Consumer Duty Fair Value Assessment: A Practitioner Methodology

TL;DR — The Consumer Duty's price and value outcome (PRIN 2A.4) requires every firm to evidence that the price the retail customer pays is reasonable relative to the benefits of the product. "Reasonable" is undefined on purpose. This piece sets out a methodology, the cohorts to segment, the evidence to retain, when to repeat the assessment, and what an exemplar fair value matrix looks like.

Table of contents

• What the rule actually says
• How the FCA has framed "value"
• A four-step methodology
• Step 1: Define the product and its target market
• Step 2: Identify benefits and price components
• Step 3: Segment by cohort
• Step 4: Benchmark and conclude
• Exemplar matrix
• Evidence the FCA will look for
• When to repeat the assessment
• Common pitfalls
• FAQ

What the rule actually says

PRIN 2A.4 requires that:

• The price paid by a retail customer for a product is reasonable relative to the benefits.
• The firm assesses this before the product is sold and at appropriate intervals afterwards.
• Manufacturers must communicate the outcome of the assessment to distributors in sufficient detail that distributors can in turn assess whether their own pricing is consistent with fair value.

The Duty applies in addition to product-specific value rules — the General Insurance Pricing Practices Rules in ICOBS 6B, the platform charges rules in COBS, and the cash savings rules emerging from the FCA's 2023 review. Where rules overlap, the higher standard applies.

FS22/5 is explicit that fair value is not the same as "cheapest." A premium product can be fair value. A cheap product can fail fair value if the benefits do not match the price. The test is reasonableness, evidenced.

How the FCA has framed "value"

The FCA's portfolio reviews since 2023 have given us a working definition. Value, in the FCA's frame, is the sum of:

• The product's intrinsic benefits (the cover, the rate, the functionality)
• Service benefits (claims handling, customer support, digital tools)
• Ancillary benefits the firm has chosen to bundle (rewards, signposting, education)

Set against:

• The headline price
• All fees, charges and commissions in the distribution chain
• Non-monetary costs — time, data, complexity — borne by the customer

The reasonableness test asks whether, in the round, a customer of this type would consider the bundle to deliver value relative to what they could get elsewhere.

A four-step methodology

This is the structure we recommend. It is the same structure most large firms have converged on after two cycles of FCA engagement.

Step 1: Define the product and its target market

• One product per assessment. Do not bundle. A "savings account" is not a product for fair value purposes; "Easy Access Saver Tier 3 (legacy book)" is.
• Target market statement, refreshed for the assessment year. State the customer types the product is designed for, with explicit exclusions (e.g., "not suitable for customers who require regular income from this product").
• Distribution map. Direct, via aggregator, via broker, via AR. Each channel may need separate treatment if economics differ.

Step 2: Identify benefits and price components

Build two tables.

Benefits table: - Intrinsic features (rate, sum insured, FX margin, fund choice) - Service features (claims SLA, branch access, app functionality) - Ancillary features (cashback, partner rewards, financial education) - For each, a measurable quantification where possible. "App functionality" is not measurable. "Mobile cheque deposit, biometric login, instant card freeze" is.

Price components table: - Headline price (rate, premium, fee) - Recurring charges (admin fee, custody fee, platform fee) - Transaction charges (FX, withdrawal, ATM, dealing) - Behaviour-driven charges (late fee, overdraft, missed-payment) - Distribution costs (commission, marketing acquisition cost) — if relevant to the customer - Non-monetary costs — time-to-onboard, document burden, exit friction

Step 3: Segment by cohort

This is where most fair value assessments fail FCA scrutiny. An aggregate "this product is fair value" conclusion masks cohorts for whom it is not. At a minimum, segment by:

• Acquisition channel (direct, aggregator, broker, AR)
• Tenure (new business, 1-2 years, 3-5 years, 5+ years)
• Vulnerability characteristics
• Pricing tier (where risk-based or behaviour-based pricing applies)
• Engagement level (engaged vs disengaged, where the product economics differ between the two — particularly relevant for cash savings)
• Product variant (legacy book vs current book, closed product vs open)

Calculate the value-to-price ratio for each cohort. Compare them. Where cohorts diverge materially, the assessment must explain why and conclude whether the divergence is justifiable.

Step 4: Benchmark and conclude

• Identify the relevant comparator set. For most retail products, this is the top quartile of equivalent products in the open market. For specialist products, a defined peer group.
• Document the data source and date. The FCA will want to know whether the benchmark is current.
• Reach a conclusion per cohort:
• Fair value
• Fair value with conditions (with named actions)
• Not fair value (with named remediation)

A "fair value with conditions" outcome must have action owners and dates. A "not fair value" outcome must have a remediation plan and an interim customer mitigant.

