Why is DVA (Debit Valuation Adjustment) controversial, and what are the main arguments for and against including own credit risk in derivatives valuation?

DVA is perhaps the most debated concept in derivatives valuation. It represents the expected gain a firm would realize from defaulting on its negative-exposure derivatives. While accounting standards require it, the economic and practical objections are significant.\n\nArguments FOR Including DVA:\n\n1. Theoretical symmetry: In a bilateral framework, both parties' default risk affects fair value. Excluding DVA creates an asymmetry where Firm A's CVA charge to Firm B doesn't match Firm B's DVA benefit from Firm A.\n\n2. Exit price consistency: IFRS 13 and ASC 820 define fair value as the price to transfer a liability. A deteriorating firm could theoretically transfer its derivatives to a lower-credit counterparty at a discount, reflecting DVA.\n\n3. Hedging reality: Dealers can partially hedge DVA by buying CDS protection on themselves (though liquidity is limited). If it's hedgeable, it should be priced.\n\nArguments AGAINST Including DVA:\n\n1. Perverse incentives: Reporting profits from credit deterioration rewards failure. During the 2008 crisis, major banks reported billions in DVA gains while approaching insolvency.\n\n2. Unrealizable gains: A firm cannot monetize DVA except by actually defaulting, which destroys the firm. The \"gain\" exists only in a hypothetical scenario the firm cannot intentionally trigger.\n\n3. P&L volatility: DVA creates earnings volatility uncorrelated with business performance. A credit downgrade produces DVA gains that mask underlying losses, misleading investors.\n\n4. Double-counting: If DVA reduces derivative liabilities, and the firm also issues debt at wider spreads (reflecting the same credit risk), the market has priced the credit risk twice.\n\n`mermaid\ngraph TD\n A[\"Bank credit
deteriorates\"] --> B[\"CDS spread widens
200 → 350 bps\"]\n B --> C[\"DVA GAIN
+$180M reported\"]\n B --> D[\"Funding costs rise
Higher bond yields\"]\n B --> E[\"Stock price falls
Market capitalization drops\"]\n C --> F[\"Paradox: Accounting
shows profit\"]\n D --> G[\"Economic reality:
firm is weaker\"]\n E --> G\n F -->|\"Contradicts\"| G\n style F fill:#ff6b6b\n style G fill:#4ecdc4\n`\n\nRegulatory Treatment:\n\nBasel III resolved this by stripping DVA from regulatory capital. Banks must calculate CVA capital charges but receive no capital benefit from DVA. This creates a permanent gap between accounting P&L (which includes DVA) and regulatory capital.\n\nPractical Compromise:\n\nMost sophisticated dealers track DVA for accounting purposes but manage risk using unilateral CVA. Traders ignore DVA when pricing individual trades, recognizing that incremental CVA is the economically meaningful metric for trade decisions.\n\nThornbury Securities' XVA desk, for example, reports quarterly DVA P&L to satisfy accounting standards but excludes it from trader compensation calculations and risk limits.\n\nExplore the DVA debate in our FRM Part II Credit Risk module.