XVA Explained: CVA, DVA, FVA, MVA, and KVA — The Complete Guide

What XVA Is and Why It Matters

Before the 2008 financial crisis, most OTC derivatives were priced using a single risk-free curve — counterparty default risk, funding costs, and capital were treated as secondary considerations that didn't belong in the mark-to-market. The crisis exposed how wrong that assumption was. Lehman Brothers' default on a massive derivatives book caused bilateral losses that had not been priced in. The cost of posting collateral became visible as funding spreads exploded. Regulatory capital requirements for counterparty credit risk multiplied under Basel III.

The industry responded by developing a unified framework of adjustments that bring these real costs back into derivative fair value. Each adjustment is an "XVA" — a valuation adjustment with its own letter prefix. Together they represent the difference between a risk-free theoretical derivative price and the full economic cost of holding the trade. For a large bank dealer, the aggregate XVA book is measured in hundreds of millions of dollars and is managed centrally by a dedicated XVA desk that prices, hedges, and reports on all adjustments.

This article covers each XVA component with precise formulas, the regulatory framework governing it, and the practical mechanics of how XVA desks use Claude for calculation, explanation, and hedging analysis. For XVA desk workflows using Claude, see also Claude AI for XVA and CVA Trading Desks.

CVA: Credit Valuation Adjustment

CVA is the market value of counterparty default risk embedded in an OTC derivative. It is the amount by which the risk-free derivative value must be adjusted downward to account for the possibility that the counterparty defaults before maturity, leaving an unrecovered positive exposure. Under IFRS 13 (effective 2013) and ASC 820, CVA is a required component of derivative fair value for financial reporting.

The standard formulation for unilateral CVA (assuming only the counterparty can default) is:

CVA ≈ LGD × ∫₀ᵀ EE(t) × PD(t, t+dt) dt

Where EE(t) is the expected exposure at time t (the expected positive MTM of the netting set, conditioned on it being positive), PD(t, t+dt) is the marginal probability of default in the interval [t, t+dt] derived from CDS spreads or credit ratings, and LGD is the loss given default (typically 40-60% for senior unsecured counterparties). In discrete form: CVA ≈ LGD × Σ EE(tᵢ) × PD(tᵢ₋₁, tᵢ) where the sum runs over quarterly intervals through maturity.

Bilateral CVA extends this to include DVA: Bilateral CVA = CVA − DVA. This makes the adjustment symmetric — both counterparties incorporate each other's default risk into fair value. The bilateral formulation is the IFRS 13 / ASC 820 standard; unilateral CVA is used for regulatory capital (Basel III CVA capital charge does not allow DVA offset).

Wrong-way risk is the critical complication: when the counterparty's probability of default is positively correlated with the exposure (e.g., an energy producer selling oil futures to you is most likely to default precisely when oil prices collapse and the contract is most in the money), standard CVA underestimates the true risk. Wrong-way risk requires simulation of joint default-exposure scenarios or specific stress multipliers.

• "CVA calculation for an IRS netting set: Counterparty is a BBB-rated industrial corporate with a 5Y CDS spread of 90bps flat and LGD = 60%. Netting set: (1) pay-fixed 5Y IRS, $50M notional, current MTM +$1.8M; (2) pay-fixed 7Y IRS, $30M notional, current MTM +$2.4M; (3) receive-fixed 3Y IRS, $20M notional, current MTM -$0.6M. Net positive exposure today: $3.6M. No CSA (uncollateralized). Using unilateral CVA: (1) extract marginal PDs from the 90bps flat CDS curve assuming LGD 60% — calculate quarterly PDs for 7 years; (2) construct an expected exposure profile for this netting set assuming linear amortization of each swap's exposure from current MTM to zero at maturity; (3) compute CVA as LGD × sum of (EE(t) × PD(t, t+dt)) over quarterly intervals; (4) express CVA as a dollar amount and as a running spread in bps on the $50M 5Y swap notional; (5) recalculate CVA if CDS spreads widen from 90bps to 150bps — show the CVA sensitivity to a 60bps credit spread widening."
• "Wrong-way risk analysis for an FX position with a commodity counterparty: We have a 2Y USD/BRL FX forward where we receive BRL and pay USD — $40M notional, currently in our favor by $2.1M. The counterparty is a Brazilian agri-exporter with CDS at 220bps (LGD 65%). Wrong-way risk: BRL typically depreciates during EM stress episodes that also widen Brazilian corporate credit spreads — so our exposure grows exactly when the counterparty is most stressed. Explain: (1) how wrong-way risk affects the standard CVA formula, (2) quantify the wrong-way risk multiplier if the correlation between BRL depreciation and counterparty CDS widening is +0.55, (3) what haircut should we apply to the close-out value to account for wrong-way risk in our CVA reserve, (4) how does BCBS guidance on wrong-way risk (Basel III Annex 4) require us to handle this?"

