OIS Discounting, FVA, and Basel Rules for Counterparty Risk

Book: Financial Risk Management: Applications in Market, Credit, Asset and Liability Management and Firmwide Risk Authors: Jimmy Skoglund & Wei Chen ISBN: 978-1-119-13551-7

The last chunk of Chapter 5 covers industry changes after 2008 and the regulatory capital framework for counterparty credit risk.

OIS discounting

During the crisis, LIBOR-OIS spreads blew out. LIBOR stopped being a clean proxy for risk-free rates because it embeds bank credit risk.

With collateralized derivatives (and increasingly all derivatives), dealers discount off OIS curves. Credit risk goes through CVA, not through the discount curve. Hull and White argue OIS should be used for all derivatives whether collateralized or not. The floating leg still references LIBOR but gets bootstrapped relative to OIS so swaps price to par.

This split matters for every desk still running legacy LIBOR-discounted systems. Migration to OIS is not a curve tweak. It changes PV, exposure profiles, and hedge ratios.

Advanced CVA and CVA Greeks

Forward simulation works for simple swaps because repricing at future dates is fast. Bermudan and American swaptions need nested simulation or approximations: represent a Bermudan as a European swaption, or use LSMC (least squares Monte Carlo) for continuation values.

Hedging CVA requires sensitivities to CDS spreads and market parameters. Credit deltas are easy under independence: bump CDS spreads, recompute survival probabilities, no resimulation. Market deltas are painful. A 20-point yield curve with +/- bumps means 40 exposure reruns per sensitivity. Adjoint algorithmic differentiation (Capriotti et al.) computes first-order Greeks at roughly 4x the cost of CVA itself, regardless of portfolio size.

Without efficient Greeks, the CVA desk runs blind. You know the charge but not how to hedge it.

Funding Value Adjustment (FVA)

When you post collateral, you fund at your borrowing rate but may earn OIS on received collateral. The spread is real money. FVA sums expected positive exposure times borrowing spread minus expected negative exposure times lending spread, weighted by joint survival probabilities.

Whether FVA belongs in fair value is debated. Hull and White argue against including it. Burgard and Kjaer defend it. Banks with lower funding spreads gain competitive edge if they pass FVA to clients.

Basel counterparty default risk

Counterparty risk has been in Basel since 1988 (current exposure method: mark-to-market plus add-ons). The 1995 amendment allowed netting of add-ons under ISDA agreements.

Basel II (2005) added three exposure methods: internal models, current exposure, and standardized approach. Banks with approval can simulate EAD paths and feed them into the same RWA formula used for banking book credit.

Post-crisis enhancements require:

• Stressed calibration of EAD models
• Backtesting of exposure forecasts against realized exposures
• Alpha factor on expected positive exposure to capture wrong-way risk (regulatory multiplier, typically 1.4, with possible internal model adjustments)

CCAR requires a largest-counterparty-default scenario for trading firms. EBA adds CVA stress from spread moves. Both feed firmwide stress alongside market risk shocks.

Basel III CVA capital charge

Basel III created a separate capital charge for CVA mark-to-market risk (not just counterparty default).

Advanced method: VaR on CVA sensitivities to counterparty CDS spreads, holding exposure profiles fixed. This is the regulatory version of the conditional CVA distribution discussed earlier in the chapter.

Standardized method: formula using exposure at netting set level, credit quality buckets, maturity, and hedge recognition. Easier to implement but less risk-sensitive.

Critics note the charge targets credit spread volatility, not full CVA tail risk from joint market-and-default moves. Banks still need internal models for the unhedged tail.

Mitigating regulatory costs

The chapter closes with practical mitigants: netting agreements, collateral (CSA), central clearing, and CDS hedging. Each reduces EAD, CVA, and capital. Portfolio design (offsetting trades within netting sets) is as important as model sophistication.

My take

This section is the regulatory payoff for everything before it. OIS discounting changed how swaps are priced. CVA capital changed how much equity desks need. FVA changed how traders think about collateral costs.

If you only read one part of Chapter 5 as a manager, read this. It explains why post-2008 derivatives infrastructure looks the way it does.

Previous: CVA Risks in Derivatives Portfolios Next: Liquidity Risk Measurement