Technology vendor of the year: Murex

Risk Awards 2024: In choppy markets, strong R&D work enabled firm to roll out a series of critical modelling and functionality changes

Didier Loiseau, Murex
Didier Loiseau

There’s never really a quiet time for a markets technology vendor – clients always want or need something. But those needs certainly change.

As inflation and interest rates surged over the past couple of years – while the Libor benchmark was also being retired – a whole family of models had to be rethought and, in some cases, replaced. In addition, the new macro environment revived a host of markets and structures that had been becalmed or gave birth to new ones.

It was against this backdrop that Murex stood out for the judges of this year’s Markets Technology Awards, with the Paris-based software company winning three of this year’s categories: best support for risk-free rates; structured products and cross-asset pricing and analytics; and XVA calculation product of the year. Platform, product and features all received their share of praise, but the firm really stood out for its research and modelling work.

Scale helps here. Murex has a team of around 120 quants, which it has added to in recent years.

“We just kept investing, adapting to the level of demand in the market but always keeping a minimum level of investment in order to stay relevant in every asset and product class, and this is the kind of period where that pays off – demand comes back, and it comes back fast,” says Didier Loiseau, global head of trading and financial engineering at Murex.

To develop that first model that deals with fading variance, we had to infer the right volatility level from the Libor vol levels that were quoted at the time. This was what we called the transition model. It was homegrown – there was no standard at all
Didier Loiseau, Murex

Arguably the biggest challenge has been the cascading series of modelling changes required in order to cope with the transition from Libor to new risk-free rate benchmarks (RFRs). For many currencies, that means going from forward-looking rates that are known at the start of a trade’s life to backward-looking rates that compound daily in arrears.

The modelling implications depend on the instrument. Linear products may require a relatively small adjustment.

“Then you get to caps and floors, and that’s where you start to get worried,” says Loiseau.

One of the challenges for options is that as the rate compounds during the observation period, the amount of variance possible gradually declines. Another challenge in the very early days was that there was no market data to use for RFR-linked trades.

“To develop that first model that deals with fading variance, we had to infer the right volatility level from the Libor vol levels that were quoted at the time. This was what we called the transition model. It was homegrown – there was no standard at all,” says Loiseau.

As RFR-linked instruments started trading, Murex was able to incorporate real data into the transition model – with a fresh round of tweaks required to account for the new market quoting conventions that were emerging.

Then in 2019, Andrei Lyashenko and Fabio Mercurio published their extension to the classic Libor Market Model, adapted for backward-looking rates – a Risk.net paper that won them the quants of the year award. Their work became the basis for Murex’s modelling of complex RFR derivatives. But only the basis.

The company still had to go out to its clients, present its approach, and explain the thinking. This was not always straightforward – it turns out that quants at different firms don’t always agree with each other.

“We had arguments with some customers about the modelling. Polite arguments about the right way to do things,” says Loiseau, who adds that at times there could be tension between providing the customer with what they wanted and nudging clients towards an emerging consensus.

“Of course, we’re not here to enforce our views. We’re here to provide solutions for practitioners, so sometimes we needed to make a compromise and follow some of the things that the customers were wanting, even if we didn’t think it was the best,” says Loiseau. “At the end of day, what financial markets are very good at is converging to some sort of consensus. The whole art of the job is to try to go in a direction to help customers while preserving a path to the standard that will emerge.”

EDDE_Maroun-2022
Maroun Edde

In effect, through these conversations, Murex helped to establish that consensus – and that is what the firm’s chief executive, Maroun Edde, expects.

“We closely monitor macro events and anticipate market shifts that rebalance customer preferences. When new RFR payoffs were emerging and XVA analytics needed to be boosted, Murex was ready to accompany 135 customers through the RFR transition, deliver 72 pre-built yield curves and ensure the RFR-readiness of 130 payoffs – also leveraged by all of our XVA customers,” he says.

Remake, remodel

The RFR-modelling push wasn’t always about finding new answers to new questions. For one specific modelling change, Murex was able to dust off work it had started almost a decade ago – further testament to the depth of R&D work at the firm.

The new calibration model was introduced to cope with severe steepening of interest rate curves and to give users greater control of curve shapes.

Much of the background work was already in place – Loiseau recalls first presenting Murex’s approach at an industry conference nearly eight years ago. The firm recognised the idea would apply well to RFR curves, where levels are closely tied to central bank rate-setting meetings and often remain stable between those points.

“It was an idea that we were playing with but never actually released. It’s become super useful in the current context,” says Loiseau.

The firm has also overhauled elements of its XVA functionality over the past 12 months. The backdrop here is the sharp increase in volatility that has accompanied rising rates – volatility that produces bigger swings in exposure sizes, default risk, funding costs and margin calls, and has therefore made these valuation adjustments a more material contributor to bank profit and loss.

In this environment, sophisticated dealers – typically those with existing XVA desks – started calling for greater flexibility when pricing trades. But demand is evolving along twin tracks. While the big dealers are pushing the envelope, a number of domestic banks in Asia and Latin America have been creating XVA capabilities from scratch.

As a result, Murex has seen its client base grow by roughly a third to around 30 clients, says Farah Cherif, head of pre-sales for risk at Murex. Within the firm, a global team of 50 now covers the topic.

In 2022, the team started working with a tier-one US dealer to develop an enhanced XVA pricer – the goal was to give the desk the same kind of powers that traders on other desks enjoy. There was a strong emphasis on being able to quickly calculate pre-deal sensitivities – offering insight into hedging costs for the trade – as well as being able to run a host of what-if scenarios.

“Some critical functionalities they wanted weren’t there [when we first spoke to them], so when we sat together, we listened to what they wanted and we came up with a design,” says Cherif.

Now when a request-for-quote is relayed to a trader, the XVA pricer displays key risk parameters, such as credit spreads, funding spreads, collateral and counterparty ratings. Traders can then adjust risk factors, conduct on-the-spot scenario analysis and measure any XVA cost alteration. Over the course of 12 months, the company also added pre-deal sensitivity tools and combinations of pricing.

Cherif says this has taken the “product to the next level”. A process that the company aims to continue across its products in the years to come.

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here