Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Asset Valuation and Impairment in Insurer Balance Sheets

Abstract. The economic function of an insurer's assets is to finance contingent obligations whose amounts, timing, and sensitivity to market conditions differ across contracts and business lines. Measurement of those assets must therefore be analyzed simultaneously as an accounting classification problem, a problem of estimating future contractual cash receipts, a market-valuation problem, and a component of the insurer's capacity to meet claims. This monograph examines the mechanisms and consequences of asset valuation and impairment in insurance balance sheets through the distinct lenses of IFRS financial reporting, European prudential valuation, United States statutory accounting, and economic risk measurement. It distinguishes amortized cost from current fair value; changes in interest rates from deteriorating creditworthiness; accounting expected credit losses from market-implied credit and liquidity premiums; and valuation uncertainty from liquidity generated by actual transactions. The analysis reconstructs the IFRS 9 business-model and cash-flow classification tests, IFRS 13 market-participant valuation logic, relevant IAS 36 and IAS 40 boundaries, and the interaction with IFRS 17 insurance contract measurement. It compares these rules with Solvency II valuation and the prudential objectives of U.S. statutory accounting, examines private credit and structured assets, and derives auditable illustrative scenarios for bond valuation, expected credit losses, asset shocks, and net-balance-sheet sensitivity. The principal conclusion is that the reliability of insurer balance sheets depends on transparent reconciliation of measurement objectives, asset-specific evidence, liability sensitivities, and liquidity assumptions. Apparently stable accounting values can coexist with material changes in economic value, and immediate fair-value losses can coexist with offsetting movements in insurance liabilities. Credible valuation therefore requires instrument-level classification, scenario-dependent cash-flow evidence, independent challenge of unobservable inputs, and explicit bridges between reported and prudential measures. Numerical examples are hypothetical teaching constructions, not empirical estimates or jurisdiction-specific capital calculations.

Keywords: Insurance finance; insurer balance sheet; asset valuation; impairment; IFRS 9; IFRS 13; expected credit losses; amortized cost; fair value; FVOCI; FVTPL; private credit; valuation uncertainty; Solvency II; statutory accounting; asset-liability management; liquidity risk; credit spread; prudential supervision.

Introduction

An insurer accepts obligations in one period and finances their discharge across uncertain future periods. Its investment portfolio supports payments to policyholders, collateral arrangements, operating liquidity, capital requirements, and, where relevant, participating or investment-linked benefits. The valuation of that portfolio enters reported profit, equity, solvency margins, investment limits, and managerial decisions. Each result depends on the measurement question that has been posed. A contractual receivable carried at amortized cost answers a different question from the hypothetical proceeds of an orderly sale. An expected-credit-loss allowance answers a different question from a market discount demanded for illiquidity or uncertainty. Their differences are economically informative when traced to the governing measurement bases.

The accounting architecture starts with the classification rules in IFRS 9, whose treatment of debt investments depends on both how the portfolio is managed and the nature of contractual cash flows. The valuation architecture of IFRS 13 asks what market participants would exchange in an orderly transaction at the measurement date. The prudential architecture of the EU Solvency II framework applies an explicit asset-and-liability economic valuation principle, while the NAIC's statutory accounting framework serves the distinct task of monitoring insurers' solvency under the applicable state regime. These systems are coherent within their respective purposes; their outputs require interpretation before they are compared.

The subject has become more consequential as insurers' invested assets have diversified beyond exchange-traded sovereign and corporate bonds. The NAIC's analysis of year-end 2025 holdings reports approximately $1.21 trillion in broadly defined U.S. insurer private credit, around 13% of cash and invested assets, while stressing that narrower categories have differing structures and risks. The underlying regulatory breakdown reports about $666 billion of private-placement bonds, bank loans and middle-market collateralized loan obligations, and another $544 billion of privately rated bonds, business development companies, and private credit funds. Those figures describe one jurisdiction, one reporting date, and the NAIC's own broad classification; they are not measurements of all insurers worldwide. The IAIS 2025 private-credit assessment likewise emphasizes variation in definitions, data coverage, and liquidity characteristics. Valuation research that collapses this heterogeneous asset universe into a single risk label becomes less informative.

