Capital and Liquidity Frameworks Driving KVA

KVA is driven by how dealers translate derivatives exposure into regulatory capital and liquidity requirements. In practice, this is governed by three interacting constraints:

• Risk-weighted capital (Basel RWA framework)
• Leverage ratio (non-risk-based backstop)
• Liquidity requirements (HQLA under LCR)

All three are ultimately driven by regulatory exposure measures that are conceptually similar to EPE, but defined by rule rather than model. Under current regimes, this is primarily determined using SA-CCR (Standardised Approach for Counterparty Credit Risk).

1. Standardised RWA for Derivatives (Basel Framework)

The core building block is Exposure at Default (EAD) under SA-CCR.

At a high level:

• Replacement Cost (RC): current MTM, floored at zero
• Plus Potential Future Exposure (PFE): driven by supervisory factors, asset class, maturity, and netting
• Adjusted for collateral, margining, and netting sets

This produces:

• EAD = 1.4 x (RC + PFE)

The 1.4 multiplier is a regulatory scaling factor applied to ensure conservatism in the standardised framework.

EAD is then converted into capital requirements:

• RWA = EAD × Risk Weight (based on counterparty credit quality)
• Capital held = ~10–12% of RWA (including buffers)

There is also a separate CVA capital charge reflecting mark-to-market volatility due to credit spread movements.  For more details, here is the full SA-CCR rule: BIS rules.

CSA impact

A two-way CSA reduces both RC and PFE:

• Variation margin reduces current exposure (RC)
• Collateralisation and netting reduce PFE

This directly reduces EAD, RWA, and therefore capital consumption. This is the primary mechanical driver of KVA reduction under the RWA framework.

2. Leverage Ratio Constraint

The leverage ratio is a non-risk-based constraint:

• Applies a flat capital requirement to total exposure
• Uses a broadly SA-CCR-derived exposure measure
• Provides more limited recognition of netting and collateral

The structure is:

• Total leverage exposure = on-balance-sheet assets + derivative exposure + off-balance-sheet items
• Capital requirement = ~3–5% of this exposure

CSA impact

A CSA reduces exposure, but less efficiently than under RWA due to constrained recognition of collateral and netting.

The Binding Constraint: “Greater Of”, Not “Sum Of”

Dealers must satisfy both RWA-based and leverage constraints. However, economically they price to the binding constraint - i.e. the higher of the two.

They do not charge for both independently. In practice:

• Some dealers are RWA-constrained (more common in Europe)
• Others are leverage-constrained (often US G-SIBs)

Implications:

• Only one constraint should drive KVA at the margin
• The same trade can generate materially different capital costs across dealers

This is a primary source of dispersion and therefore negotiable value.

___________________________________________________________________________________________________________

Below we address some of the negotiating tactics applied during CSA negotiations, and outline the correct approach to challenge the push backs

“Capital is a sunk cost”, ”Capital cannot be hedged" arguments. And why they are wrong

It is often argued that:

• CVA and FVA can be dynamically hedged over the life of a trade
• KVA cannot, as it represents an accrual of capital costs over time rather than a tradable risk factor

From this, dealers sometimes draw a further conclusion:

• Because KVA is not hedgeable, its evolution over time cannot be locked in
• Therefore any subsequent reduction in KVA (e.g. from a CSA) cannot be “returned” in the same way as CVA/FVA

This line of reasoning is superficially appealing, but economically incorrect.

What Is True (but Limited in Scope)

It is correct that:

• KVA is not directly hedgeable in the way CVA (credit spreads) or FVA (funding spreads) are
• There is no market instrument that allows a dealer to lock in future capital requirements
• As a result, KVA is realised as an accrual over time, rather than a fully crystallised P&L item - either at trade inception and/or subsequent change in portfolio or CSA terms

It is also true that:

• Attempting to “hedge” KVA using proxy instruments would introduce accounting mismatches
• The realised KVA over time will differ from initial estimates due to market evolution, portfolio changes, and regulatory updates

Why This Does Not Prevent KVA Release

The above is an internal risk management and accounting issue, not an economic constraint.

