Negative Interest Under the ISDA 1995 Credit Support Annex: What the (2019) Court of Appeal Decision Means for 2025—and How TRW Structures Your CSAs

Audience: Bangladesh-origin banks, NBFIs, corporates, and funds that enter into OTC derivatives (FX, rates, commodities) with EU/UK counterparties and continue to hold or renegotiate legacy ISDA 1995 English law Credit Support Annex (CSA) frameworks.
Core message: The English Court of Appeal (2019) confirmed that the standard form ISDA 1995 CSA does not oblige a Transferor of cash collateral to pay “negative” interest. That holding still matters in 2025 for legacy books, back-to-back hedges, and close-out valuations—especially where treasury teams confront rate-regime pivots, cross-currency funding, and mixed documentation (1995 CSA on some lines; 2016 VM CSA on others). This article explains the decision, dissects its reasoning, and translates it into practical drafting and operational playbooks that TRW implements for clients in Dhaka, London, and Dubai.

(Per your policy, we include only internal TRW links where helpful—for deeper context see Secured Lending & Syndication and Regulatory (Bangladesh Bank).)

1) The Business Problem That “Negative Interest” Exposes

Cash collateral posted under a CSA accrues interest. In “normal” rate environments, interest is positive, and the contract typically spells out who pays whom. But when benchmark rates dive below zero, a conceptual tension arises: should the receiver of cash collateral (Transferee) pay interest to the poster (Transferor)? Or does the contract contemplate only positive interest flowing in the opposite direction?

Why this matters in 2025, even after rate normalization in many economies:

• Legacy contracts: Thousands of relationships still run under ISDA 1995 CSAs (some amended piecemeal), especially for long-dated infra hedges and back-to-back structures.
• Volatility is cyclical: Rate regimes can slip below zero again under stress. “We’ll never see negatives again” is not a legal position.
• Pricing & disputes: If your CSA is silent (or asymmetric) on negative interest, P\&L transfers, margin calculations, and close-out amounts can shift materially during a rate shock.
• Cross-border books: Bangladesh-origin treasuries hedging in USD/EUR/GBP with EU/UK dealers need clarity so ops, audit, and lenders can trust the collateral engine across cycles.

2) A Quick Refresher: What the ISDA 1995 English Law CSA Does

The 1995 English law CSA is a title-transfer collateral annex. In the vanilla setup:

• The Transferor posts cash (or securities) to the Transferee when exposure calls arise.
• The Transferee typically owes the Transferor positive interest on posted cash at a defined benchmark ± a spread (the Price Differential / Interest Amount concept).
• Collateral is re-transferred when exposure drops or upon termination, subject to netting and thresholds.

Crucial limitation: The 1995 form was drafted with a positive-rate paradigm. It contains a detailed mechanism for paying positive interest on cash collateral but is silent on the reverse flow that would operationalize negative interest.

3) The 2019 Court of Appeal Decision—Plainly Stated

In 2019, the English Court of Appeal confirmed that the standard ISDA 1995 CSA does not provide for payment of “negative” interest by a Transferor of cash collateral. The Court reached the same end result as the High Court but criticized the lower court for being “too simplistic” in how it got there. The appellate court’s three pillars:

1. Textual anchor in paragraph 5(c)(ii)
   The clause that deals expressly with positive interest is the “obvious place” one would expect to see negative interest if intended. It isn’t there. That asymmetry strongly indicates no obligation to pay negative interest arises under the standard form.
2. User’s Guide and background materials
   The court treated the ISDA User’s Guide (1999) and surrounding best-practice commentary as background showing the market did not contemplate negative interest payments under the 1995 CSA. Even post-contract materials (near contemporaneous best-practice notes) were considered relevant to understanding market thinking.
3. Contract read as a whole / business common sense
   Taking the CSA in its entirety, there is nothing signalling that negative interest was meant to be paid. If rates later turned negative, that was an unforeseen market development—not a gap for the court to fill by implying reciprocal negative interest obligations.

Bottom line: Under the standard 1995 CSA, no contractual obligation arises to pay negative interest on cash collateral from the Transferor to the Transferee.

4) Why the Decision Still Matters in 2025

Even though many relationships migrated to the 2016 ISDA VM CSA (which better reflects modern margin regimes), you may still have:

• Legacy hedges under a 1995 CSA (possibly with bespoke amendments).
• Back-to-backs where one leg uses a 1995 CSA (e.g., an old project finance hedge) and the other uses a modern VM CSA.
• Mixed portfolios where operational teams rely on shared procedures across different annex types.

For any book where the 1995 CSA remains relevant, the 2019 decision offers litigation-grade clarity: unless you amended the annex or adopted a specific protocol, negative interest is not contractually owed by the Transferor.