This study addresses six questions: what economic object is being valued; how measurement bases allocate market and credit changes among accounting statements; how expected loss connects default probabilities to contractual exposures; what evidence supports prices for assets without active trading; how asset shocks transmit through an insurer's total balance sheet; and how a decision-maker should reconcile these views. It uses a comparative reading of original accounting and supervisory texts, documented scholarly finance literature, algebraic examples, and critical examination of scope and timing. Source documents are linked at the precise claims they support, and the research ledger identifies what was verified and the limitations of access.

The cutoff for this analysis is November 15, 2026. European Solvency II revisions adopted in the 2025 reform program are relevant as future requirements, with the new framework applying January 30, 2027. The analysis of rules currently operating in November 2026 therefore uses the existing Solvency II requirements. IFRS recognition, EU supervisory valuation, U.S. generally accepted accounting principles, and U.S. statutory accounting remain separately identified; no assertion assumes that one jurisdiction's measurement rule automatically governs another's reporting entity. Where an issuer's full licensed standard is unavailable, official public standard summaries and directly available supervisory texts are used within their observable scope.

The argument proceeds from instrument classification to valuation, impairment, illiquidity, scenarios, and reconciliation. Original analysis consists of comparing economically different measures across jurisdictions, constructing matched numerical illustrations, clarifying sources of apparently contradictory asset movements, and identifying the evidence necessary for a reproducible insurer-specific assessment. It does not propose a new theory or assert novel accounting rules.

Financial Asset Measurement Bases

The economic role of invested assets

Invested assets supply cash flows across multiple states of the world. For life insurers, fixed-income holdings may be selected to support long-duration guarantees, annuity payments, and policyholder options; for property-and-casualty insurers, investment duration and liquidity demands interact with claims development and catastrophe exposure. An investment's value affects the insurer's ability to discharge obligations only through the conjunction of cash-flow recoverability, permissible use, transferability, and the amount and timing of liabilities. A high current market price does not, by itself, provide immediate unencumbered cash; a stable carrying value does not, by itself, establish protection against a market shock.

Four analytically distinct asset questions recur. Legal ownership and contractual rights determine whose cash flows belong to the insurer. Recognition asks whether those rights are assets within a particular accounting perimeter. Measurement asks which valuation attribute applies to the recognized item. Usability asks whether and when the item can meet obligations, including collateral, local-law restrictions, and liquidation constraints. Each question can change the numerator of a policyholder-protection assessment without altering the investment's contractual face amount.

An insurer may hold government bonds, public corporate debt, residential or commercial mortgages, private loans, funds, structured securities, equities, derivatives, investment property, reinsurance recoverables, and cash. The meaning of fair value varies with instrument structure and with what market participants can observe. Credit spread analysis is central to corporate loans and defaultable bonds; duration and convexity dominate rate-sensitive risk-free claims; volatility and optionality matter in callable securities; loan covenants and collateral recoveries matter for private credit; property-specific cash flows and capitalization assumptions affect investment real estate. Analytical evidence must correspond to the economic instrument.

Classification by rights, business model, and measurement objective

The IFRS 9 classification framework evaluates whether qualifying debt assets are managed to collect contractual cash flows, to both collect and sell, or under another model, and whether the cash flows satisfy the solely-payments-of-principal-and-interest requirement (SPPI). A qualifying hold-to-collect debt asset can be measured at amortized cost. A qualifying collection-and-sale debt asset normally enters fair value through other comprehensive income (FVOCI). Other financial assets enter fair value through profit or loss (FVTPL), subject to the standard's specific elections, designations, and instrument rules. Equity instruments are not assimilated to conventional debt classification; particular irrevocable FVOCI elections for qualifying equity instruments have different recycling consequences.

The instrument's commercial label does not determine its accounting category. A note with contingent returns linked to equity, leveraged exposure, or non-basic lending features may fail SPPI even if management intends to hold it to maturity. Conversely, a plain-vanilla bond held for active trading does not become an amortized-cost investment merely because contractual principal is repayable at maturity. Classification requires inspection of both contractual terms and actual portfolio management. Reclassifications of financial assets follow a change in business model rather than opportunistic reactions to unrealized losses, as specified in the same financial-instrument measurement standard.

