Collateral Terms and Their Impact on XVA and Pricing

Collateral terms within a CSA are not operational detail — they are primary economic drivers of XVA. Parameters such as thresholds, MTAs, eligibility schedules, and termination triggers directly shape exposure profiles, funding requirements, and capital consumption. These, in turn, determine CVA, FVA, and KVA, and therefore the all-in price a dealer charges.

The critical point in practice is that these terms are often negotiated with an operational lens rather than an economic one. This leads to systematic mispricing and a persistent transfer of value from the client to the dealer unless actively managed.

Thresholds

The threshold defines the amount of unsecured exposure that can accumulate before collateral must be posted.

A positive threshold allows the client to be out-of-the-money to the dealer without posting collateral. Economically, this is equivalent to granting the dealer an unsecured credit exposure.

Impact on XVA:

• CVA increases materially, as expected positive exposure (EPE) rises directly with the unsecured MTM profile
• FVA increases, as the dealer must fund this exposure internally
• KVA increases due to higher regulatory capital requirements associated with unsecured counterparty risk

Thresholds are one of the most powerful levers in CSA negotiation. Reducing or eliminating them compresses all three XVA components simultaneously. Dealers typically price thresholds conservatively — often assuming stressed exposure profiles — so tightening thresholds can generate disproportionately large pricing improvements.

In practice, moving toward a zero-threshold (fully collateralised) structure is often the single largest driver of XVA release.

Minimum Transfer Amount (MTA)

The MTA defines the minimum size of collateral movements. Exposure below this level is effectively uncollateralised.

In theory, MTA introduces a band of unsecured exposure and therefore contributes to CVA, FVA, and KVA. In practice, the economic significance depends on magnitude and symmetry.

Refined view:

• FVA impact is largely symmetric and second-order. Small MTAs create short-lived, oscillating exposures that tend to net out over time, resulting in limited funding cost
• CVA impact is not symmetric. Even small unsecured exposures contribute to expected positive exposure, particularly under stress where volatility increases
• KVA impact follows a similar pattern to CVA, as capital is driven by tail exposure and stress calibration rather than average exposure

Practical conclusion:

MTA should be kept low to maintain tight collateral discipline, but it is not typically a primary pricing lever. Its role is best understood as avoiding “sloppy collateralisation” rather than generating material XVA savings.

Dealers will rarely concede meaningful pricing improvements purely for MTA tightening — and correctly so. It is hygiene, not alpha.

Eligible Collateral (and Haircuts)

The eligibility schedule defines which assets can be posted as collateral and the associated haircuts. This is one of the most misunderstood — and most consistently mispriced — aspects of CSA design.

The standard industry approach, often encouraged by large dealers and custodians, is to maximise the eligible collateral set:

• Include government bonds, corporates, equities, funds
• Provide maximum flexibility to the posting party
• Optimise for operational convenience

This is typically presented as best practice. In reality, it is often economically suboptimal.

Core issue:

Optionality only has value if it is exercised.

If the client is operationally or structurally constrained to post a narrow set of assets (typically cash or a limited pool of HQLA), then a wide eligibility schedule provides no real benefit. However, dealers will still price the optionality embedded in that flexibility.

They assume:

• The posting party will deliver cheapest-to-deliver collateral (from the dealer’s perspective)
• Collateral will be selected in a way that maximises the dealer’s funding cost

This results in an FVA charge that reflects worst-case collateral usage, not actual behaviour.

The outcome is structurally inefficient:

• The client pays for optionality
• The client does not use that optionality
• The dealer captures the economic value

This is a classic mistake and is frequently reinforced by outsourced collateral management models, including those provided by JPMorgan Chase, State Street, Goldman Sachs, and Bank of New York Mellon. These frameworks tend to optimise for operational robustness and flexibility rather than economic efficiency.

Correct approach:

• Align the eligibility set with actual posting behaviour
• Do not include assets that will never realistically be delivered
• Simplify the CSA rather than broadening it unnecessarily

Collateral optimisation should instead be managed externally via repo or securities lending markets, where the client retains control of funding decisions.

Rating-Triggered Termination Rights

These clauses allow one party — typically the dealer — to terminate trades if the counterparty’s credit rating falls below a specified level.

From the dealer’s perspective, this reduces exposure by truncating the tail of the distribution:

• CVA decreases due to reduced exposure in stress scenarios
• FVA decreases due to a shorter effective exposure horizon
• KVA decreases as tail risk and stress losses are capped

However, the economic effect is asymmetric.

The client is effectively selling a credit option to the dealer:

• The dealer can terminate precisely when the client’s credit deteriorates
• This is typically when the trades are most valuable to the client
• The clause embeds wrong-way optionality in favour of the dealer

This optionality is rarely priced transparently.

Negotiation implication:

• These clauses have clear economic value and should result in measurable pricing benefit
• In practice, this benefit is often under-delivered
• Removing or softening rating triggers (e.g. higher thresholds, cure periods, mutuality) can be valuable

In many cases, collateral-based mechanisms (such as increased margining) are preferable to hard termination rights

Cross-Term Interaction and Strategic Framing

These terms operate as a system rather than in isolation:

• Thresholds and MTA define the baseline unsecured exposure profile, driving CVA and KVA
• Eligible collateral defines the cost of funding that exposure, driving FVA
• Termination rights reshape the tail of the exposure distribution, affecting CVA asymmetry and embedded optionality

Dealers convert these features into a composite XVA charge embedded in pricing, typically using internal models, conservative assumptions, and limited transparency.

From a negotiation perspective, the key insight is that improving collateral terms is often more effective than negotiating headline pricing. However, improvements only create value if the economic benefit is explicitly extracted.

Bottom Line

Collateral terms are the mechanism through which credit risk, funding cost, and capital usage are transferred and priced.

• Thresholds are a primary driver of XVA and the most powerful negotiation lever
• MTA should be kept low for discipline but is not a major economic driver
• Eligible collateral is frequently mis-specified, leading to priced but unused optionality and silent value transfer
• Rating triggers embed asymmetric optionality and must be explicitly valued

The consistent pattern in practice is that dealers retain a significant portion of the economic benefit unless challenged.

The governing principle is simple:

• Only negotiate for optionality you will actually use. Everything else is a transfer of value to the dealer.