Structuring a Recovery Plan Under SS9/17 Without Resolvability Drag
A practical guide to building a Recovery Plan submission that evidences credible, executable recovery options while avoiding content that undermines the firm's resolvability position. Covers the sequencing, stakeholder choreography, and judgement calls that separate a plan the PRA accepts from one that triggers follow-up.
The hardest part of a Recovery Plan is not demonstrating options exist. It is demonstrating they are credible, timely, and executable under stress, without inadvertently signalling that your preferred resolution strategy would fail. SS9/17 asks firms to show recovery capacity; the Resolvability Assessment Framework asks whether the firm could fail safely. These two documents are read together by the same supervisors, and inconsistencies between them are where submissions come unstuck.
Start with the indicator framework, not the options menu
Most weak plans begin with a catalogue of recovery options and bolt indicators on afterwards. Strong plans work the other way. The indicator framework, capital, liquidity, asset quality, profitability, market based, and macroeconomic, is the spine of the document. Each indicator needs a calibrated early warning level and a recovery level, with evidence that the thresholds would trigger before the firm loses optionality. If your liquidity recovery indicator fires at the same point your LCR breach is reported to the PRA, the framework is not doing its job.
The common failure is calibration that looks prudent in isolation but, when back-tested against the 2020 or 2023 stress episodes, would not have triggered management action in time. Run that back-test and include it. Supervisors notice when it is missing.
Build options around capacity, not optimism
Each recovery option needs four things evidenced: quantified capacity (in pounds and basis points of CET1 or days of liquidity survival), time to execute, dependencies, and the specific governance step that authorises execution. "Sale of non-core portfolio" is not an option. "Sale of the X portfolio, carrying value Y, buyer universe of Z identified counterparties, estimated execution 90 to 120 days, CET1 impact +45bps, requires board approval and PRA notification under SS9/17" is an option.
Stress each option against the scenario it is meant to address. A capital raise is not credible in an idiosyncratic scenario where the market has lost confidence in the firm. Disposals are not credible in a system wide scenario where buyers are retrenching. Showing that you have thought about this, and still have residual capacity, is what credibility looks like.
The resolvability trap
This is where senior teams most often create problems for themselves. If your recovery plan relies on intragroup support, liquidity upstreaming, or operational services that would be disrupted in resolution, you are telling the Bank of England that your resolvability work has gaps. The same applies to options that assume access to central bank facilities on terms that would not survive a loss of market confidence.
The fix is not to hide the dependency. It is to evidence that the option works in recovery (where the firm is a going concern) and to acknowledge, in the resolvability documentation, that the same lever is not available post resolution weekend. Consistency across documents matters more than cleverness within one.
Governance evidence the board actually owned it
The PRA reads minutes. If the plan arrives with board approval but no record of challenge, that is a flag. Show where the board pushed back on scenario severity, where indicator calibration was contested, where management revised assumptions. A playbook that has been dry-run, with findings incorporated, carries more weight than one approved on the nod.
Scenarios that stress the actual business
Generic scenarios produce generic plans. The idiosyncratic scenario should name the specific exposure, concentration, or conduct event that would most plausibly damage your firm. The combined scenario should compound in ways that reflect real correlations in your book. If your scenario narrative could be copy pasted into a peer's plan, it is not doing its job.
The decision point
Before submission, ask one question: if the PRA reads this plan and our resolvability submission side by side, do they tell the same story about how this firm would behave under stress? If the answer is no, or you are not sure, that is the work to do before the document goes in. Everything else is presentation.
Frequently Asked Questions
How many recovery options is enough?
There is no fixed number. What matters is total quantified capacity under each scenario, with sufficient redundancy that the loss of the two largest options still leaves the firm recoverable. For most mid-sized firms this means six to ten substantive options, not twenty tokens.
Should we include options we are unlikely ever to use?
Only if they are genuinely executable. Including an option you know would fail in practice, to pad capacity, is the fastest way to lose credibility with your supervisor. Better to show fewer options, fully evidenced.