The business-model assessment is a discipline of observable behavior: investment mandates, management performance metrics, histories of sales, funding needs, and the frequency and reasons for disposals. Long-duration insurance investment does not automatically imply a hold-to-collect model. An insurer may systematically sell bonds to rebalance cash flows under updated liability estimates, or it may hold individual instruments primarily to earn contractual interest. In both circumstances the cash-flow structure may be identical, while the accounting basis changes because the management objective differs.

An economic research dossier must keep a record of instrument-level SPPI assessments and portfolio-level management objectives. Without this record, estimates of what proportion of a portfolio is insulated from current market movements in recognized earnings become circular: the accounting basis would be inferred from the observed smoothness of its own outputs.

The valuation perimeter and instrument taxonomy

Accounting scope extends beyond investments usually labeled bonds or equities. Recoverables and receivables can introduce counterparty exposure. Derivatives frequently carry embedded optionality and collateral settlement requirements. Investment property is governed by a distinct standard, and cash-generating nonfinancial assets have their own impairment regime. Distinguishing the instruments is therefore a precondition for identifying the relevant measurement and loss-recognition rules.

The classification of financial assets under IFRS 9 should be separated from impairment under IAS 36. The IAS 36 scope explanation explicitly excludes financial assets subject to IFRS 9, as well as investment property measured using the IAS 40 fair-value model and various other assets covered by specialized rules. It would be a technical error to calculate a standard IAS 36 recoverable amount for an ordinary IFRS 9 loan portfolio as though the specialized financial-asset expected-loss model did not apply.

The IAS 40 property measurement framework permits a cost-model or fair-value-model policy, with associated disclosure and measurement consequences. A property held to earn rentals or for capital appreciation thus has a different source of valuation evidence from a fixed-income claim: rents, occupancy, market yields, comparable transactions, and operating costs matter. An insurer may have economic exposure to the same property through direct ownership and through a mortgage or collateralized security, yet the legal claim and valuation technique are different.

The research taxonomy used here identifies for each position: instrument type; unit of account; contractual cash-flow map; currency; interest-rate profile; seniority and collateral; optionality; accounting classification; valuation method; observable input quality; impairment framework; encumbrance; and linked insurance liabilities. The list is an organization of existing measurement questions rather than a separate theoretical classification system. It permits a consistent comparison without treating a portfolio average as though it were an instrument-level assertion.

Comparative international valuation regimes

Financial reporting, prudential solvency and managerial economics need distinct descriptions. The IFRS financial reporting basis seeks useful information for financial-statement users under its standards; it recognizes certain assets at amortized cost and others at fair value, and it separates certain gains and losses among profit or loss and other comprehensive income. Prudential valuation under the EU Solvency II system generally emphasizes market-consistent assets and liabilities: Article 75 of the Directive defines arm's-length exchange or transfer concepts for both sides of the insurance balance sheet.

The EU delegated regulation's Article 10 valuation hierarchy privileges quoted prices for identical assets in active markets, then comparable assets with adjustments, and then alternative valuation techniques that maximize relevant market information. Unobservable inputs can be used when relevant observable inputs are not available; they must reflect the assumptions market participants would employ. This supervisory instruction is stronger than a generic call to mark all assets using whatever internal forecast is available. The source and hierarchy of inputs are substantive constraints.

In the United States, most authorized insurers prepare statutory financial statements using a framework summarized in the NAIC accounting principles overview. Statutory accounting concentrates on a conservatively stated balance sheet and policyholder protection. Its admissibility, recognition and valuation rules differ from U.S. GAAP and can vary through state-prescribed and regulator-permitted practices. It follows that a U.S. statutory carrying amount cannot be assumed identical to an IFRS amortized-cost or Solvency II economic amount. The proper statutory measurement of an individual bond or other asset must be established from the applicable SSAP, designation, insurer type, state practice and reporting period.

