LONDON, May 26 (IFR) - Post-trade processing firm TriOptima has expanded its triCalculate counterparty credit risk analytics service to include margin valuation adjustment (MVA) calculations.
The new tool enables swaps counterparties to more accurately calculate the long-term cost associated with new rules that require collateral to be posted against over-the-counter derivatives that are not centrally cleared.
The MVA tool adds to a range of capital-related swaps valuation charges already live on the web-based service, including CVA, DVA, FVA and potential future exposure. Capital valuation adjustment (KVA) analytics will be added over summer.
The platform, which went live for testing 18 months ago and is now being used by a handful of clients, enables swaps counterparties to conduct full analysis of credit, capital, and funding valuation adjustments across their derivatives portfolios. Those costs have become increasingly complex in the face of stringent capital requirements and sweeping derivatives reforms, such as uncleared margin rules, which went live for the first wave of dealers in the US and Japan in 2016 and in Europe earlier this year. Additional market participants will be brought into scope in waves by 2020.
“With the margin requirements coming into play for an increasing number of market participants, they are now incurring a cost that they didn’t previously and most of them don’t have a mechanism for accurately calculating the cost of the lifetime initial margin of a trade,” said Mireille Dyrberg, COO of TriOptima. “With triCalculate, clients get far more clarity on the drivers of cost and risk in their portfolio and therefore are better able to manage them.”
While KVA analytics were initially at the top of the development plan for 2017, reflecting interest from German bank clients, MVA leapfrogged the timeline following demands from Japanese banks that had already gone live with initial margin on their bilateral portfolios.
While large dealer banks typically run their own internal systems and models, triCalculate is being viewed for scenario analysis and validation purposes. Second tier banks are viewing the platform as a cost-effective primary system for managing counterparty risk on the CVA desk, or for sales desks to calculate the cost, risk and margin before offering prices to their swaps clients.
Corporates are also beginning to test the service for pricing transparency and determine whether it would be advantageous to enter into a new trade with one dealer versus another.
“Historically, corporates have been in a situation where they are having to rely on banks for a lot of information so we’re trying to empower them to do these calculations themselves," said Tom Griffiths, business manager for triCalculate. “It’s really all about transparency and giving our clients all the tools they need to achieve this.”
The triCalculate service is the NEX Group-owned firm's first foray into analytics. It uses matrix probability methodology - technology designed for the gaming industry – to achieve greater pricing precision. The technology can incorporate over 100,000 simulations – more than 20 times the number that would go into a typical Monte Carlo simulation exercise.
A two-step process runs client portfolios through a graphics processing unit to establish the potential value at all points in a time series. Unlike a more commonly used central processing unit, which can handle six-to-eight extremely complex tasks in parallel, a GPU is capable of processing a large quantity of simple addition and multiplication tasks at once.
The GPU step calculates vast sheets of data to create a look-up table that is referenced by the CPU Monte Carlo simulation in the second stage.
A traditional CPU-based simulation would typically limit the number of simulations to 1,000-5,000, resulting in lower pricing accuracy. (Reporting by Helen Bartholomew)