KVA is:

• The present value of expected future capital consumption
• Determined by exposure profiles (EAD), regulatory rules, and target return on capital

A CSA that reduces exposure:

• Mechanically reduces future EAD
• Mechanically reduces future capital requirements

Therefore reduces the expected future capital cost stream

This effect is:

• Real
• Quantifiable at the point of renegotiation
• Independent of whether the bank previously hedged or accrued for it

The Key Distinction: Portfolio Accounting vs Marginal Economics

The dealer argument implicitly relies on the following conflation:

“Because we could not perfectly hedge or lock in KVA over time, we cannot realise or return changes in KVA today.”

This is incorrect because:

• Banks manage KVA at the portfolio level, across thousands of trades and counterparties
• Accrual variability is naturally diversified and absorbed within the broader balance sheet
• Capital is continuously reallocated across businesses based on marginal returns

From an economic perspective:

• What matters is the current forward expectation of capital consumption, not historical accrual paths
• A reduction in expected future capital usage has an immediate marginal economic value

That value manifests as:

• Released capacity to support new trades
• Reduced need to raise or allocate capital
• Improved return on capital for the business

Implication for CSA Negotiation

Therefore:

• The inability to hedge KVA does not invalidate its valuation
• Nor does it prevent a before-and-after comparison when a CSA is introduced
• Nor does it eliminate the economic value of capital relief

The correct framing is:

KVA is not hedgeable, but it is still a forward-looking economic cost.
If that cost is reduced, the benefit is real - even if it is realised over time.

Failure to compensate for this reduction is not a reflection of economic reality, but of:

• Internal accounting conventions
• Desk-level P&L smoothing
• And, in practice, a negotiation stance

Bottom Line

• KVA accrual mechanics affect timing and attribution of P&L, not economic value
• The reduction in future capital consumption from a CSA is genuine and monetisable
• The fact that it is realised over time does not negate its present value today

Accordingly:

The “non-hedgeability” of KVA is not a barrier to value transfer - it is a framing device that obscures it. In practice, this argument is applied selectively: dealers are willing to charge KVA upfront based on forward-looking assumptions, but resist returning it when those same assumptions improve. The asymmetry is commercial, not economic.

This creates a clear inconsistency:

• If KVA cannot be returned because it is an accrual and not locked in or hedgeable, then by the same logic it should not be charged upfront on new trades or on CSA changes that increase capital consumption.
• In reality, dealers do charge KVA on a forward-looking basis whenever it increases expected capital usage.

Therefore:

• KVA is clearly treated as a present value economic quantity when it is favourable to the dealer, and as a non-realisable accrual when it is not.

The correct and consistent treatment is:

• KVA should be valued symmetrically as the present value of expected future capital costs, and any reduction in those costs should be reflected in pricing, irrespective of hedgeability or accounting treatment.

 To be more aggressive in negotiation settings, the key argument is:

“If it can be charged, it can be returned.”

Practical Interpretation for Clients

From a negotiation perspective:

• SA-CCR is the regulatory equivalent of EPE - but rule-based and conservative
• Capital is priced to the binding constraint, not both frameworks
• Capital relief from a CSA is real, marginal, and transferable

Liquidity requirements:

• Translate directly into funding costs via HQLA
• Are often embedded within FVA or presented as opaque liquidity add-ons

Critically:

• FVA is often visible in MTM
• CVA is partially visible
• KVA and liquidity costs are embedded, model-driven, and opaque

This creates:

• Material dispersion across dealers
• Scope for inconsistent or overstated assumptions
• A structural information asymmetry

The combination of:

• Opaque modelling
• “Sunk capital” arguments
• Mischaracterisation of accrual effects
• Non-transparent liquidity add-ons

means that KVA and liquidity costs are often where the largest share of value is retained unless actively challenged.