The IAIS insurance supervisory standards place valuation within prudential risk assessment and the integrity of insurance supervision. The international institution helps articulate supervisory objectives; it does not create a single globally effective financial-statement accounting classification. The comparison therefore records objective, population, perimeter, accounting location, and reporting date before analyzing differences in numbers. This solves a persistent interpretive problem: two legitimate balance sheets may carry the same economic claim at different values, and even identical asset values may support different prudential conclusions when liability and eligibility rules differ.

Amortized Cost and Fair Value

The effective interest method

Amortized cost is a contractual cash-flow accounting measure shaped by initial recognition, repayments, interest accrual under the effective-interest method, premiums or discounts, and the applicable expected-credit-loss allowance. It is useful for describing the realization of a qualifying lending asset's yield through time. The IFRS 9 framework links this measurement to the business-model and SPPI conditions. It does not assert that a lending asset's current economic sale price equals its amortized cost.

For an uncomplicated zero-coupon asset with promised face amount F, initial contractual yield y, and remaining contractual maturity T, the basic illustration is P = F/(1+y)^T. If F is 100 monetary units, y is 4% per year, and T is five years, initial contractual present value is 100/(1.04)^5 = 82.1927. One year later, absent a payment, modification or credit loss, accretion at 4% produces 85.4804. This is a mechanical illustration of carrying-value growth in a qualifying zero-coupon investment. Taxes, transaction costs and credit allowances are omitted from the gross amount to make the mechanism transparent.

Suppose market-required yield then rises to 6% with four contractual years remaining and the promised 100 unchanged. The corresponding illustrative market value is 100/(1.06)^4 = 79.2094. The difference between market value and gross amortized cost is approximately -6.2710. A shift in an appropriate market discount rate can therefore generate a meaningful sale-price loss without implying that the borrower will necessarily default. This comparison is between unlike measurement bases; the gross carrying figure remains subject to IFRS 9 impairment where applicable.

The example also shows why categorization matters to managerial signals. A long-duration insurer holding a bond for collection may recognize interest income over time while a separate market-value disclosure and a prudential balance sheet reveal rate sensitivity. The presence of two values calls for a reconciled explanation, not a conclusion that one number is universally more accurate. Both can be computed consistently against their own question, but their comparability is conditional on the cash flows and assumptions used.

Market-participant pricing and credit spreads

IFRS 13 defines fair value in terms of the price received to sell an asset, or paid to transfer a liability, in an orderly transaction between market participants at the measurement date. The standard's approach asks for current market-participant assumptions, including risk, rather than an insurer-specific intention to hold a position to maturity. Fair value can be produced by an observed transaction price or by a valuation technique where a suitable direct market quotation is absent.

For a simple fixed-cash-flow debt claim, one may write an economic value as V = sum over t of E[C_t | information]/(1+r_t+s_t)^t, with adjustments as required for the consistent definition of expected cash flows, discount rates, and timing. The notation is a teaching simplification: risky cash flows and discount spreads cannot be assembled arbitrarily without double-counting credit risk. An expected-cash-flow approach that already reduces promised payments for default must not indiscriminately add the full credit risk premium appropriate to promised cash flows. Risk-neutral pricing, real-world expected cash flows, and required-return discounting are different mathematical parameterizations. The valuation technique must specify which one is being used.

Classic financial research explains the economic structure of defaultable debt. Merton's corporate-debt analysis treats risky debt claims through the firm's underlying asset value and limited-liability payoff; it establishes why credit deterioration affects bond prices before a contractual default occurs. Duffie and Singleton's defaultable-term-structure work develops an approach to pricing term structures under default risk. Both are research tools and economic explanations rather than substitutes for an issuer-specific accounting measurement policy or independently verified probability forecasts.

A corporate bond's yield spread is influenced by credit risk and additional market effects. The empirical work of Longstaff, Mithal, and Neis (2005) examines default and liquidity components of corporate spreads using credit-default-swap information. In an illiquid market, a rise in observed spreads can include a fall in expected recovery, an increased probability of distress, a rise in compensation for bearing risk, and diminished willingness or ability of intermediaries to hold inventory. A research analyst should investigate those mechanisms instead of calling every price decline an impairment of contractual cash-flow recoverability.

