IHS Markit Rolls Out New Buy-Side Derivatives Risk Modeling Platform

When Mark Findlay, global head of financial risk analytics at IHS Markit, joined the data giant in January, he had three major goals in mind, none of which were tied to the Covid-19 pandemic that has since gripped the world. Now they’re inextricable.

Those goals—to make an immediate improvement to product coverage, leverage financial engineering and critical data across the firm, and fill gaps in the buy-side risk market—have led to the introduction of a new product, dubbed Risk Bureau, aimed at helping buy-side firms calculate and model their risk using alternative data, machine learning, and cloud computing, which are already employed by top sell-side banks.

By leveraging GPUs run on the AWS Cloud and incorporating machine learning, IHS Markit has reduced the time it takes to calculate valuation adjustments (XVAs) for complex and simple derivatives portfolios by 200% to 250%, compared to traditional Monte Carlo models. XVAs include credit valuation adjustments, funding valuation adjustments, collateral valuation adjustments, and capital valuation adjustments.

Where a Monte Carlo simulation uses a forward-looking stochastic process, Risk Bureau works backward with a regression technique. By pre-computing all the different simulated parts with machine learning, users who are plotting and moving data points around on a graph can cut the lag time associated with calculating paths of single lines down to milliseconds.

Beyond looking only at traditional market risk factors and counterparty risk factors, Findlay and his team of risk analysts are trying to understand alternative risk factors as well by overlaying those factors with alternative data points to create what they’re calling a credit default swap (CDS) disruption spread, which acts as a metric that illustrates CDS spreads historically and in potential forecasts. Rather than forecasting from a T-0, or current, standpoint, users will be able to forecast from a future point in time of their choosing—a feature meant to aid companies that want to know not how Covid-19 is affecting them right now, but how it might impact them months down the road.

“In terms of the alternative data, we’re now starting to overlay things like mobility data so that our users can see: What does that mean for a motor company? What does that mean for an airplane? If the transportation is moving differently, how does that affect the credit spread? Is it a leader of the credit spread? Is it a follower of the credit spread? What happens without PMI [Purchasing Managers’ Index] indexing? Is that an early signal?” Findlay says, offering some examples of questions clients may be trying to answer as some jurisdictions begin to lift pandemic-induced lockdowns.

Added to this are new indexes created by the risk business, such as the Covid-19 vulnerability index, which uses an advanced set of algorithms to measure a country’s capability to respond well—and not so well—to the emergence from the pandemic. The analysis part is purposely open and non-prescriptive at this point, as Findlay knows clients will use it in tandem with their own models and methodologies.

“I think the next genesis of this is for us to pick out what resonates with our clients and start to say, ‘Well, how can we pick those risk factors and start to model them in such a way that we can act as kind of a driver for our CDS prediction?’” he says.

Crisis and Beyond

At the core of the package are big data and analytics, which are indispensable in the midst of the pandemic, but Findlay hopes the product will be useful beyond the Covid-19 era. He envisions that clients will want to run the calculations weekly, for example, until further notice, but in some cases, predicts clients may decide to request the tool be run as a managed service, which IHS Markit offers for other platforms across its various product suites.

Risk Bureau is also poised to help buy-side firms coming under scope of the next phase of the uncleared margin rules—phase five of which was recently delayed until September 2021—by helping them understand what their XVA costs are, as well as capital implications involved. Firms affected by the upcoming Fundamental Review of the Trading Book (FRTB), which is slated to go live at the start of 2022, will be able to gauge sensitivities to their top counterparties, run sets of “what-if” scenarios around certain stresses, or get a second market-risk opinion on a tricky derivatives portfolio.

Machine learning will guide the further development of the tool, especially as it starts to include oil pricing and inflation, Findlay says. Currently, it shows a correlation between risk factors on-screen in the credit forecasting utility. Eventually, it will allow users to override that with what their firm decides the correlations are, and machine learning will provide the speed needed to accomplish that, so that the real-time component of the risk model stays intact.

“Everyone is looking for more critical analytics and data, more responses to what’s happening in the markets. Most of us have worked through a number of these different financial crises over the years—whether it was the Russian ruble crisis in 1998, the dot-com bubble bursting in 2000 to 2001, or the financial credit crisis in 2008—and we’ve seen different aspects of lessons to be learned from those, but this is different yet again,” Findlay says. “The thing I certainly thought we would be able to help in is being able to look at really, really big data analytics and being able to bring together things like machine learning to try and make those analytics more readily available and quicker.”