Valuation Adjustments: CVA, DVA, FVA

Every derivative trade you price has a hidden layer of costs that never shows up in the Black-Scholes output: the risk that your counterparty defaults before settlement, the cost of funding uncollateralized positions, and (controversially) the accounting benefit of your own potential default. These valuation adjustments, collectively called XVAs, now routinely swing tens of basis points on a single trade and drive billions in quarterly P&L volatility across major dealer banks. The practical point isn't just understanding the formulas. It's recognizing that XVA charges increasingly determine whether a trade gets done at all, and that the XVA desk has quietly become one of the most powerful functions on a derivatives trading floor.

The XVA Family (What Each Adjustment Actually Does)

Before diving into the three core adjustments, here's the full landscape. You'll encounter six XVAs in practice, though CVA, DVA, and FVA dominate pricing conversations:

The point is: these aren't academic abstractions. When a dealer quotes you a swap, the "mid-market" rate has already been shifted by the XVA desk's charge. On an uncollateralized five-year interest rate swap, XVA costs can add 40-60 basis points upfront, sometimes more than the trading desk's profit margin.

CVA: The Adjustment That Started It All (And Now Requires Capital)

CVA captures a straightforward concept: if your counterparty might default before the trade matures, the trade is worth less than its risk-free value. You're essentially selling credit protection embedded in every uncollateralized derivative.

The calculation: CVA = (1 - Recovery Rate) x Sum of (Expected Exposure at time t x Marginal Default Probability at time t)

The inputs you need:

• Credit spreads from the CDS market (these imply default probability)
• Expected positive exposure (EPE) from Monte Carlo simulation of the portfolio
• Recovery rate, typically assumed at 40% for senior unsecured (though actual recoveries vary enormously)

The takeaway: CVA isn't just a pricing concept anymore. Under Basel III's revised framework, banks must hold regulatory capital against CVA volatility. The two approaches available (SA-CVA and BA-CVA) create very different capital charges, and the choice between them materially affects a bank's willingness to trade with you.

How the Capital Charge Works

The Basel III finalized framework (with EU implementation beginning January 2025 and UK/US targeting mid-2025) introduced two methods for computing CVA capital:

Why this matters: SA-CVA produces significantly lower capital charges for banks that actively hedge CVA (because it recognizes exposure hedges that BA-CVA ignores). But getting SA-CVA approval requires sophisticated infrastructure. The result is a two-tier market: banks with SA-CVA approval can price more aggressively on complex, long-dated trades, while BA-CVA banks face higher capital costs that get passed directly to clients.

A Worked CVA Example (So You Can See the Numbers)

Your situation: You're a corporate treasurer entering a five-year, $100 million interest rate swap (receiving fixed) with a bank. The bank needs to calculate CVA on your exposure.

Your credit profile:

• Five-year CDS spread: 150 basis points
• Recovery rate assumption: 40%

Step 1: Extract default probability Annual hazard rate = Spread / (1 - Recovery) = 1.50% / 0.60 = 2.5% per year Five-year cumulative default probability = 1 - e^(-0.025 x 5) = ~11.75%

Step 2: Estimate exposure profile Monte Carlo simulation across thousands of rate scenarios produces an average expected positive exposure (EPE) of $4.5 million over the life of the swap. Exposure peaks around year two to three (that's typical for interest rate swaps, where the mark-to-market grows as rates diverge from the fixed rate, then converges back toward zero at maturity).

Step 3: Calculate the charge CVA = 60% (LGD) x $4.5M (EPE) x integrated marginal default probabilities = ~$320,000

Interpretation: The bank will embed roughly 32 basis points upfront into your swap rate to compensate for the risk that you default while owing money. If you post daily variation margin under a CSA, that exposure drops dramatically (often to $50,000 or less), because collateral eliminates most of the credit exposure window.

DVA: The Adjustment Nobody Likes (But Accounting Requires)

DVA is the mirror image of CVA, calculated from the counterparty's perspective. When your own credit deteriorates, your liabilities are worth less (because there's a higher chance you won't pay), and accounting standards require you to record a gain.

The practical problem: you book profits precisely when your firm is in financial trouble. During the 2011 European debt crisis, several major banks reported billions in DVA "gains" as their credit spreads widened, which looked absurd to investors and analysts. The data pattern repeated during COVID-19 volatility in early 2020 (when bank CDS spreads spiked and DVA gains appeared across the industry).

DVA = Sum of (Own Default Probability x Expected Negative Exposure x Own LGD)

The controversy runs deep:

• For DVA: It's mathematically symmetric to the counterparty's CVA, and accounting standards (IFRS 13, ASC 820) require fair value to reflect all market participants' views, including your own credit risk
• Against DVA: You can't monetize your own default. There's no hedge. The "gain" is unrealizable, and it creates perverse incentive signals
• The compromise: IFRS 9 now allows (and most banks use) routing DVA changes through Other Comprehensive Income (OCI) rather than P&L, which removes the earnings volatility but keeps it on the balance sheet

The signal worth remembering: most sophisticated investors and bank analysts strip DVA out when evaluating performance. If you see a bank's trading revenue boosted by a large DVA gain, treat it as noise (or worse, as a warning signal about the bank's credit quality). Many institutions now exclude DVA from internal performance metrics entirely, even though they must report it for accounting compliance.

FVA: The Adjustment That Divided Quantitative Finance

FVA reflects something intuitive: if a bank enters an uncollateralized derivative and must fund that position by borrowing (at a spread above the risk-free rate), that funding cost is real and should be reflected in the price.