Measurement pathways and financial statement effects

Under IFRS 9, a qualifying amortized-cost debt asset ordinarily follows the effective-interest method, subject to an expected-credit-loss allowance, rather than recognizing every market interest-rate movement through the balance-sheet carrying amount. A qualifying debt asset measured at FVOCI remains at fair value on the statement of financial position, with the standard's relevant impairment and interest effects recognized in profit or loss and other fair-value movements generally presented in OCI. Debt instruments at FVTPL route fair-value changes through profit or loss; a separate ECL allowance is not simply superimposed on the same asset as though it were held at amortized cost. Accounting entries have to follow the actual classification and the detailed recognition rules.

For the earlier hypothetical bond, if it were a qualifying amortized-cost asset, the gross amortized-cost balance and impairment allowance would be assessed separately, and its disclosed current fair value might fall to 79.2094. If it were measured at FVOCI, the asset's recognized measurement would reflect the current fair value, with the required allocation of interest, credit impairment, and residual remeasurement through the accounting statements. At FVTPL, market movements enter recognized earnings under the applicable rules. These pathways imply different reported earnings volatility even while the same bond has one market exchange opportunity at a given date.

This pattern has two important insurance implications. First, a mismatch between asset accounting category and insurance-liability financial presentation may cause accounting volatility that differs from the economic volatility of the matched assets and liabilities. Second, an accounting category with more stable carrying values can obscure changes in the portfolio's exit value unless current values are disclosed and monitored through a distinct economic risk report. The IFRS 7 disclosure requirements provide a financial-instrument reporting context for evaluating exposures and measurement information.

Credit events, changes in forecast receipts, rate shocks, and portfolio sales can generate interacting effects. The analyst should separate the original effective-interest trajectory, current expected credit losses, contractual modifications, current fair values, transfers among presentation categories permitted under the standards, and realized sale outcomes. A one-line category labeled investment losses may conceal five different economic mechanisms.

Investment property, nonfinancial assets, and other impairment boundaries

Insurer asset valuation is not confined to bonds. Under IAS 40 investment-property guidance, selection and application of a permissible measurement model affects the way changing market rental yields, vacancies and comparable property prices appear in financial reporting. A property with apparently predictable annual rent can suffer an economic valuation decline as required capitalization yields rise. A cost-model carrying amount and a fair-value-model carrying amount therefore convey different information, subject to the applicable standard and disclosures.

The nonfinancial impairment boundary matters for assets such as operational property and certain intangible assets. The IAS 36 impairment model uses recoverable amount based on the higher of value in use and fair value less costs of disposal for assets within its scope. It includes cash-generating-unit tests for certain assets where independent cash inflows cannot be identified. Its rules cannot be transferred mechanically to financial assets within IFRS 9 or investment property measured at fair value. The economic phenomena of deteriorating value may be similar, yet the recognition test, unit of account, and measurement objective differ.

Fair value is also conceptually distinct from proceeds in a distressed forced sale. An orderly market-participant exit price at the measurement date is the benchmark of the fair-value definition. A stress exercise may separately ask what proceeds could be raised under urgent liquidation. The two results can differ because the second incorporates the insurer's specific time constraint, volume and funding need. Keeping them separately reported supports better decision-making than silently substituting a fire-sale price for fair value or assuming an orderly-market price would always be realizable under stress.

Duration, convexity, and insurance liabilities

For a fixed-income portfolio, first-order interest-rate sensitivity can be approximated by the negative of modified duration multiplied by a small parallel yield change: ΔV/V ≈ -D_mod Δy. A second-order approximation adds 0.5 C(Δy)^2, where C describes convexity under the specified price-yield definition. These are local mathematical approximations, not guarantees for large or nonparallel movements. Callable bonds, mortgage prepayments, floating-rate coupons, and embedded options can alter effective duration dynamically.

