Capital Calculations For a Million Trades Takes Seconds

Successful proof-of-concept of Markit FRTB Studio at European bank enables firm to quickly understand how business decisions affect all FRTB capital measures

Markit FRTB Studio, an aggregation and capital analytics engine, has processed a portfolio of one million trades in single-digit seconds on a European bank’s Hadoop infrastructure as part of a successful proof-of-concept. This high-speed performance enables banks to understand the capital implications of business decisions in near real time in relation to the FRTB standards.

Part of IHS Markit’s broader FRTB solution, FRTB Studio is built on big data technology and can be used as a standalone component, integrating easily with banks’ existing risk management infrastructure. The European bank was up and running on FRTB Studio in just two days. Front office and risk teams are using the aggregation and capital analytics engine to assess – almost instantly – how decisions relating to P&L attribution (PLA), desk restructuring, and modelability impact their capital charges.

“The financial industry stands at the next frontier of adopting emerging technologies around big data, artificial intelligence and machine learning” said Cubillas Ding, Research Director, Finance and Risk Practice at Celent. “All of these technologies, in the context of FRTB, can result in augmented delivery of risk analytics, stronger data supply chains and innovative IT models at performance and cost frontiers previously not possible.”

“Markit FRTB Studio can calculate the capital charges for a Standard and Internal Model Approach (SA and IMA) in mere seconds based on a portfolio of one million trades” said Andrew Aziz, Global Head of Markit Analytics at IHS Markit. “We are able to integrate FRTB Studio with banks’ existing valuation engines, enabling firms to perform all FRTB capital calculations at scale and speed, and most importantly, on a highly cost-effective technology.”

As firms look to manage growing costs, Markit FRTB Studio enables banks to reuse existing valuation engines, either homegrown or vendor supplied, to produce the required inputs for consistent aggregation and capital calculation, without having to build that capability in-house.

Benchmarks show that FRTB Studio can support the capital charges for IMA and SA in single-digit seconds based on a portfolio of one million trades with nine million sensitivities, 18 million P&L vectors, and on hardware costing just USD20k.

FRTB Studio can be implemented as a single component or as part of IHS Markit’s broader solution for FRTB.

Related Posts

Misys Supports Banks with FRTB Compliance Misys is supporting the banking community as it prepares for Basel’s Fundamental Review of the Trading Book (FRTB). The regulation, which overhaul...
Ed Thorp: How Did You Do It? By Nicholas G. Colon A few times in every generation a person comes along that makes a significant impact on the world.  Rare or lucky are those th...
Julia Seminar for Quants in London 2017 is turning out to be the year that Julia moves from early adopters within the finance community into the mainstream Last week’s Julia in Fina...
Julia In Finance Seminar 16 March, 6.00pm - 9.00pm GMT Fitch Learning, 55 Mark Lane, London, EC3R 7NE In partnership with Julia Computing, the CQF Institute invites yo...
Jobseekers, Welcome to the Wilmott Jobs Board Welcome to the Wilmott Jobs Board! If you are a job seeker then to have full access to job-seeking functionality you must first Register on wilmo...
Julia Enters Top 50 Programming Languages   Julia, a numerical computing language, took its place among the top 50 computer programming languages in September 2016, according to th...
Mathematica: Going To Eleven Stephen Wolfram’s recent blog post on the launch of Mathematica and Wolfram Language 11 presents a number of innovations that Quants will find worth e...
Capital Calculations For a Million Trades Takes Se... Markit FRTB Studio, an aggregation and capital analytics engine, has processed a portfolio of one million trades in single-digit seconds on a European...