Why NPL portfolios cannot be valued from the headline balance

The unpaid balance of a non-performing loan portfolio is an accounting and contractual reference point. It is not its investment value.

For a prospective purchaser, value depends on the cash that can realistically be recovered, the cost of producing that recovery, the time required and the uncertainty around each assumption. Two portfolios with the same gross book value can have very different economics because their collateral, lien priority, documentation, borrower circumstances and enforcement routes differ.

A useful working expression is:

Portfolio value = probability-weighted recoveries, discounted for time, less servicing, legal, enforcement, tax, holding and capital costs.

This is not a substitute for a transaction-specific model. It is a reminder that “discount to face value” is not a measure of cheapness unless the underlying recovery assumptions are known.

A seven-stage NPL due-diligence process: mandate, data integrity, legal rights, collateral, servicing, scenarios and investment decision.

The regulatory and data context in Europe

European NPL secondary markets now operate within a more standardised framework than they did after the global financial crisis.

Directive (EU) 2021/2167 establishes a framework for credit servicers and credit purchasers, with the dual aim of supporting secondary markets and protecting borrowers’ rights. National implementation still matters: licensing, notification, servicing and conduct requirements should be verified in every relevant jurisdiction.

For qualifying sales by EU credit institutions, Commission Implementing Regulation (EU) 2023/2083 specifies standard transaction data templates. The required information covers counterparties, credit agreements, relationships, collateral, guarantees, enforcement and historical collections.

The templates reduce information asymmetry, but they do not eliminate it. A populated field may be stale, inconsistent or based on an assumption unsuitable for the buyer’s strategy. Missing information can itself be material. Due diligence must therefore test the data rather than merely receive it.

Stage 1: Define the asset and the mandate

Before analysing the data tape, establish exactly what is being sold and what the buyer is permitted and equipped to own.

Questions include:

• Is the transaction a sale of loans, creditor rights, shares or assets?
• Are accrued interest, guarantees, hedges and ancillary rights transferred?
• Are any loans syndicated, participated or subject to intercreditor arrangements?
• Does the buyer require a locally authorised credit servicer?
• Are consumer, residential or vulnerable-borrower protections relevant?
• Are there restrictions on foreign ownership of the collateral or creditor rights?
• What representations, warranties and post-closing assistance will the seller provide?

This stage prevents a common error: building a recovery model for rights the purchaser will not actually receive or cannot practically exercise.

Stage 2: Test data-tape integrity

The first analytical task is not valuation. It is reconciliation.

The loan-level tape should be tested against control totals and supporting records. Gross balance, principal, accrued interest, fees, cash collections and collateral values should reconcile across portfolio, borrower, loan and collateral tables. Unique identifiers should map consistently where one borrower has several loans or one property secures several facilities.

Useful tests include:

1. Completeness: What proportion of mandatory and investment-critical fields is populated?
2. Freshness: How old are valuations, borrower contacts, legal-status updates and collection records?
3. Internal consistency: Do default dates, days past due and legal-stage fields agree?
4. Relationship integrity: Can each loan be linked to the correct borrower, guarantee, collateral and enforcement file?
5. Duplicate detection: Are borrowers, loans or properties repeated under variant identifiers?
6. Outlier review: Which balances, valuations, interest rates or recovery histories fall outside plausible ranges?

The EBA framework is particularly useful because it requires granular information and relationship mapping. But its scope is a regulatory baseline, not a guarantee that every data point needed for a specific acquisition is present.

Stage 3: Establish legal rights and enforceability

Real estate value is relevant only if the security is valid, enforceable and connected to the loan being acquired.

Legal review should consider:

• evidence of debt and executed facility documentation;
• identity and capacity of the obligor and guarantors;
• validity, perfection and registration of mortgages or charges;
• lien ranking and prior-ranking debt;
• amendments, waivers and forbearance agreements;
• limitation periods and procedural deadlines;
• insolvency stays or restructuring plans;
• litigation, borrower defences and conduct risks;
• assignment formalities and borrower-notification requirements; and
• the expected route, duration and cost of enforcement.

The legal workstream should feed the financial model directly. An enforceable first-ranking mortgage with a current valuation is not economically equivalent to a contested junior claim, even when both records show the same collateral address.

Stage 4: Re-underwrite the collateral

Seller valuations are inputs, not conclusions.

The RICS Valuation – Global Standards provide a recognised professional framework for valuation instructions, investigations, methods and reporting. For NPL underwriting, the buyer should understand the basis of value, valuation date, inspection scope, special assumptions and limitations attached to each figure.

The collateral review should cover:

Title and physical identity

Confirm that the property inspected and valued is the property over which security was granted. Check title boundaries, tenure, access, easements and any unregistered interests identified by local counsel.

Income and occupancy

Review leases, rent rolls, arrears, incentives, break options, expiries and tenant concentration. Contracted rent is not automatically recoverable rent, and passing rent is not automatically market rent.

Capital expenditure

Estimate urgent repairs, deferred maintenance, tenant works, safety expenditure, planning obligations and energy-performance improvements. These costs affect both the amount and timing of recovery.

Market liquidity

A reported market value does not state how quickly the asset can be sold or what discount may be required in a constrained process. Comparable transactions should be relevant to the property type, location, condition and likely sale route.

