliver the full amount, but only credited for the portion that the seller has failed to send them, leading to an imbalance on the penalty.

Vaughan believes that higher penalties may deter buyers from enabling partial settlement. “One of the biggest issues I see with the progressive rates that they’re looking to impose is that there’s no incentive for the end buyer to adopt auto partial, because they’re going to receive very attractive penalties. In fact, in some ways, it’s an incentive to not accept delivery,” he says.

In the instances where the CSD is unable to correctly identify the “at fault” counterparty, market participants can initiate claims against each other. As it stands, Afme advises its members to only initiate claims if the value of the penalty would exceed €500 ($542), reflecting the operational cost of processing a claim. 

Do we really expect that settlement efficiency is going to go up by applying those additional ways of penalizing settlement failures? That’s where we are not convinced

Dirk Loscher, Clearstream

Vaughan believes the changes to CSDR would see the number of bilateral claims increase significantly. “With the progressive rates they’re looking to hit, it’s going to bring so many more penalties above that threshold, and you’re going to have to increase your workforce to manage those claims. And any time you claim, it’s never a straightforward process; it’s always a time-intensive process. So administratively, it’s going to be really heavy,” he says.

Similarly, asset managers or hedge funds that do not currently see any impact from CSDR penalties might begin to take an interest if they rise materially. Currently, some prime brokers choose to absorb all penalties and credits. 

Even for asset managers who know that they are not responsible for the majority of their fails, and who should therefore be net up, penalties at their current rates could still not be worth pursuing, AccessFintech’s Cassells explains. But if the penalties are going up tenfold, those asset managers might want to see their credits hitting their account. 

“It could be that the primes then actually come under increasing pressure from the asset managers, and then you could end up with more claims being raised. … I think we’re going to see a ripple effect across the street,” Cassells says.

Structural substitutes

Replies to Esma’s consultation paper are due by the end of February, and the regulator will submit technical advice to the European Commission by September.

Sources who spoke to WatersTechnology for this article universally emphasized that increased penalties alone will not cure Europe’s efficiency ills. They said structural reform was also needed to simplify the post-trade process.

“We believe it’s an important part of the toolkit toward achieving improved settlement efficiency, but we would encourage the authorities to look at broader structural changes which might also deliver improvements to settlement efficiency. It’s not just about creating behavioral incentives in the marketplace; it’s also about addressing those structural issues,” says Afme’s Tomlinson. “Not just how do you punish settlement fails, but how do you prevent them in the first place?”

Several sources had examples of ways market participants and infrastructures can boost settlement efficiency under their own steam.

If you’re in double-digit fails on a T+2 settlement cycle, and you’re being impacted by penalties for all of these fails, how do you survive a T+1 shift in Europe?

Pardeep Cassells, AccessFintech

It can be a case of identifying and addressing weak points in the settlement process. AccessFintech’s Cassells says she reviewed a client’s data before a meeting at their office recently. “I was able to walk in and say, ‘Do you realize that every single day, on average, you have between 60 and 80 trades where, with this custodian and every [executing broker] that you’re working with, you have an issue in the Spanish market where your custodian is constantly flowing through Euroclear, your brokers are consistently trying to settle domestic, and if you’re not correcting that, every single one of these trades is failing?’ And that’s just the tiny population that I’ve looked at in one market,” she says.

Euroclear Sweden has had noticeable success in improving settlement efficiency rates for equities. Between Q1 2021 and Q3 2023, the CSD increased settlement efficiency for equities from 85% to over 96%.

“More or less when the central counterparty (CCP) concept was introduced in the Swedish market, we saw that the efficiency started to slowly drop in equities. And we reached levels that raised the attention of the regulators,” says Quentin Berillon, product manager at Euroclear Sweden. 

In 2015, the CSD started issuing SEK500 ($48) fines for failed deliveries to CCPs. The idea was to improve the efficiency on regulated flow executed on the Stockholm Stock Exchange and cleared by the CCP, as these trades were likely to create chain reactions when they failed.

In 2019, seeing that there was no significant improvement, Euroclear Sweden raised the penalty fee to SEK1500 ($145). This worked for a period, but the efficiency rate later degraded again.

Then, in 2021, Euroclear Sweden enabled participants to instruct a partial settlement of positions held in omnibus accounts for the first time. Later that year, the CSD implemented changes to its partial logic, facilitating the simultaneous partial settlement of a chain of linked transactions.

“We improved the partial logic that was used, so that in the same batch, we could do three iterations. So taking a new snapshot, looking at the available position, then doing the partial, then taking a new snapshot. ‘Okay, now I see that 20 has landed on the account, we can use those 20 to settle onward.’ And that was very useful for a chain of transactions. Because, for example, you could have participant X delivering to CCP A, then CCP A to CCP B, then CCP B to participant Y. That way, we could cure the entire chain in one batch instead of waiting one or two hours for the next batch,” Berillon says.

After launching this functionality, Euroclear Sweden saw a big increase in partial settlement. Over 15% of the value of equities now settle through partial, up from just 5% before the change to partial logic.

“Together with the partial functionality improvements, I think one of the key elements was prolonging our delivery-vs.-payment settlement window,” Berillon adds. “Prior to that, we had the cutoff at 2 pm. That was one of the earliest in Europe. In discussion with the market, we increased that by 1.5 hours to 3:30,” he says.

Another CSD that has made inroads boosting settlement efficiency is Clearstream. From 2022 to 2023, Clearstream reduced the value of fails by around 90%. Dirk Loscher, Clearstream’s head of custody and investor solutions, attributes this to client interaction and education.

“Historically, there have been challenges around having the right standard settlement instructions. Transactions have been unmatched for some time because there was a difference in the instruction on one side. And even if just one of the parameters is wrong, which can be the counterparty, for example, it’s going to take some time to find out and react to it. We report status update ‘unmatched,’ and then clients have to take care of it. It can all be done by straight-through processing, but it requires someone to really take action and correct it,” Loscher says.

But there is still room for improvement. Loscher points out that CSDs are not in a position to mandate auto partials, for example. “Some market participants would argue, ‘Well, with a partial, I have a higher cost for getting the settlement done, because it’s not just one transaction, but multiple transactions.’ And at this point, it is not something where we have a mandate to tell them, ‘You have to do it.’ The mandate we would rather have is to support our clients by offering the necessary tools for their use,” he says.

Loscher adds that any changes to CSDR should be assessed not only in terms of feasibility from an IT perspective, but also in terms of reasonability from a cost-benefit perspective. In that respect, the case for increasing penalties is yet to be demonstrated.

“The big question is what exactly is the benefit of that? Do we really expect that settlement efficiency is going to go up by applying those additional ways of penalizing settlement failures? That’s where we are not convinced,” he says.

The extent of changes to CSDR will not be clear until Esma issues its final report in the summer. But with T+1 now looming large in Europe—it could reportedly be introduced as early as 2026—market participants are worried that the twin blow of increased penalties and a shorter settlement cycle could create a perfect storm for the back office.

“If you’re in double-digit fails on a T+2 settlement cycle,” says Cassells, “and you’re being impacted by penalties for all of these fails, how do you survive a T+1 shift in Europe?”