FVA = Funding Spread x Expected Funding Requirement x Duration

The two components:

• FCA (Funding Cost Adjustment): The cost when you need to post collateral or fund a positive exposure to an uncollateralized counterparty
• FBA (Funding Benefit Adjustment): The benefit when you receive funding (the counterparty's uncollateralized negative exposure provides you cheap funding)

Why the Debate Still Isn't Settled

The academic objection (led most prominently by John Hull and Alan White) is that FVA represents double-counting: if you've already priced CVA and DVA correctly, adding FVA charges the counterparty twice for credit risk. In Modigliani-Miller terms, a firm's funding costs shouldn't affect the value of individual projects.

The practitioner response (which has largely won the operational argument) is simpler: treasury desks charge funding costs, and those costs are real cash flows that traders must recover. Whether or not FVA is theoretically correct, ignoring it means trading at a loss to the funding desk.

The point is: the debate is less about who's right in theory and more about what goes into the price in practice. Every major dealer bank now includes FVA in derivatives pricing. The remaining disagreement is whether to use symmetric funding curves (same rate for borrowing and lending) or asymmetric curves (reflecting the bid-offer spread in actual funding markets). Most banks have moved toward asymmetric FVA, which better reflects real-world funding conditions but adds computational complexity.

Collateral Transforms Everything

The single most powerful lever for reducing XVA charges is collateralization. Here's how the numbers change for that same $100 million, five-year swap:

The practical antidote to high XVA costs isn't negotiating the charge down. It's upgrading your collateral agreement. Moving from uncollateralized to daily margining reduces your total XVA by roughly 80-85%, which often dwarfs any benefit from shopping swap rates across dealers.

The XVA Desk (Where Pricing, Risk, and Capital Collide)

At major dealer banks, XVA management has evolved from a back-office accounting adjustment into a centralized desk that sits between trading, risk management, and treasury. The XVA desk owns the pricing, hedging, and P&L of all valuation adjustments across the entire derivatives book.

What the XVA desk actually does:

• Prices new trades: Every derivative gets an XVA charge before the trader can quote a client
• Hedges the book: The desk actively trades CDS, interest rate products, and funding instruments to manage aggregate XVA risk
• Manages regulatory capital: Particularly under SA-CVA, the desk optimizes hedges that reduce capital consumption
• Runs netting optimization: Trades within the same netting set partially offset each other's exposures (a new trade that reduces net exposure can actually generate an XVA benefit)

Why this matters: understanding the XVA desk's incentives explains otherwise puzzling dealer behavior. A trade that looks profitable to the trading desk might get killed by a large XVA charge. Conversely, a marginally profitable trade might get done because it reduces net exposure within a netting set (and thus lowers overall XVA).

Wrong-Way Risk (The Hidden Multiplier)

Standard CVA models assume that your counterparty's default probability is independent of the trade's exposure. Wrong-way risk is what happens when that assumption fails, and the two become positively correlated.

The classic example: You sell a put option to a counterparty. Markets crash, the option goes deep in-the-money (your exposure spikes), and simultaneously the counterparty's credit deteriorates (because they're exposed to the same market decline). Your exposure is highest precisely when default is most likely.

Wrong-way risk can multiply CVA by 2-5x beyond what standard models suggest. It's notoriously difficult to model (because it requires joint simulation of market risk and credit risk), but ignoring it is how banks have historically underpriced risk in correlated stress events. Sovereign-bank wrong-way risk during the European debt crisis was a painful case study (bank exposure to sovereigns spiked just as sovereign default probabilities climbed).

The test: for any major counterparty exposure, ask yourself whether the scenarios that increase your mark-to-market are the same scenarios that stress the counterparty's credit quality. If the answer is yes, standard CVA is insufficient.

XVA Mitigation Checklist (Tiered)

Essential (high ROI)

These four actions eliminate the majority of XVA costs:

• Negotiate daily variation margin under a CSA/ISDA agreement (reduces CVA and FVA by 80%+)
• Consolidate trades into netting sets aligned with ISDA master agreements (netting can reduce exposure by 40-60%)
• Clear eligible trades through CCPs (eliminates bilateral CVA entirely, though introduces MVA)
• Monitor counterparty credit spreads weekly for material widening (a 100bp spread move can double CVA on long-dated trades)

High-Impact (workflow and infrastructure)

For institutions managing significant derivatives portfolios:

• Build or buy Monte Carlo exposure simulation (the foundation of accurate CVA)
• Implement wrong-way risk scenarios for concentrated counterparty exposures
• Establish daily CVA/FVA P&L attribution to separate XVA moves from trading P&L
• Optimize collateral allocation across agreements to minimize FVA

Advanced (for dealer banks and large buy-side)

If you're running or negotiating with an XVA desk:

• Evaluate SA-CVA vs. BA-CVA capital treatment and build infrastructure for the preferred approach
• Implement portfolio compression to reduce gross notional and exposure within netting sets
• Run netting-set-aware pre-trade analytics so new trades are priced with incremental (not standalone) XVA

Next Step (Put This Into Practice)

Pull up your largest uncollateralized derivative position and estimate the CVA embedded in it.

How to do it:

1. Find the counterparty's CDS spread (or bond spread as proxy) and calculate the implied annual default probability: Spread / (1 - 0.40)
2. Estimate your expected positive exposure (for a plain swap, roughly 2-5% of notional at peak)
3. Multiply: CVA = 60% x EPE x cumulative default probability over the remaining life

Interpretation:

• Under $50,000: Likely already collateralized or short-dated, XVA is well-managed
• $50,000-$500,000: Material, worth reviewing your CSA terms to see if tighter margining would reduce costs
• Over $500,000: Significant drag on the trade's economics, investigate collateral upgrades, novation to a CCP, or portfolio compression

Action: If your estimated CVA exceeds the annual carry on the trade, the counterparty credit risk cost is eating your return. That's the point where upgrading collateral terms or restructuring the trade becomes the highest-ROI move available to you.