DVA: Debit (Debt) Valuation Adjustment

DVA is the mirror of CVA: it is the value of the bank's own default risk from the counterparty's perspective, applied to the bank's derivative liabilities. Where CVA reduces the asset value of a positive-MTM trade (because the counterparty might default), DVA increases the value of a negative-MTM trade (because the bank itself might default before paying). Bilateral fair value = Risk-free MTM − CVA + DVA.

The accounting logic is consistent but the economic interpretation is contentious. A bank with deteriorating credit sees its DVA increase — effectively booking a gain from its own creditworthiness declining. IFRS 13.42 requires that the fair value of a financial liability reflect the credit risk of the entity, which produces this outcome. The "DVA paradox" — a firm profits on its own approaching insolvency — prompted intense criticism from regulators and investors who saw it as obscuring true financial performance.

The regulatory response under Basel III: DVA gains are not recognized in CET1 capital (CRR Article 33 in Europe, corresponding US rules). A bank can book DVA in P&L for accounting purposes but must deduct any DVA gains from Tier 1 capital. This creates a permanent divergence between accounting fair value and regulatory capital treatment.

• "DVA calculation and regulatory capital treatment: Our bank has a 5Y EUR IRS portfolio where we are net payers — we owe $8.4M MTM to our counterparties on aggregate (from the counterparty's perspective, they hold a $8.4M positive exposure to us). Our own 5Y CDS spread: 55bps. LGD for senior bank debt: 45%. Expected loss on our own obligations: calculated symmetrically to CVA but using our own default probability. Calculate: (1) DVA on our net derivative liability of $8.4M using our 55bps CDS spread and LGD 45%, (2) the DVA as a dollar amount and as a bps adjustment on the aggregate notional, (3) explain the IFRS 13 basis for including DVA in our fair value hierarchy Level 2 derivative valuations, (4) how we account for the divergence between our accounting DVA gain and the Basel III requirement to deduct DVA from CET1, (5) what happens to our DVA if our own CDS spreads widen from 55bps to 90bps?"

FVA: Funding Valuation Adjustment

FVA captures the cost (or benefit) of the funding mismatch created by an uncollateralized derivative. When a dealer bank has an uncollateralized trade with a corporate client and hedges it in the interbank market under a CSA, the hedge gains cash collateral when it moves in the money — but the client trade generates no collateral inflow, creating a funding gap. The dealer must fund that gap at its own unsecured funding cost (SOFR + bank funding spread), which is the FVA.

FVA has two components: FCA (Funding Cost Adjustment) — the cost of funding positive exposure on uncollateralized trades where the dealer funds itself at above-risk-free rates — and FBA (Funding Benefit Adjustment) — the funding benefit when the dealer receives cash from a CSA-collateralized hedge against a negative-MTM uncollateralized client trade. Net FVA = FCA − FBA.

The academic debate is notable: Hull and White (2012) argued that in a Modigliani-Miller frictionless world, FVA should be zero because the funding cost is already reflected in firm value and double-counting it distorts prices. Practitioners universally rejected this — in a world with regulatory constraints on capital, asymmetric funding access, and real desk-level transfer pricing, FVA is a genuine cost that must be priced and hedged. All major dealer banks began booking FVA in the 2012–2014 period.

