MVA Using Algorithmic Differentiation

At the end of 2018, the financial industry has still not yet established a consensus methodology for calculation of Margin Value Adjustment (MVA) on non- centrally cleared derivatives. MVA represents the expected funding cost of initial margin over the lifetime of a trade/portfolio, and is particularly relevant today due to BCBS-IOSCO 261 – more commonly referred to as the “Swaps Margin Rules”, it requires financial entities to exchange sufficient collateral to cover potential losses over a 10-day period with 99% confidence, and is complementary to the margin typically used to settle daily mark-to-market changes (variation margin), phasing-in to cover most derivatives market participants by September 2020. MVA is the most recent valuation adjustment (“xVA”), joining similar calculations for counterparty credit risk (CVA), the funding cost of variation margin (FVA), and the cost of capital (KVA), among others. MVA is particularly difficult to calculate due to the requirements for trade sensitivities along each Monte Carlo simulation path as inputs to ISDA’s Standard Initial Margin Model (SIMM). AD provides the most efficient and robust calculation of these in-simulation sensitivities.

Download the Whitepaper