5) What the Courts Actually Valued in the Reasoning

Understanding the legal method helps you draft stronger positions:

• Location matters: When a contract has a specific clause for interest (paragraph 5(c)(ii)), courts expect both sides of the coin (positive and negative) to be spelled out there if intended. Silence where you’d expect text is meaningful.
• Market documentation context: English courts give respectful weight to industry standardization, especially for ISDA forms shaped by thousands of practitioners.
• Common sense over mechanical symmetry: The court refused to force “symmetry” just because positive interest exists; it looked at intended economics and document design, not math for math’s sake.

6) Practical Implications for Bangladesh-Origin Parties

A) If you hold a 1995 CSA (unamended):

• No negative interest payable by the Transferor on posted cash, absent bespoke drafting to the contrary.
• Dispute avoidance: Ensure your ops and counterparty teams align on this; memorialize it in a side letter or portfolio memo so staff turnover doesn’t resurrect old misunderstandings in a future rate shock.
• Pricing awareness: Dealers may factor the asymmetry into pricing, especially if they perceive one-way economics when rates dive. Understand the spread trade-off if they push you to modernize.

B) If you face a dealer asking to “turn on” negative interest under a 1995 CSA:

• Resist “implied term” arguments. The appellate decision is strong authority that implied reciprocity isn’t there.
• Consider bargaining: You can negotiate commercial give-and-take (e.g., adjustments to Minimum Transfer Amounts, eligibility, or haircuts) if you agree to negative interest prospectively—but put it in writing with precise drafting.

C) If you are migrating to a 2016 VM CSA:

• Treat the 1995 annex’s negative-interest position as legacy risk and design the cutover explicitly.
• Train treasury on which annex governs which trades, especially during transition periods when both annexes sit side-by-side.

D) If you run back-to-backs (intragroup or client-to-street):

• Mismatches between a 1995 annex on one leg and a 2016 VM CSA on the other can generate P\&L noise in stress scenarios. TRW maps the cash-flow stack, sets internal transfer pricing, and—if needed—re-papers the vulnerable leg.

7) The 2014 ISDA Collateral Agreement Negative Interest Protocol—What It Did (and Didn’t)

The 2014 Protocol was ISDA’s pragmatic answer to market uncertainty. It gave counterparties a standardised way to amend interest provisions in certain collateral agreements so that negative interest could be recognized contractually.

But note:

• Protocol adherence is voluntary and relationship-specific. If you didn’t adhere (or your counterparty didn’t), your 1995 CSA may remain unamended.
• Even where adhered, operational clauses (valuation times, netting sequences, rounding, floor logic) still need to be understood by treasury and middle office.

TRW’s position: Don’t assume your CSA reflects the Protocol. We audit the signed adherence list, the Annex inventory, and any bespoke amendments your teams made across years of renewals.

8) The 2016 ISDA VM CSA: Designed for Modern Margin, Not for Guesswork

The 2016 VM CSA (English law) modernised the daily variation margin framework and sits comfortably with post-crisis risk mitigation regimes. Among the benefits:

• Cleaner alignment of daily VM calculations, collateral eligibility, and haircuts.
• Improved scaffolding for interest mechanics, reducing ambiguity across rate regimes.
• Operational clarity for call windows, settlement cut-offs, and dispute processes, which matter for Dhaka–Dubai–London time-zone choreography.

However: If you run a mixed estate (some 1995 CSAs; some 2016 VM CSAs), you still need policy and process differentiation. The 2019 decision remains relevant for any pocket where the 1995 form persists.

9) Drafting Tactics TRW Uses in 2025

When we (re)paper your CSA stack, we focus on clarity, symmetry where intended, and operational truth:

1. Define interest economics expressly

• If negative interest is intended, draft it plainly: who pays whom, at which benchmark ± spread, with what floor.
• If not intended, state a zero floor explicitly to eliminate interpretive drift.

1. Disaggregate VM vs IM

• VM interest mechanics can differ from segregated IM (where interest often belongs to posted collateral owner, subject to custodian/platform terms).
• Keep IM economics and ops (segregation, reuse bans, control agreements) out of VM clauses to avoid cross-contamination.

1. Currency-aware haircuts and floors

• If VM posts occur in USD against EUR exposures (or vice versa), fix FX haircut logic; decide if interest rates follow the collateral currency, exposure currency, or a specified benchmark.

1. Rounding, MTA, and day-count conventions

• These “small” terms drive real cashflows. Align day-count to the benchmark used for interest; ensure Minimum Transfer Amounts prevent noise but don’t cause cliff-edge calls.

1. Dispute mechanics

• Name the valuation agent(s) and data sources; set a tolerance; provide escalation and interest on adjustments to avoid relationship damage during volatility spikes.