• "FVA calculation for uncollateralized USD IRS with a corporate: We have a 5Y pay-fixed USD IRS, $75M notional, with an unrated mid-size corporate (no CSA). The trade is ATM today. We hedge it with a receive-fixed 5Y IRS under a bilateral ISDA/CSA with a primary dealer (zero threshold, daily margining). Our internal funding cost: SOFR+40bps. As rates move and our hedge becomes positive MTM, we receive cash from the CSA but cannot pass it to the corporate client. The average positive exposure on the hedge side over the 5Y life is estimated at $1.8M (expected positive exposure, EPE). Calculate: (1) FCA on this trade — annual funding cost on average uncollateralized exposure, (2) FVA as a present value using a flat 40bps funding spread applied to the exposure profile at SOFR discount, (3) FVA as a running spread in bps on the $75M notional, (4) should this FVA be charged upfront to the client or included as a spread in the fixed rate? What are the commercial implications of each approach, (5) if we can execute the hedge with a CCP-cleared IRS instead of bilateral, what happens to the FVA?"
• "FVA reporting for P&L decomposition: Our FVA book had a daily P&L of -$340,000 yesterday. Break down this move across the following drivers: (1) FVA theta (daily time-decay of FVA as trades age one day and expected exposures roll down), (2) funding spread movement — our 5Y senior unsecured spreads widened 3bps intraday from 38bps to 41bps, (3) rates movement — 5Y USD SOFR moved +8bps, shifting expected exposure profiles on our uncollateralized IRS book, (4) new trades — we added $100M notional of 3Y uncollateralized IRS with a new corporate client, (5) trade maturities — two short-dated FX forwards expired. Estimate the contribution of each factor and identify what drove most of the -$340k move."

MVA: Margin Valuation Adjustment

MVA is the present value of the cost of posting initial margin (IM) over the life of a trade. It emerged as a material XVA component with the phased rollout of the BCBS/IOSCO Uncleared Margin Rules (UMR), which mandated bilateral initial margin for uncleared OTC derivatives between covered entities. Phase 1 (September 2016) covered the top dealers by AANA (aggregate average notional amount); Phase 6 (September 2022) captured entities with AANA above €8 billion.

For trades subject to UMR or cleared through a CCP, the dealer must post IM that it cannot use for other purposes — the opportunity cost of that tied-up capital is MVA. The formula:

MVA = ∫₀ᵀ [FundingSpread × E[IM(t)]] × P(τ > t) dt

Where E[IM(t)] is the expected initial margin at future time t (typically modeled using ISDA SIMM or CCP IM model sensitivities), FundingSpread is the dealer's marginal cost of funding, and P(τ > t) is the survival probability. For long-dated trades with significant interest rate or credit sensitivity, MVA can exceed CVA in magnitude — a 10-year IRS with $45,000 DV01 might generate several hundred thousand dollars in SIMM-based IM, the funding cost of which is the MVA.

MVA optimization is now a standard desk function: comparing cleared vs. bilateral execution, portfolio compression to reduce IM, and SA-CCR netting set optimization to minimize EAD (and through it, initial margin at some CCPs).

• "MVA calculation for a 10Y EUR IRS subject to UMR: New trade: receive-fixed 10Y EUR IRS, €80M notional, ATM at 3.15%. Counterparty is a large asset manager subject to UMR Phase 5. Bilateral IM calculated under ISDA SIMM. Estimated SIMM IM for this trade: €2.4M (based on IR delta risk weight for 10Y EUR bucket × adjusted notional). Our funding cost for IM: EURIBOR+35bps. Trade life: 10 years. (1) calculate MVA using the simplified formula: MVA ≈ FundingSpread × average SIMM IM over life × trade tenor, assuming SIMM IM decays linearly from €2.4M to zero as the trade approaches maturity, (2) express MVA as a bps spread on the €80M notional over 10 years, (3) compare: if this trade is executed via LCH SwapClear instead, IM is calculated under SwapClear's PAIRS model — assume LCH IM for this trade would be €2.9M. Calculate the MVA under clearing and compare to bilateral, (4) what does this MVA comparison say about whether to clear or execute bilaterally, ignoring other factors?"

KVA: Capital Valuation Adjustment

KVA is the present value of the stream of regulatory capital costs that a trade locks up over its lifetime. Every OTC derivative generates regulatory capital requirements — for counterparty credit risk (SA-CCR EAD × RWA × 8% × capital surplus), for CVA capital (under Basel IV BA-CVA or SA-CVA), and potentially for market risk (FRTB SA or IMA). KVA quantifies the cost of holding that capital.