Environmental and sustainability risks

The fourth edition of the RICS ESG and sustainability standard for commercial property valuation, effective from April 2026, reinforces the consideration of significant sustainability factors and related cost information in valuation. For a distressed asset, energy performance and physical climate risks may influence lettability, financeability, insurance and future expenditure.

Stage 5: Design the servicing and workout strategy

Recoveries do not occur automatically after closing. They are produced through servicing, negotiation, litigation, asset management and sale execution.

The operating plan should identify:

• who will service the loans and under what authorisation;
• the capacity and experience of the servicer by jurisdiction and asset type;
• borrower-contact and conduct procedures;
• delegated settlement authority;
• legal-panel coverage and fee arrangements;
• reporting frequency and data ownership;
• property-management capability after possession; and
• incentives linking servicer remuneration to sustainable net recovery rather than activity alone.

Historical collections should be segmented by source. Voluntary payments, discounted settlements, asset sales, enforcement proceeds and cures have different implications for future performance.

Stage 6: Model recoveries as scenarios, not a single forecast

A base case alone creates false precision. Each material exposure should have alternative routes with explicit probabilities, timing and costs.

A practical structure may include:

Consensual case

The borrower refinances, sells the asset, agrees a discounted payoff or enters a viable restructuring. This route may be quicker and preserve more value, but it depends on cooperation and funding capacity.

Enforcement case

The creditor exercises security and recovers through court or out-of-court procedures. The model should reflect jurisdiction-specific steps, appeal risk, auction mechanics, taxes and legal costs.

Insolvency case

Recovery occurs through an insolvency or restructuring process. Priority, administrator costs, competing creditors and stays may affect both value and timing.

Asset-control case

The buyer or its vehicle obtains control of the property, funds works, stabilises income and exits later. This can increase gross recovery while introducing operating, capital and market risk.

For each route, the model should state:

• gross proceeds;
• probability of achieving them;
• time to cash receipt;
• legal and servicing costs;
• taxes and transfer costs;
• required protective advances or capital expenditure;
• interim income and operating costs; and
• residual uncertainty at the assumed exit date.

Sensitivity analysis should focus on variables that genuinely drive value: enforcement time, collateral proceeds, capital expenditure, cure probability and discount rate. Changing every input by the same percentage can obscure rather than clarify risk.

Stage 7: Convert uncertainty into bid discipline

The final investment decision should preserve a traceable link from source data to price.

An investment committee should be able to see:

• which data were verified and which were assumed;
• how missing fields affected sampling or pricing;
• the proportion of value concentrated in the largest borrowers or properties;
• which legal opinions are conditions to closing;
• how much of the return depends on asset appreciation rather than execution;
• what liquidity and follow-on capital are required; and
• which seller protections survive closing.

Bid protections can include price adjustments, exclusions, conditions precedent, escrow arrangements, repurchase provisions or targeted warranties. Their availability depends on transaction dynamics and bargaining power, but uncertainties should be allocated consciously rather than lost inside a blended discount.

Common analytical errors

Treating gross book value as economic value

The outstanding balance may exceed, equal or fall below recoverable value. It is not the correct valuation anchor by itself.

Applying one recovery rate across the portfolio

Blended assumptions can hide concentration, lien and jurisdiction differences. Segmentation should reflect the factors that actually change recovery behaviour.

Using collateral value without deduction for time and cost

Gross property value is not net present recovery. Enforcement, tax, capex and delay can materially change the result.

Assuming missing data are neutral

Data gaps may be random, but they can also be concentrated in older, disputed or operationally difficult cases. The effect should be tested through sampling and conservative assumptions.

Separating legal due diligence from the model

Legal conclusions determine recovery routes and timing. They should alter cash-flow assumptions rather than remain in a standalone report.

Conclusion

Real-estate-backed NPL investing is an exercise in connecting rights, data, property and execution. The most useful due-diligence process moves from tape integrity to legal enforceability, collateral re-underwriting, servicing design and probability-weighted recoveries.

The objective is not to remove uncertainty; that is rarely possible. It is to identify uncertainty, decide who controls it and pay a price consistent with the capital and time required to resolve it.

Molyneux Capital supports investors across this lifecycle through private deal origination, acquisition strategy, market analysis, valuation, feasibility work and asset management.

Related: Explore Molyneux Capital’s services or contact us privately.

Sources and methodology

This article is an original analytical framework based on public regulatory and professional materials. It is not copied from any source. Sources were checked on 25 August 2026.

• European Union, Directive (EU) 2021/2167 on credit servicers and credit purchasers.
• European Union, Commission Implementing Regulation (EU) 2023/2083.
• European Banking Authority, Implementing Technical Standards on NPL transaction data templates.
• European Banking Authority, NPL transaction templates and supporting materials.
• European Central Bank, Guidance to banks on non-performing loans.
• RICS, Valuation – Global Standards.
• RICS, ESG and sustainability in commercial property valuation, fourth edition.

Important notice

This article is provided for general information only. It does not constitute investment, legal, tax, credit-servicing or valuation advice. NPL acquisition and servicing requirements vary by jurisdiction. Specialist legal, regulatory, tax, valuation and servicing advice should be obtained for each proposed transaction.