The ability to extract economic value from the heterogeneous funding (or repo) characteristics of collateral eligible under a CSA was first properly understood during the transition from LIBOR to OIS discounting. That transition forced the market to confront a fundamental reality that had previously been obscured: collateral is not economically neutral. The identity and funding value of the collateral posted directly determines the appropriate discount curve, and therefore the present value of any derivative position.

Once discounting became explicitly collateral-linked, it became clear that a CSA is not just a credit mitigation document—it is a funding and valuation instrument. Optionality embedded in eligible collateral sets, haircuts, and substitution rights translates directly into economic value.

In the early phase of this transition (circa 2009–2010), leading dealers recognised this ahead of the buy-side. At least one major Wall Street institution is widely understood to have generated a significant proportion of its FICC revenues by systematically pricing trades off funding curves that were advantageous relative to those implicitly assumed by counterparties. In effect, they monetised the gap between their own funding reality and the client’s pricing framework.

As awareness of multi-curve discounting increased, the dealer community invested heavily in infrastructure—upgrading pricing libraries, collateral optimisation engines, and XVA frameworks—to internalise and neutralise these arbitrage opportunities. What had initially been an opaque source of dealer alpha became embedded into standard pricing. The economic effect did not disappear; it was simply crystallised and transferred to clients who had not adapted their frameworks accordingly.

However, the system today is still not fully efficient—and in specific areas remains structurally inconsistent.

One of the most persistent and underappreciated sources of inefficiency arises in cleared or “give-up” trading models involving an executing broker (EB) and a clearing agent (CA).

In these structures:

• The executing broker (EB) prices the trade using discount curves aligned to its own clearing relationship, collateral terms, and internal funding stack
• The trade is then passed (“given up”) to the clearing agent (CA)
• The clearing agent (CA) values the same position using its own discount curves, reflecting its own collateral agreements, funding costs, and optimisation constraints

Critically, the post-trade approval and allocation processes governing this workflow are typically economically blind to discounting assumptions. They validate trade economics in nominal terms (rate, spread, structure), but not in present value terms under consistent discounting frameworks.

The result is a structural inconsistency:

Identical contractual cashflows are assigned different present values depending solely on which balance sheet is performing the valuation.

This creates a persistent valuation wedge at inception—one that is not driven by market risk, but by differences in funding curves, collateral eligibility, and internal optimisation frameworks.

For trades with limited funding sensitivity, this effect is second order.
For trades with material collateral optionality or long-dated cashflows, it is economically significant.

In these cases, the mismatch can create what is effectively a manufactured backwardation:

• The trade is executed at a level reflecting one discounting framework
• It is subsequently valued under another, systematically different framework
• The resulting PV difference represents an embedded, repeatable transfer of value
• This is not theoretical. It is observable, measurable, and -critically -actionable.

Where value is lost (or captured):

• Misalignment between EB and CA discount curves
• Differences in collateral eligibility sets and substitution rights
• Funding asymmetries across dealers (balance sheet, repo access, regulatory constraints such as leverage ratio or SLR)
• Operational processes that fail to enforce economic consistency at trade acceptance

Why this persists:

• Market structure fragmentation (multiple dealers, multiple clearing relationships)
• Opaque dealer pricing frameworks (particularly around funding and XVA components)
• Buy-side focus on headline pricing (rate/spread) rather than PV-consistent execution
• Post-trade processes that are not designed to enforce discounting consistency

Why it matters:

Because in an environment where outright credit risk can be largely eliminated through collateralisation, funding becomes the dominant driver of economics. Small differences in discounting assumptions, when applied to large notional and long-dated flows, translate into meaningful value transfer.

Our perspective

This is not an abstract inefficiency—it is a systematic source of value leakage for clients that do not actively manage it, and a source of repeatable value capture for those that do.

The key is not simply to “be aware” of discounting differences, but to:

• Diagnose where mismatches arise across execution and clearing chains
• Quantify the PV impact under alternative discounting assumptions
• Control execution pathways to align pricing and valuation frameworks
• Exploit residual asymmetries where alignment is not enforced

This requires a combination of:

• Deep understanding of dealer funding and XVA pricing frameworks
• Practical experience of how trades are priced, approved, and processed across EB/CA structures
• The ability to challenge dealer assumptions and reverse-engineer pricing where necessary
• The reality is that this part of the market remains highly opaque and poorly arbitraged.

Clients that rely on surface-level pricing or standard execution processes will systematically give up value.
Clients that approach this with a disciplined, insider-informed framework can extract it.