How do we handle group dependencies for a UK subsidiary?
Be explicit about what parental support is committed, what is discretionary, and what assumptions underpin availability in stress. Letters of support that have not been tested in a real downturn should be described as such.
How often should the plan be genuinely rewritten versus refreshed?
Annual refresh is the minimum. A full rewrite is warranted when the business model shifts materially, after a significant stress event in the sector, or when supervisory feedback indicates the structure itself is not working. Rolling forward the same document for three years is visible and unhelpful.
What triggers a resolvability concern from the recovery plan itself?
Most commonly: recovery options that depend on capabilities the firm has told the Bank it would retain in resolution, timelines that imply the firm would not identify stress early enough for resolution authorities to act, or indicator calibration that defers action past the point of no return.
Frequently asked questions
How many recovery options is enough?
There is no fixed number. What matters is total quantified capacity under each scenario, with sufficient redundancy that the loss of the two largest options still leaves the firm recoverable. For most mid-sized firms this means six to ten substantive options, not twenty tokens.
Should we include options we are unlikely ever to use?
Only if they are genuinely executable. Including an option you know would fail in practice, to pad capacity, is the fastest way to lose credibility with your supervisor. Better to show fewer options, fully evidenced.
How do we handle group dependencies for a UK subsidiary?
Be explicit about what parental support is committed, what is discretionary, and what assumptions underpin availability in stress. Letters of support that have not been tested in a real downturn should be described as such.
How often should the plan be genuinely rewritten versus refreshed?
Annual refresh is the minimum. A full rewrite is warranted when the business model shifts materially, after a significant stress event in the sector, or when supervisory feedback indicates the structure itself is not working. Rolling forward the same document for three years is visible and unhelpful.
What triggers a resolvability concern from the recovery plan itself?
Most commonly: recovery options that depend on capabilities the firm has told the Bank it would retain in resolution, timelines that imply the firm would not identify stress early enough for resolution authorities to act, or indicator calibration that defers action past the point of no return.
Related guides
How to Structure an Operational Resilience Self-Assessment Under PRA SS1/21
A practical guide to building a self-assessment that evidences impact tolerances with intellectual honesty, including how to treat mapping gaps as findings rather than hide them. Readers will finish with a clear structure, a credible approach to residual weaknesses, and a sharper view of what the PRA actually expects boards to own.
How to Structure a MIFIDPRU ICARA That Evidences Adequate Own Funds
A practical guide to building an ICARA submission under FCA IFPR that demonstrates genuine capital adequacy and sound risk judgement. Readers will understand how to sequence the assessment, calibrate own funds and liquid assets credibly, and engage supervisors in a way that stands up to challenge.
How to Structure an FCA Business Model Assessment Response
This guide explains how to construct a Business Model Assessment (BMA) response that credibly evidences strategic viability and prudent risk ownership. Readers will learn how to sequence the narrative, align financial projections with stated strategy, and address weaknesses in a way that demonstrates genuine control rather than defensiveness.
How to Structure a Pillar 3 Disclosure That Protects Competitive Position
A practical guide to building a CRR Pillar 3 disclosure that meets market transparency expectations in full while handling genuinely sensitive information through the mechanisms the framework provides. Readers will finish with a clearer view of how to sequence drafting, apply the proprietary and confidential carve-outs properly, and defend disclosure judgements to auditors, the PRA, and the board.
Structuring a Section 165 Response That Protects Privilege
A practical guide to responding to FCA Section 165 information requests in a way that meets the firm's statutory disclosure duties in full while preserving legal professional privilege. Readers will understand how to scope, sequence, and document a response that is both credible to the supervisor and defensible in later proceedings.
Where a specific question needs an outside answer, quickly
Polar Insight's Expert Network connects leadership teams with practitioners who can speak to a precise regulatory, commercial, or stakeholder question before a decision is finalised.
Explore Expert Network