The standard formulation:

KVA = ∫₀ᵀ CoC × E[K(t)] × P(τ > t) dt

Where CoC is the cost of capital (typically the difference between the bank's target return on equity — say 12-15% — and its risk-free funding cost), E[K(t)] is the expected regulatory capital requirement at future time t generated by the trade, and P(τ > t) is survival probability. For a 5Y IRS, K(t) includes SA-CCR EAD × 8% capital ratio × RWA density, plus the Basel IV CVA capital charge contribution of that trade.

KVA is the most controversial XVA component. The theoretical argument for including it is that regulatory capital is a scarce resource allocated at a cost; a trade that consumes capital for 5 years imposes a real economic cost that should be priced in. The counterargument: capital costs are a firm-level cost of doing business, not a trade-level charge; double-counting capital costs inflates derivative prices and disadvantages dealer banks versus less-regulated entities. In practice, large dealer banks include KVA in internal pricing but not always in the initial price shown to clients — it is more commonly reflected in hurdle rates and desk capital allocation.

• "KVA pricing for a new 7Y cross-currency swap: New trade: 7Y USD/EUR cross-currency basis swap, $50M notional, paying SOFR flat vs receiving EURIBOR-8bps. Counterparty: large European corporate, BBB rated, no CSA. SA-CCR EAD for this trade: estimated at $3.8M (RC $0.8M + PFE $3.0M, using FX supervisory factor 4% × maturity factor × adjusted notional, then × α=1.4). CCR RWA: EAD × risk weight 100% (BBB unrated corporate) = $3.8M RWA. CVA capital (BA-CVA, simplified): approximately $0.38M capital. Total capital required for this trade: $0.38M (CVA capital) + $0.30M (CCR capital at 8%) = $0.68M. Cost of capital: 13% return on equity target, risk-free rate 4.5%, so CoC = 8.5%. (1) calculate KVA as the present value of CoC × capital requirement over 7 years, assuming capital declines linearly as exposure rolls off, (2) express KVA as a running spread in bps on the $50M notional, (3) decompose KVA into CCR capital component and CVA capital component, (4) how does the KVA compare to the CVA charge on this trade (use BBB 100bps CDS, LGD 60%)?"
• "KVA sensitivity analysis for a desk-level capital review: Our rates derivatives desk runs $15B notional of OTC IRS with a mix of CSA and non-CSA counterparties. Current SA-CCR EAD (non-CSA book): $185M. CVA capital (BA-CVA): $28M. CCR RWA: $1.48B. Cost of capital target: 12% RoE, current funding cost SOFR+60bps (approximately 5.6%), so CoC = 6.4%. (1) estimate the annualized KVA cost for the desk's non-CSA book, (2) if we can novate 20% of non-CSA trades to centrally cleared LCH (reducing EAD by 35% due to netting), what is the KVA saving, (3) if the BA-CVA charge increases by $5M following a regulatory review, what is the KVA impact, (4) at what level of CoC does the KVA on a new non-CSA IRS exceed the FVA, making capital cost the dominant pricing driver?"

ColVA: Collateral Valuation Adjustment

ColVA captures the value of optionality embedded in credit support annexes (CSAs). Many CSAs do not specify a single collateral currency — they specify a set of eligible collateral (typically G10 currencies, high-quality government bonds) and allow the posting party to choose which eligible asset to post at each margin call. This optionality to post the "cheapest to deliver" (CTD) collateral has positive value to the poster and negative value to the receiver. ColVA is the present value of that optionality.

ColVA is material when CSAs are old, drafted before the era of systematic XVA pricing, and allow unusual collateral. A CSA permitting posting in USD, EUR, or GBP government bonds, for example, gives the posting party the option to post whichever currency is currently cheapest relative to the OIS discounting currency — and those currency bases (the cross-currency basis between OIS rates) move meaningfully. For a 20-year IRS under a multi-currency CSA, the ColVA can run to several hundred basis points of notional.

