Jun 4, 2018

Putting Quantitative Innovation into Focus for XVAs

By Dr. Andrew McClelland, Director of Quantitative Research, Numerix

 

Managing xVAs is complex, and that complexity is growing as banks extend xVA to reflect new costs, such as those associated with posting initial margin, or to more-accurately reflect the costs of capital utilization. There is a raft of conceptual difficulties one must navigate when addressing xVAs, such as what should be captured in accounting statements and what should be charged on to clients, etc. Simultaneously, there are enormous computational and modelling challenges to overcome, owing to the sheer difficulty of forecasting quantities such as initial margin requirements or capital impacts.

At Numerix we have a strong focus on finding solutions for xVA problems, and recognize that any such solutions must be flexible and as future-proof as possible. Indeed, derivative markets practitioners regularly identify new trading realities, such as costs, offsets and overlaps for collateral, funding and hedging, that require a tweaking of an xVA framework. Ensuring that this can be done rapidly via simple configuration is key to ensuring our clients can either capture relevant costs or provide more competitive pricing immediately.

This is why quantitative innovation is critical to supporting the financial industry as it confronts challenges in the xVA world. At Numerix, we work continuously to introduce new angles for solving xVA problems, and always leverage our industry-recognized advanced quantitative expertise in doing so.

That’s one reason why I was proud and honored to be asked to chair the xVA stream (a full day of xVA-related presentations and discussions) at the QuantMinds International 2018 conference, held in Lisbon on May 14-18. I was also pleased to be one of the presenters at this conference, attended by about 500 global quants experts from both academia and industry.

My presentation showcased some unique insights into MVA, or Margin Valuation Adjustment, and how algorithmic differentiation can be used to dramatically reduce the costs of MVA calculations. I also highlighted some efficiencies in algorithmic differentiation that can both simplify and reduce the cost of computing CVA Greeks. You can access the presentation below:

AD-on-LSMC for MVA and CVA Greeks: Simplifications and Efficiencies

I enjoyed a number of interesting discussions with my peers at the conference, and welcomed the opportunity to catch up with clients in attendance. We received some great feedback regarding new directions our xVA solutions are taking and I’m looking forward to working together with clients to see our vision through. For me, the event was exciting, collaborative, and fulfilling, and it was motivating to see the degree of innovation in the space of pricing and risk, as well as the potential for moving the industry forward to open up new opportunities.

Here's a short video recapping some highlights from the event:

 

Need Assistance?