SRT Market: Systemic Risk Profile (Europe)

Synthetic securitisation designed to generate regulatory capital relief without loan sale. Systemic risk concentrates in protection seller liquidity, supervisory recognition, and procyclical feedback loops.

As of2026‑01‑28 RegionEurope CategoryStructured Credit / SRT StatusGrowing issuance; disclosure limited
Non‑bank intermediation System overlay
Market size (proxy outstanding)
~€500bn
Estimated
Annual issuance (2024)
€141.5bn
Estimated
Capital relief (2022)
€5.6bn
Confirmed
Protection seller concentration (top 2)
75%
Confirmed (survey)ESRB (ECB survey, Jun 2023)

Executive Summary

Badges show provenance

Europe’s Significant Risk Transfer (“SRT”) market is a private, bank‑to‑investor synthetic securitisation channel designed to generate regulatory capital relief without loan sale. Issuance has trended higher (2024: €141.5bn) and the outstanding stock is not centrally disclosed; a rolling issuance proxy implies ~€500bn scale. Estimated

The trade is capital optimisation with a regulatory gate: banks retain loans, but transfer defined loss slices via funded CLNs or unfunded CDS/guarantees. In stress, capital relief can become uncertain if supervisors re-assess whether “significant” risk transfer criteria are met, creating cliff risk through RWA restoration. Modeled

Concentration is a systemic feature: credit protection providers are predominantly credit funds (~45%) and asset managers (~30%), ~75% combined (ESRB/ECB survey, Jun 2023). Confirmed (survey) These providers can be procyclical via margin, repo, and redemption mechanics—so liquidity risk can dominate credit risk in fast moves. Modeled This concentrates tail risk in entities whose funding can evaporate under margin/redemption pressure. Modeled

Non‑bank intermediation concentration
This refers to concentration of SRT risk in non‑bank sellers with procyclical liquidity structures. The dominant marginal risk absorbers are non-banks with heterogeneous liquidity structures (repo/margin/redemptions), creating procyclical funding stress. Modeled
€500bn market, limited disclosure
There is no position-level public tape for SRTs; proxy sizing relies on rolling issuance sums and assumed deal life. Opacity is a systemic variable (collateral and funding chains are hard to map). Estimated
Liquidity risk > credit risk in stress
Spread widening can trigger MTM losses and collateral calls before realized defaults exhaust tranches, forcing deleveraging in protection sellers. Modeled
Regulatory arbitrage dynamics
Capital relief depends on supervisory acceptance. Interpretation shifts can cause non-linear “capital cliff” effects (RWA restoration) and spread repricing. Modeled

Definitions

Quick reference

SRT vs synthetic securitisation: SRT is the regulatory test (risk has moved). Synthetic securitisation is the structure often used to achieve it (risk slices transferred while loans remain on balance sheet).

Funded CLN vs unfunded guarantee/CDS: Funded CLNs post cash up-front into an SPV; losses write down principal. Unfunded protection can introduce collateral/CSA dynamics and performance risk under stress.

Protection seller: The counterparty taking the credit-risk slice (fund/asset manager/insurer/supranational). This note uses “seller” as shorthand for “protection seller.”

Capital relief vs RWA reduction: Capital relief is the ratio impact; RWA reduction is the mechanical driver. Relief is conditional on supervisory recognition of “significant” risk transfer.

Recognition cliff: supervisory re-assessment can restore RWAs despite unchanged underlying loan performance.

Operating Snapshot

Issuance aggregates ECB/ESRB

Issuance Trend

Estimated
MetricValueProvenance
SRT issuance — 2021€68.1bnEstimated
SRT issuance — 2022€96.0bnEstimated
SRT issuance — 2023€113.4bnEstimated
SRT issuance — 2024€141.5bnEstimated
SRT issuance — 2025 (YTD)€91.1bnEstimated
Source: AFME aggregation (methodology via SCI/RTRA). Estimated

Outstanding Proxy + Capital Relief

Derived
ItemValueProvenance
Rolling 4Y proxy (2022–2025YTD)€442.0bnDerived
Rolling 5Y proxy (2021–2025YTD)€510.1bnDerived
Working market size anchor€500bnEstimated
Capital relief (record, 2022)€5.6bnConfirmed
Proxy assumes average effective risk-transfer life of ~4–5 years (varies by asset class and triggers). Proxy ignores amortisation and early unwind; intended as order-of-magnitude.Modeled
Synthetic securitisations delivered ~70–90% of total capital relief over 2018–2022 (Euro area). Confirmed