• "ColVA analysis for a multi-currency CSA: We have a 15Y USD IRS portfolio with a European sovereign wealth fund under a CSA permitting collateral in USD cash, EUR cash, or UK Gilts. The USD/EUR cross-currency basis (3M SOFR vs 3M EURIBOR) is currently -18bps (EUR borrowing costs more than USD, so EUR collateral is cheaper to post). The USD/GBP basis is -12bps. The expected collateral balance over the life of the trades (net MTM projection under current rate curve): $45M average positive exposure to us, meaning the counterparty will be posting collateral. The counterparty will optimize to post EUR cash when EUR/USD basis is most negative. (1) explain the mechanics of the CTD optionality in this CSA and who holds it, (2) estimate ColVA on a simplified basis: if the counterparty can save 18bps per year by posting EUR vs USD and the average collateral balance is $45M, what is the present value of that saving over 15 years? (3) how should this ColVA affect our discount curve for this netting set — should we use USD OIS or a blended discount rate? (4) what is the market practice for booking ColVA?"

The XVA Desk: Managing the Full Framework

In a large dealer bank, all XVA components are managed centrally by an XVA desk (also called the CVA desk or counterparty risk trading desk). The XVA desk does not take proprietary positions; its job is to price XVA into trades before they are executed, hedge the market risk of XVA positions (credit spread delta via CDS, rates delta via IRS, FX delta via FX forwards), and manage the regulatory capital implications of the aggregate book.

The XVA desk interacts with front office in a specific way: when a rates trader wants to execute a $100M uncollateralized 10Y IRS with a BBB corporate, the XVA desk calculates the CVA, FVA, MVA, and KVA for that trade and charges the rates desk as an internal cost. The rates desk then bids the client price adjusted for XVA — the XVA charge is why an uncollateralized trade is always priced wider than a CSA-collateralized equivalent. Front office traders who don't understand XVA pricing either undercharge clients (absorbing costs themselves) or over-charge and lose business to dealers with better XVA infrastructure.

• "XVA pricing for a new trade — front office briefing: A rates trader has booked a new 8Y pay-fixed USD IRS, $80M notional, with an AA-rated insurance company that has a one-way CSA (they post collateral to us if we're in the money, but we don't post to them if they're in the money). XVA charges for this trade: CVA $45,000, FVA $22,000 (we can't call collateral from them when trade is negative MTM so we must fund that exposure), MVA $0 (one-way CSA, they post to us — we receive IM when trade is positive), KVA $18,000. Total XVA charge = $85,000. (1) explain to the trader in plain language why a one-way CSA still generates FVA and CVA, (2) why does a one-way CSA where we receive collateral eliminate our MVA but not our CVA, (3) convert the $85,000 XVA charge into a running spread on the 8Y IRS — what does this mean for the all-in pricing of this trade vs. a fully collateralized bilateral trade, (4) the trader says a competitor is quoting the same IRS 2bps tighter — what XVA assumption could explain the difference?"
• "XVA hedging strategy for a CVA book: Our CVA book has the following aggregate sensitivities as of today: CS01 (credit spread 01 — change in CVA for a 1bp widening in all counterparty CDS): +$420,000 (long credit risk — we lose if spreads widen). IR DV01 (change in CVA for a 1bp parallel rate shift): -$85,000. FX delta on EUR counterparties (change in CVA for 1% EUR/USD move): +$28,000. Current hedges: $500M notional of 5Y CDX.IG (iTraxx Europe equivalent) protection. (1) calculate our net CS01 after the CDX hedge (CDX DV01 approximately $500 per $1M notional for 5Y contract), (2) is our CDX hedge sufficient, over-hedged, or under-hedged on credit spread delta? (3) what additional single-name CDS positions should we consider given the counterparty breakdown (assume 60% IG financials, 25% IG industrials, 15% HY), (4) how should we hedge the IR DV01 — what IRS tenor and notional? (5) comment on whether the FX delta is worth hedging given transaction costs."

XVA Interaction Effects and the "XVA Debate"

XVA components interact in ways that make the aggregate adjustment non-additive. FVA and CVA overlap for uncollateralized trades — the funding cost of carrying positive exposure includes a credit component (if the counterparty defaults, you lose the exposure you were funding). DVA and KVA interact because DVA is excluded from CET1 but affects P&L that then drives capital ratios. MVA reduces CVA (because margin reduces expected exposure) but increases the capital cost of posting IM.