10) Operational Reality for Bangladesh Treasuries

A) Time-zone choreography
Daily calls often reference London close; settlements must clear through correspondent banks before cut-offs. We map Dhaka banking windows against Dubai and London to stop settlement fails.

B) Liquidity staging
Posting VM in USD/EUR/GBP while revenues accrue in BDT requires offshore pools and standing lines. TRW’s Dubai/London teams help set funding rails that comply with Bangladesh Bank rules and avoid “daylight” liquidity gaps.

C) Board-level governance
Update the Derivatives Use Policy: state whether negative interest can arise (and under which annexes), define authorized signatories, and require quarterly reporting on collateral P\&L.

For complementary reading on policy alignment and lender interfaces, see Regulatory (Bangladesh Bank) and Secured Lending & Syndication.

11) “What If” Scenarios (2025 Lens)

Scenario 1: Rates lurch negative in one currency for six months. You run a 1995 CSA (unamended).

• Effect: No negative interest obligation from you (as Transferor) under the standard form.
• Action: Communicate position early; issue a relationship note to counterparties; monitor for pricing pushback.

Scenario 2: Back-to-back hedging with a client: upstream 2016 VM CSA (negative interest allowed), downstream 1995 CSA (silent).

• Effect: Possible P\&L mismatch in a negative-rate window.
• Action: Insert inter-affiliate adjustment or transition downstream annex to a modern form; until then, set internal accrual to cushion basis risk.

Scenario 3: Counterparty proposes a quick side letter “recognizing negative interest going forward.”

• Effect: Might be fine, but watch for spill-overs (e.g., changes to benchmark fallback, floors, FX haircuts) hidden in drafts.
• Action: TRW redlines to confine the change to exactly what’s intended; consider a pricing concession in your favour.

Scenario 4: Close-out during a negative-rate month.

• Effect: Annex asymmetries can feed into the Close-out Amount calculus.
• Action: Keep valuation statements, interest journals, and call logs clean and time-stamped; ensure the determination method under your ISDA Master is aligned with your records.

12) FAQs (2025)

Q1: Does the 2019 appellate decision mean negative interest is never payable under any CSA?
No. It means the standard 1995 English law CSA doesn’t create that obligation. Parties can draft it in (or adhere to a protocol). Many 2016 VM CSAs articulate interest mechanics that handle zero floors or negatives explicitly.

Q2: If my 1995 CSA is silent, can a court imply a term for symmetry?
The appellate court’s reasoning strongly disfavors implication where the text and market materials don’t support it. If you want symmetry, write it in.

Q3: We are an NFC- (non-financial counterparty below clearing thresholds) hedging with a UK dealer. Does any of this change?
Your margin obligations, trading lines, and collateral economics still depend on the annex you actually signed. NFC status doesn’t override the contract.

Q4: If we adopt a 2016 VM CSA now, should we also retrospectively “fix” old trades under the 1995 CSA?
Typically you don’t rewrite historic VM interest for closed periods. You can set a cutover date for interest mechanics and leave history as-is, unless there’s a bilateral commercial reason to restate.

Q5: What’s the simplest way to neutralize the risk without a full repaper?
A short bilateral amendment that (i) floors interest at zero or (ii) defines negative-interest flows cleanly, with aligned benchmark/day-count and FX choices. Keep it tight; avoid accidental scope creep.

13) TRW’s 10-Point Playbook for 1995 CSA Estates

1. Inventory every annex by type, governing law, and counterparty; flag any 2014 Protocol adherence.
2. Classify annexes by interest mechanic (positive-only vs explicit negative treatment).
3. Map product and tenor to annex type; prioritize high-volatility or long-dated pockets.
4. Stress test interest flows under sample negative-rate paths (USD, EUR, GBP).
5. Identify mismatches in back-to-back hedges; propose internal transfer pricing or repaper where basis risk bites.
6. Draft a standardized zero-floor or negative-interest rider (two variants) for quick bilateral adoption.
7. Set ops SOPs: valuation times, call windows, escalation contacts; codify interest journals and audit trails.
8. Board policy: update the Derivatives Use Policy to state the official position on negative interest by annex type.
9. Train treasury/legal/middle office; run a table-top drill simulating a 6-week negative-rate episode.
10. Review annually (or on rate regime alerts); keep a live CSA dashboard.

14) Bangladesh, London, Dubai—Why TRW’s Tri-Hub Model Works

• Dhaka: Local regulatory fit, Bangladesh Bank interfaces, board approvals, lender communications, and audit readiness.
• London: English-law drafting, ISDA negotiation at dealer desks, and real-time handling of UK/EU market shifts.
• Dubai: Liquidity staging in USD/EUR with time-zone overlap; custodian onboarding; contingency routes if London cut-offs collide with Dhaka banking hours.