Market Topology

System map

SRT Network (3-level)

Derived view
European Banks (Originators) Corporate/SME/CRE loan books · Capital & RWA constraint
SRT Structures SPV / Issuer
Funded CLN Mezz (funded)
Unfunded protection (guarantee/CDS) Often first-loss / senior protection
Credit funds 45% Confirmedsurvey estimate
Asset managers 30% Confirmedsurvey estimate
Supranationals 15% Confirmedsurvey estimate
Insurers 5% Confirmedsurvey estimate
Pension funds 5% Confirmedsurvey estimate
Concentration
Top two seller types (credit funds + asset managers) represent ~75% of protection sold, concentrating liquidity and funding-structure risk outside the banking perimeter. Confirmed (survey)
Source note (seller split)
Investor-type shares are taken from ESRB Occasional Paper No 23 (survey ECB data, June 2023; Table 2). The paper reports estimated shares of total SRT market credit protection sold; scope and category definitions can vary (euro area focus; funded vs unfunded; classification overlap). Confirmed

Scenario Analysis (A/B/C)

Assumptions Computed

Scenario Comparison

Externally specified scenario inputs
Metric Scenario A
Benign
Scenario B
Cycle stress
Scenario C
Tail downturn + seller failure
Expected
Illustrative weight Modeled 40% 35% 25% 100%
Portfolio loss rate Derived 0.675% 2.75% 14.0%
First-loss recovery Derived 91.56% 65.63% 0.00%
Mezz recovery Derived 100.00% 100.00% 50.00%
Senior recovery Derived 100.00% 100.00% 100.00%
RWA restoration (system overlay) Modeled €0bn €100bn €300bn €110bn Derived
Protection performance event ModeledNoLowHigh
Weights are stylised for intuition (not fitted to historical default series or base rates). Scenario C combines deep credit stress with seller performance/regulatory intervention mechanisms.Modeled

Why SRT Is Procyclical

Mechanism map

Two feedback loops

Diagram
Spreads widen Margin calls Forced selling Spreads widen ↺
Supervisor tightens Recognition risk RWA snapback Capital constraint Credit tightening ↺
Wrong‑way risk (“circles of risk”)
If banks provide financing (repo/lines) to the same funds selling protection, stress is wrong‑way: the bank is exposed to the seller at the same time it depends on that seller for capital relief. Modeled

Limits of Inference

Disclosure constraints
  • No public, position-level transaction tape for SRTs; most deals are private.
  • Survey evidence is partial and category definitions vary across jurisdictions and institutions.
  • Funded vs unfunded protection can behave very differently under stress (principal write-down vs collateral/CSA dynamics).
  • Supervisory recognition is judgment-based; guidance can evolve, creating “cliff” effects in capital relief.

These constraints are why the note emphasizes order-of-magnitude sizing and mechanism mapping over false precision. Modeled

Recovery Waterfall

Representative tranches

Scenario A

Benign
First-loss91.56%
 
Mezz100.00%
 
Senior100.00%
 

Scenario B

Cycle stress
First-loss65.63%
 
Mezz100.00%
 
Senior100.00%
 

Scenario C

Tail + seller failure
First-loss0.00%
Zero
Mezz50.00%
 
Senior100.00%
 

Event Timeline

Catalysts
2026 Potential tightening
2026–27 Credit cycle turn risk

Risk Scoring Matrix (0–100)

Modeled
Entity
Credit
Liquidity
Counterparty
Regulatory
Contagion
Total
European banks
60
45
55
65
70
59
Insurance sellers
55
50
40
55
60
52
Hedge fund sellers
75
85
60
45
80
69
Pension fund sellers
55
30
35
35
50
41
CLO / levered credit channel
70
75
55
45
75
64
Total is the unweighted average of the five dimension scores (0–100). Modeled
CLO/levered credit is included here as a correlated funding channel (spread/margin loop), not as a primary SRT protection seller category.Modeled