The practitioner consensus has settled on a hierarchy: CVA and DVA are required by accounting standards (IFRS 13/ASC 820) and are non-negotiable. FVA is economically justified and universally adopted by dealer banks. MVA became mandatory once UMR made initial margin costs real and unavoidable. KVA is adopted by most large dealers for internal pricing but remains debated for client quoting. ColVA is treated as a valuation input to CSA documentation rather than a desk-level charge in most institutions.

• "Aggregate XVA for a netting set — comparing components: We have a 3-trade netting set with a BBB corporate (no CSA, subject to UMR Phase 6): (1) 10Y pay-fixed IRS, $50M, MTM +$2.1M; (2) 5Y EUR/USD cross-currency swap, €30M, MTM -$0.8M; (3) 2Y FX forward, $15M, MTM +$0.3M. No collateral agreement. Bank funding spread: 42bps over SOFR. Counterparty CDS: 95bps, LGD 60%. SIMM IM for the netting set: estimated $1.6M. Cost of capital: 11% RoE target, risk-free 4.8%, CoC = 6.2%. Calculate each XVA component: (1) CVA using the unilateral formula with quarterly discrete approximation, (2) FVA using the EPE profile and 42bps funding spread, (3) MVA using SIMM IM of $1.6M and funding cost over the weighted average life, (4) KVA using SA-CCR EAD (calculate approximately) × 8% × 100% risk weight × 6.2% CoC. (5) Sum the XVA components and express as a total adjustment in dollars and as a basis point running spread. Which component dominates?"
• "XVA sensitivity report for risk management: Our XVA book shows aggregate sensitivities to the following market factors as of end of month. Format a structured XVA sensitivity report covering: (1) CVA CS01: sensitivity to a 1bp parallel shift in all counterparty CDS curves (current: +$380k per 1bp), broken down by credit quality bucket (AAA/AA: $45k, A: $92k, BBB: $180k, HY: $63k), (2) CVA IR DV01: sensitivity to 1bp parallel shift in SOFR (current: -$72k), (3) FVA funding spread sensitivity: change in FVA for 1bp widening in our own funding spread (current: +$28k), (4) MVA SIMM sensitivity: change in MVA for 10% increase in aggregate SIMM IM (current: +$115k per 10%), (5) KVA RWA sensitivity: change in KVA for $100M increase in SA-CCR EAD (current: +$62k per $100M EAD). Format as an executive risk dashboard suitable for the weekly XVA risk review."

XVA Regulatory Capital: Basel IV CVA Capital Framework

Under Basel IV (final framework, effective across jurisdictions through 2025-2028), CVA risk has its own dedicated capital framework within the standardized approach — separate from the CCR capital for counterparty default. Banks must calculate CVA capital using either the Basic CVA (BA-CVA) approach or the Standardized CVA (SA-CVA) approach (internal models for CVA are no longer permitted under Basel IV).

BA-CVA is the simpler approach for banks without active CVA hedging: SCVA = Σ (ρ × SNSCVA_c)² + (1-ρ²) × Σ SNSCVA_c², where ρ=0.5 is the systemic correlation factor. SA-CVA requires a formal CVA hedging desk and allows recognition of CDS hedges with a more granular calculation using sensitivities. The SA-CVA capital benefit (vs BA-CVA) for an active hedging desk can be 30-50%. For a detailed SA-CCR walkthrough — which feeds into SA-CVA capital — see SA-CCR Explained. For market risk capital under FRTB, which interacts with CVA capital for traded CVA positions, see FRTB and Claude AI.

• "Basel IV CVA capital calculation — BA-CVA approach: Our institution has the following aggregate CVA exposure by credit quality: AAA/AA counterparties: $45M EAD (RW 0.7%), A counterparties: $80M EAD (RW 1.0%), BBB counterparties: $120M EAD (RW 1.5%), HY counterparties: $35M EAD (RW 5.0%). All EAD calculated under SA-CCR. No CDS hedges recognized. BA-CVA formula (simplified full CVA): Capital_CVA = 0.5 × Σ_c [RW_c × EAD_c × M_c] where M_c is effective maturity (assume 3Y average for all buckets). (1) calculate the BA-CVA capital requirement, (2) apply the scaling factor for banks below the threshold (reduced BA-CVA if applicable), (3) what RWA does this generate (multiply capital by 12.5), (4) if we were to implement single-name CDS hedges on our top 5 HY counterparties ($25M of the $35M HY EAD), estimate the BA-CVA capital reduction using the reduced BA-CVA formula that recognizes single-name hedges."

