Alethograph

xVA hedging strategies

6 min read

xVA hedging encompasses the strategies used by an xVA desk to manage the mark-to-market volatility of valuation adjustments across a derivatives portfolio. The fundamental challenge is that xVA depends on a high-dimensional set of risk factors spanning every asset class, credit spreads, funding costs, and capital requirements, many of which are illiquid or unhedgeable. The general representation of xVA sensitivities (Equation 21.1 in Gregory) decomposes changes into three components: sensitivity to credit spreads (dxVA/dS), sensitivity to exposure-driving market variables (dxVA/dE), and cross-gamma terms (d^2 xVA / dS dE), plus time decay (theta). Market risk is hedged in aggregate across xVA terms and across counterparties, not at the individual counterparty level.

Hedging instruments fall into four categories of decreasing practicality: (1) liquid hedges (interest rate and FX deltas), which are routinely hedged by most banks; (2) illiquid hedges (volatility via swaptions and FX options), hedged selectively for major currencies; (3) proxy hedges (credit indices for credit delta, out-of-the-money options for tail risk), which provide partial protection subject to basis risk; and (4) no hedge (correlations, cross-gamma), which are typically unhedgeable except via bespoke instruments. The beta hedging framework quantifies the efficiency of proxy hedges: for a single counterparty hedged with a 50%-correlated index, the residual standard deviation is 87%, but for a portfolio of 50 equally-weighted counterparties the residual drops to approximately 23%, making index CDS hedging materially effective at the portfolio level.

Cross-gamma — the joint sensitivity to simultaneous moves in two risk factors — is a distinctive and dangerous feature of xVA portfolios. It is typically realised after significant correlated market events (e.g., the 2016 EU referendum caused GBP rates tightening and credit spreads widening simultaneously, creating losses for delta-hedged CVA books). Cross-gamma can be partially mitigated by direct hedges (e.g., CDS in two currencies), over/underhedging deltas based on known directionality, or buying out-of-the-money options. Jump-to-default (JTD) risk — the impact of a sudden counterparty default on a delta-hedged position — depends on current exposure, CDS hedge notional and maturity, and xVA release. JTD risk can only be managed by adjusting the credit spread delta by tenor.

The regulatory capital implications of xVA hedges are critical. Under current and future (FRTB-CVA) rules, CVA hedges and non-CVA hedges are treated separately: eligible CVA hedges (single-name CDS, index CDS, and under SA-CVA most market risk hedges) reduce CVA capital, while ineligible hedges must be captured in the trading book market risk framework, where they may increase capital. This creates a perverse incentive where hedging accounting xVA volatility can increase total regulatory capital. Under the future SA-CVA methodology, most CVA hedges will be eligible, better aligning accounting and regulatory incentives.

Key Details

  • xVA Greeks for a single cross-currency swap include: interest rate risk (two currencies), basis risk, interest rate vega (two currencies), FX spot, FX vega, cross-currency basis, credit delta, gamma/JTD, and cross-gamma
  • Market risk hedges are most frequently hedged (liquid instruments). Interest rate futures are cheapest for directional risk; swaps are liquid up to 10 years in major currencies
  • Credit CS01 hedging uses index CDS (liquid, spread hedge only, no JTD protection), proxy single-name CDS (limited liquidity, partial JTD), or single-name CDS (very limited liquidity, full JTD). Banks generally prefer to warehouse credit risk due to inherent risk premiums
  • Beta hedging: optimal hedge ratio driven by correlation between counterparty exposure and hedging instrument. Residual variability = sqrt(1 - rho^2) for a single name; for a portfolio of n equal counterparties, idiosyncratic risk diversifies away, making index hedging significantly more effective (Figure 21.5)
  • JTD P&L = -(current exposure x assumed LGD) + (CDS notional x assumed CDS LGD) - (current xVA contribution) (Equation 21.2)
  • P&L explain decomposes daily xVA changes into theta, rates delta, FX delta, credit delta, vega, gamma, cross-gamma, defaults, funding/capital cost changes, portfolio changes, and counterparty changes
  • xVA hedges interact with regulatory capital: under current rules, only single-name and index CDS are eligible for CVA capital relief; market risk hedges for CVA are ineligible and may increase trading book capital. Under future SA-CVA, most hedges become eligible
  • Table 21.9 provides a detailed comparison of hedge eligibility under current (standardised, advanced) and future (BA-CVA, SA-CVA) regulatory regimes
  • For KVA hedging (Section 21.3.2), three methods of releasing KVA profit are compared: Method A (immediate release, variable ROC), Method B (scheduled release, arbitrary), Method C (full KVA management with hedging, locks in ROC at cost of variable P&L). Figure 21.10 shows KVA hedge P&L in different market scenarios
  • The optimal index CDS hedge for capital relief depends on the spread regime: at low spreads, hedging improves ROC; at high spreads, the cost of protection exceeds the capital benefit (Figure 21.8)