Asset duration by itself does not measure an insurer's solvency interest-rate risk. Let A and L denote economic present values of assets and insurance liabilities and E = A - L a simplified surplus. For small shocks, ΔE ≈ ΔA - ΔL. A liability with longer effective duration than its backing assets can experience a larger value increase when risk-free rates fall, putting pressure on surplus even if bond prices rise. Conversely, higher rates can reduce the economic values of both assets and liabilities, with the net movement depending on duration, convexity, option behavior, margins, and how regulatory technical provisions are constructed.

The liability comparison cannot be treated as an arbitrary offset. IFRS 17's insurance contract standard measures insurance obligations within a contract-specific accounting framework, including fulfillment cash flows, risk adjustment and, where applicable, a contractual service margin. The EIOPA implementation study documents that IFRS 17 and Solvency II serve different objectives and produce both synergies and significant liability-measurement differences. Thus an interest-rate scenario must state which liability convention it is shocking before attributing the net result to the insurer's economic resilience.

Credit Impairment and Expected Loss

Credit deterioration, default risk, and impairment

Credit impairment concerns a reduction in the expected recoverability of promised financial-asset cash flows under the relevant accounting convention. Default is one possible event within a wider set of deteriorating credit states. A borrower can continue current payments while becoming materially more vulnerable to a future decline in revenue, refinancing disruption, collateral depreciation, or covenant acceleration. Credit assessment must therefore connect contractual structure to changing repayment capacity rather than wait for an observable missed payment.

Market prices frequently adjust before accounting indicators become decisive. The structural perspective developed in Merton (1974) establishes a way of seeing default risk as an economically valued contingent claim. Its lesson for insurance research is that the current market price of a bond reflects more than the next scheduled coupon. Asset value, leverage, maturity and uncertainty influence the value of the debt promise. The model's simplifying assumptions must be disclosed when it is used for practical valuation; an insurer's investment in a private loan cannot be credibly priced from the structural model alone without sufficiently reliable borrower inputs.

Credit-risk measurement has at least three separate outputs. An accounting impairment allowance reflects expected losses under the applicable impairment standard. A market price reflects the compensation and discount demanded by market participants, including premia beyond probability-weighted credit loss. An economic capital or solvency assessment considers the distribution of losses and the effect of adverse states on the insurer's whole balance sheet. Each output is relevant, but their numerical equality is neither required nor generally expected.

Credit quality is multidimensional. Relevant information includes borrower cash-flow projections, leverage, collateral coverage, refinancing need, sector concentration, covenant protection, guarantees, debt seniority, historical restructurings, payment-in-kind features, currency mismatch, geographic conditions and macroeconomic changes. The importance of each variable depends on the claim. A senior secured floating-rate loan can exhibit limited pure interest-duration sensitivity while remaining exposed to borrower deterioration and refinancing stress. A long-duration investment-grade bond can exhibit substantial market-price sensitivity to risk-free rates while expected default losses remain modest.

IFRS 9 expected credit loss mechanics

The IFRS 9 impairment requirements introduce expected credit loss recognition for financial assets within the standard's impairment scope, including relevant amortized-cost and FVOCI debt instruments. For the general approach, the recognized allowance ordinarily reflects 12-month expected credit losses before a significant increase in credit risk and lifetime expected losses when credit risk has increased significantly since initial recognition. Financial assets that become credit-impaired have additional interest-revenue and reporting consequences. The general approach coexists with specific simplified and purchased-or-originated credit-impaired treatment where applicable.

The phrase 12-month expected credit loss requires precision: it concerns the present value of lifetime cash shortfalls arising from defaults that may occur within the next 12 months. It does not restrict the cash shortfall itself to payments due during that calendar year. This interpretation follows from the expected-credit-loss framework in IFRS 9 and is essential for long-term insurance investments. A default occurring during the first year can impair cash receipts in several subsequent years.

Expected credit loss is a probability-weighted estimate taking account of reasonable and supportable information on past events, current conditions and forecasts of future economic conditions. In the simple single-exposure approximation, ECL = PD × LGD × EAD × DF, where PD is a consistently defined probability of default over a specified horizon, LGD is loss given default, EAD is exposure at default and DF is an appropriate discount factor. This is an explanatory approximation, not a complete replacement for the standard's probability-weighted present-value requirements. Multi-period exposures require survival-adjusted marginal probabilities, exposure evolution, recovery timing and, where relevant, contractual features and scenarios.