Where to Go from Here

For practitioners building XVA capabilities with Claude, the most productive starting point is the CVA calculation workflow — the discrete quarterly approximation is accurate enough for most desk purposes and the parameter inputs (CDS spreads, LGD, exposure profiles) are readily available. Layer in FVA once the uncollateralized exposure inventory is understood. MVA should be calculated at trade level for any new trade subject to UMR or CCP clearing. KVA is most useful for desk-level capital allocation decisions rather than trade-by-trade pricing until the desk has consistent SA-CCR EAD data.

The XVA and CVA Trading Desk article covers daily P&L attribution, hedge performance analysis, and SA-CCR exposure calculations in more operational detail. The Derivatives AI guide covers broader derivative pricing and structuring workflows including Greeks, scenario analysis, and client documentation. For all CVA and counterparty risk templates, the Compliance & Risk category has the full XVA toolkit.

Frequently Asked Questions

What is the difference between CVA and the CVA capital charge?

CVA (Credit Valuation Adjustment) is a fair-value adjustment to derivative pricing — it appears in the P&L of the XVA desk when credit spreads move, trades are added, or existing exposures change. The CVA capital charge is a separate regulatory requirement under Basel III/IV: it is the amount of regulatory capital a bank must hold against the risk that its CVA moves adversely (i.e., counterparty credit spreads widen). You can think of CVA as the mark-to-market reserve for expected counterparty credit losses, and the CVA capital charge as the capital buffer against unexpected CVA volatility. Both are required, but they appear on different parts of the balance sheet — CVA in fair value reserves, CVA capital in the risk-weighted asset calculation.

How does a CSA reduce XVA?

A credit support annex (CSA) reduces expected exposure by requiring frequent collateral exchange — when the trade moves in your favor, the counterparty posts collateral, reducing the amount you would lose if they defaulted. Lower expected exposure directly reduces CVA (fewer uncollateralized dollars at risk) and FVA (less funding gap to manage). A standard two-way zero-threshold CSA with daily margining can reduce CVA by 70-90% versus an uncollateralized trade. However, a CSA with a material threshold (e.g., $10M), minimum transfer amount, or independent amounts will only partially reduce XVA. MVA, paradoxically, is zero for uncollateralized trades (no initial margin posted) and material for CSA-governed trades subject to UMR or CCP clearing — so adding a CSA reduces CVA and FVA but creates MVA.

What model does ISDA SIMM use and how does it relate to MVA?

ISDA SIMM (Standard Initial Margin Model) is the industry-standard sensitivity-based model for calculating bilateral initial margin under UMR. It computes IM as a function of delta, vega, and curvature sensitivities to prescribed risk factors (IR rates, credit spreads, equity prices, FX rates, commodity prices), aggregated using specified within-bucket and cross-bucket correlations. For MVA calculation, the expected SIMM IM at each future date is estimated by projecting the trade's sensitivities under Monte Carlo paths and computing the SIMM IM at each path and date. The funding cost of that projected IM profile — discounted back to today — is the MVA. ISDA SIMM is updated annually (the current version since 2.6) and the correlations and risk weights are recalibrated each year to reflect recent market volatility.

Can XVA be explained to front office traders who are skeptical of the charges?

Yes, and the most effective explanation uses the funding gap analogy. When a rates trader executes an uncollateralized IRS with a corporate client and hedges it in the interbank market under a CSA, the hedge generates and consumes cash as rates move — but the client trade generates no cash flows until settlement. The trader has created a funding gap that the bank's treasury must finance at the bank's actual funding cost (not the risk-free rate). FVA is simply the present value of that funding cost. CVA is the expected cost of the corporate defaulting before maturity with the trade in-the-money to the bank. Neither is an internal tax — both are real economic costs that, if not charged to the trade, would be absorbed by the bank's shareholders. Framing XVA as the cost of doing uncollateralized business — and showing the client how the cost disappears with a two-way CSA — is usually the most productive approach.