Textbook References

The xVA Challenge (Gregory, 2020)

  • Section 21.2.1 (pp. 619—621): Overview of xVA hedging. Market risk on derivatives is hedged by trading desks; xVA market risk is hedged separately by the xVA desk. xVA is analogous to managing a book of options (Sorensen-Bollier analogy). Changes in funding/capital costs are largely unhedgeable.
  • Section 21.2.2 (pp. 621—624): xVA sensitivities (Equation 21.1): dxVA/dS (spread), dxVA/dE (exposure), dxVA/dt (theta), d^2 xVA/dSdE (cross-gamma). Table 21.4 gives Greeks for a 100M receive-fixed 10Y swap across CVA, DVA, FCA, FBA. AAD can compute all Greeks at a fixed ~4x cost (Section 21.3.3).
  • Section 21.2.3 (pp. 625—627): Gamma, cross-gamma, and tail risk. Figure 21.2 shows CS01 varying with interest rate moves. Cross-gamma is realised during correlated market events (e.g. EU referendum: GBP rate tightening + credit spread widening). Mitigation: direct hedges, over/underhedging deltas, buying OTM options.
  • Section 21.2.4 (pp. 627—628): Market practice. Figure 21.3: market risk delta is fully hedged by most banks; credit spread delta is discretionary; vega, gamma, cross-gamma are rarely hedged. Table 21.6 summarises hedging practicality and practice. Most banks hedge CVA market risk jointly with FVA.
  • Section 21.2.5 (pp. 629—630): JTD risk. Equation 21.2 gives the JTD P&L formula. Figure 21.4 shows JTD P&L as a function of credit curve shift for 5Y vs. 10Y CDS hedges. JTD can only be hedged by adjusting credit delta by tenor.
  • Section 21.2.6 (pp. 630—631): Beta hedging. Residual variability = sqrt(1 - rho^2) for single name. Figure 21.5 shows portfolio benefit: 50 counterparties at 50% correlation yields ~23% residual variability. Beta estimation uses historical relationship between xVA volatility and index.
  • Section 21.2.7 (pp. 631—633): Risk limits and P&L explain. Table 21.8 gives an example P&L explain decomposition. UX (unexplained P&L) should be small relative to daily P&L; if not, additional sensitivity calculations are needed.
  • Section 21.2.9 (pp. 634—637): Impact on capital. Figure 21.8 shows ROC as a function of index CDS hedge for different spread regimes. Table 21.9 details hedge eligibility under current and future CVA capital rules. Non-eligible hedges (FVA market risk hedges) consume trading book capital. Future SA-CVA will better align accounting and regulatory hedge incentives.
  • Section 21.3.2 (pp. 640—641): KVA hedging. Table 21.10 compares three methods for releasing KVA profit: immediate (Method A), scheduled (Method B), full management with hedging (Method C). Method C locks in ROC but with variable P&L from KVA hedges (Figure 21.10).
  • Section 21.3.4 (pp. 645—647): xVA optimisation. Portfolio compression, restrikes (resetting ITM trades to ATM), CSA renegotiation, voluntary bilateral margining, and backloading to CCPs. Regulatory arbitrage can arise from converting ITM derivatives into ATM derivatives plus loans.

concept

Graph View

Backlinks