Exemplar matrix

A simplified matrix for a credit card might look like:

Cohort	Benefits score (0-10)	Total cost (£/year, indicative)	Comparator (top quartile)	Conclusion
New customer, prime, 0% intro	8.5	£24 (annual fee)	£0-£36	Fair value
New customer, near-prime	7.0	£180 (interest + fees)	£150-£220	Fair value
Long tenure, post-intro, prime, low utilisation	6.0	£24	£0 (no fee comparators dominant)	Fair value with conditions — review fee retention vs reward redemption
Long tenure, persistent revolver	5.0	£640 (interest + fees)	£420-£560	Not fair value — remediation in train: rate review, persistent debt rule application
Vulnerable customer flagged	Variable	Variable	n/a	Reviewed individually; specialist team

The point of the matrix is not to produce a single "score" but to force the firm to confront cohorts where value diverges. The persistent-revolver row is where the FCA's eye will go first.

For worked examples of these patterns appearing in BNPL, motor insurance, savings and wealth, see our FCA Consumer Duty examples.

Evidence the FCA will look for

In a section 166 or supervisory engagement, expect requests for:

1. The fair value assessment document for each material product, dated.
2. The minutes of the committee that approved it.
3. The underlying data — typically a CSV extract or a model output — that supports each cohort conclusion.
4. The MI run since the last assessment showing whether the conclusions remain valid.
5. The customer communications generated as a result (rate-rise notices, product-switch invitations, fee waivers).
6. The distributor pack issued to intermediaries, where the firm is a manufacturer.
7. The board paper that summarised the conclusions, which should also surface in the Consumer Duty board report.

If any of these are not available within five working days, the firm has an evidentiary problem irrespective of the substantive answer.

When to repeat the assessment

The Duty requires assessment "at appropriate intervals." Practice has settled around:

• Annually at minimum, for all material products.
• On any material change — pricing change, target market change, new distribution channel, material change in benefits.
• On any market shock — material rate moves, regulatory change, sustained inflation deviation, peer competitor pricing change.
• On any trigger from outcomes data — spike in complaints, cohort drift in default rates, sustained drop in persistency.

Smaller firms often try to run the cycle on a calendar basis only. We would caution against that — the FCA's expectation is that the firm responds to evidence, not just to the calendar.

Common pitfalls

• Treating fair value as a price-only test. Benefits matter equally. A product can be fair on price alone but fail on benefits delivery (e.g., a cheap policy with a 14-week claims SLA).
• Aggregating cohorts to reach a comfortable conclusion. The cohort split is the test.
• Stale benchmarks. A benchmark from 18 months ago is not credible in a rate-moving environment.
• No action from a "fair value with conditions" outcome. The conditions must be tracked and closed, or the conclusion is not credible.
• Distributor packs that are too thin. Manufacturers must give distributors enough to discharge their own duty; a target-market summary alone is not enough.
• Treating loyalty as inherently positive. Long tenure with a disengaged book can be a flag, not a credential. The FCA's framing has shifted on this and most firms have not caught up.

For firms shopping for tooling that can ingest the underlying data and surface cohort drift automatically, we look at the market in FCA Consumer Duty compliance software.

FAQ

Is a fair value assessment the same as a value-for-money assessment under COBS 19? Related but not identical. The COBS 19 VFM rules for workplace pensions are more prescriptive on metrics. The Duty's fair value assessment is broader in scope but less prescriptive in form. Most firms run the COBS 19 process and use it as one input to the Duty assessment.

Does fair value apply to professional clients? Generally no — the Duty is for retail customers. But some products that are sold to retail and professional clients in parallel will need a retail-cohort fair value view.

How granular do cohorts need to be? Granular enough that material divergence is visible. If the firm cannot answer the question "do vulnerable customers experience the same value as non-vulnerable customers?" the cohorting is not granular enough.

Can a firm rely on a third-party benchmark? Yes, provided the source is credible, current, and the firm has documented why the benchmark is appropriate. Aggregator data, FCA-published market data and reputable industry surveys all qualify in principle.

What happens if the firm concludes a product is not fair value? The Duty requires remediation. That can include repricing, withdrawal, customer migration, redress for past harm, or some combination. The remediation plan must be specific and time-bound.

Does fair value cover commission? Yes, where commission affects the customer's price or the distribution incentive. Manufacturers must consider whether the commission structure they pay influences distribution behaviour in a way that affects customer value.

How does AI / model risk interact with fair value? Where pricing or affordability is driven by a model, the firm must satisfy itself that the model is not producing systematically worse outcomes for protected or vulnerable groups. Outputs are the test, not just inputs.

Get a baseline of your fair value risk

Most firms know which products carry the most fair value risk; few have it quantified against the patterns the FCA is calling out in supervision. Sedric's free Enforcement Risk Scorecard is a 12-question diagnostic that benchmarks your firm against the most common fair value and Duty findings — savings pass-through, loyalty premium, distribution-chain gaps, vulnerable cohort outcomes. You get a written risk profile within 24 hours. Take the Enforcement Risk Scorecard.