Our multi-hub practice ensures the legal text and the daily plumbing (calls, interest booking, reconciliations) are consistent—so your documentation isn’t elegant on paper but brittle in operations.

15) Implementation Timeline (Typical)

• Weeks 1–2: Diagnostic & inventory
  Annex census; protocol check; quick stress test; back-to-back mapping.
• Weeks 3–6: Documentation
  Draft/select zero-floor or negative-interest rider; targeted counterparty outreach; negotiate and sign.
• Weeks 6–8: Ops integration
  Update SOPs, day-count, accrual codes; refresh treasury ledger mapping; conduct training.
• Ongoing: Monitoring
  Quarterly dashboard; rate-regime trigger alerts; annual board review; incident playbooks.

16) Case Study (Hypothetical; names generic)

“Rahman Power & Textiles Ltd.” maintained a legacy 1995 CSA with a European dealer for project-linked USD swaps. During a eurozone mini-shock, front-office feared negative EUR depo might reopen an old debate about negative interest on cash VM.

TRW actions:

• Confirmed no negative interest obligation under the 1995 CSA; prepared a relationship note memorializing shared understanding with the dealer.
• Proposed a concise zero-floor rider for prospectively clearer drafting (accepted without pricing change due to long-term relationship).
• Updated SOPs and interest journals; ran a table-top drill.
  Outcome: The six-week episode passed with no disputes; auditors praised the documented policy and ledger clarity.

17) Board-Ready Checklist

• [ ] Inventory annex types (1995 vs 2016 VM; English vs NY law) and protocol adherence.
• [ ] Decide: maintain zero floor or adopt explicit negative-interest language prospectively.
• [ ] Align benchmark, day-count, rounding, MTA with operations.
• [ ] Update Derivatives Use Policy and lender communications.
• [ ] Drill a negative-rate scenario across Dhaka–Dubai–London teams.
• [ ] Install a CSA dashboard for oversight and incident logs.

18) Conclusion: Contract Clarity Beats Market Guesswork

The 2019 appellate ruling provided a durable, commercially sensible answer for a standard document drafted in a different era: under the ISDA 1995 English law CSA, negative interest is not payable by the Transferor of cash collateral—unless the parties say so. In 2025, the right approach is not to litigate metaphysics but to decide your economics and write them down—with interest floors, currencies, day-counts, and dispute mechanics tailored to your treasury’s reality.

TRW will map your annex inventory, design crisp riders, negotiate them in London, stage liquidity and operations in Dubai, and embed governance in Dhaka—so your collateral engine works the same on paper, in ledgers, and in court.

For further internal reading that complements this topic, see:

• Secured Lending & Syndication
• Regulatory (Bangladesh Bank)

Structured Summary Table

Topic	What the Court of Appeal confirmed	Why it matters in 2025	TRW’s practical fix
1995 CSA & negative interest	Standard form does not oblige payment of negative interest by Transferor	Legacy books persist; rate regimes can pivot; close-out math depends on it	Keep legacy position; issue relationship notes; train ops
Paragraph 5(c)(ii)	Clause addresses positive interest; silence on negative flows is telling	Avoid “implied symmetry” claims in disputes	If symmetry desired, draft it expressly
Background materials	User’s Guide + near-contemporaneous best practices show market didn’t intend negative interest	Supports your legal posture; informs audits and controls	Cite in internal policy; include in board packs
Mixed estates (1995 + 2016 VM)	Different annexes = different interest mechanics	Back-to-back P\&L noise in stress	Map books; add zero-floor riders; standardize SOPs
Protocol reliance	2014 Protocol was optional	Not all relationships adhered	Verify adherence; don’t assume; remediate gaps
Ops alignment	Time zones, cut-offs, day-counts drive real cashflows	Reduces disputes; avoids settlement fails	Dhaka–Dubai–London choreography; ledger codes
Governance	Policies must declare the house view	Audit and lender comfort	Update Derivatives Use Policy; quarterly dashboard

Contact TRW Law Firm

Tahmidur Remura Wahid (TRW) Law Firm
Dhaka: House 410, Road 29, Mohakhali DOHS
Dubai: Rolex Building, L-12 Sheikh Zayed Road
London (UK): 330 High Holborn, London WC1V 7QH, United Kingdom

Call Us: +8801708000660 / +8801847220062 / +8801708080817
Email: info@trfirm.com | info@trwbd.com | info@tahmidur.com

How we can help—next steps

Send us your CSA inventory (even a simple list of counterparties and annex dates). We’ll return a one-page gap memo flagging where you stand on negative interest, and a draft zero-floor or negative-interest rider tailored to your treasury and documentation stack.