Credit risk at origination remains an essential benchmark for significant-increase-in-credit-risk assessments. A portfolio can consist of individually low absolute default probabilities, while several assets have suffered a pronounced relative deterioration since purchase. Mechanical reliance on a single numerical rating grade or delinquency threshold can therefore fail to capture meaningful changes. Conversely, market spread widening driven by liquidity or risk-premium shifts need not demonstrate the same change in accounting expected losses. A disciplined staging decision triangulates internal credit ratings, borrower evidence, forward-looking scenarios, observable prices, and the reason for spread moves.

The IASB's IFRS Interpretations Committee materials show why issues of recognition and application arise from specific contractual and measurement provisions rather than from a generic expectation of credit loss. Published authoritative summaries are useful starting evidence; the exact IFRS paragraph and applicable facts must govern an actual financial statement. In this study no bank-only prudential interpretation is represented as automatically binding on insurers.

Probability-weighted loss calculations

Consider an illustrative amortized-cost bond with exposure at default of 100 monetary units, a one-year default probability of 1.5%, loss given default of 40%, and an effective discount rate of 4%. Under an intentionally simplified default-and-recovery assumption with one-year payment timing, ECL = 100 × 0.015 × 0.40 / 1.04 = 0.576923. If credit conditions deteriorate such that the comparable one-year probability becomes 5% and loss given default 50%, the same calculation yields 100 × 0.05 × 0.50 / 1.04 = 2.403846. The difference, 1.826923, measures the change in this illustrative one-year loss estimate. Actual expected credit losses depend on all relevant timing and scenarios rather than merely replacing two inputs in a static formula.

For a multi-period teaching example, suppose the contract is exposed to default at the end of each of the next three years, with marginal unconditional default probabilities p1 = 1%, p2 = 1.8%, and p3 = 2.2%; exposure at default remains 100 and loss given default is 40%. At a 4% annual discount rate, simplified lifetime ECL is 40 × [0.01/1.04 + 0.018/(1.04)^2 + 0.022/(1.04)^3], approximately 1.833. Marginal probabilities here are unconditional shares of the original exposure population and already reflect survival. If conditional year-by-year default hazards were supplied instead, survival probabilities would have to be applied before summation. The example intentionally excludes coupons, amortization, recovery delays and multiple macroeconomic paths.

A second scenario shows how macroeconomic mixtures alter the estimate. Suppose a two-state one-year forecast assigns 70% probability to a baseline state with PD 1.5% and LGD 40%, and 30% probability to a recession state with PD 6% and LGD 55%, each with EAD 100 and the same discount factor 1/1.04. The mixture estimate is [0.70 × 100 × 0.015 × 0.40 + 0.30 × 100 × 0.06 × 0.55]/1.04 = 1.355769. The difference from the baseline-only 0.576923 is 0.778846. This is a demonstration of scenario-weighting mechanics, not a credible forecast of an insurer's portfolio. The scenario probabilities and severity assumptions are chosen arbitrarily and disclosed as such.

The example also exposes a common analytical error: using an unconditional probability of default for a future year and then multiplying it again by an independently estimated survival probability. That would double-adjust for survival. A model must state whether its PD series contains conditional period hazards, unconditional marginal defaults, or cumulative default probabilities. A separate error occurs if a lifetime credit-loss figure is added to an otherwise complete market credit spread deduction without defining the compatible cash-flow and discounting convention. Both errors can create materially distorted valuations.

The uncertainty of ECL is broader than statistical parameter uncertainty. Borrower-specific collateral may be valued using assumptions that are themselves sensitive to the same macroeconomic scenarios driving default. Defaults can become correlated in recessions. Recovery can be delayed during restructuring. Probability estimates derived from liquid public credit histories may be poorly calibrated for private, covenant-lite or structured instruments. Models must therefore be reconciled with file-level borrower evidence and tested for coherent dependence among PD, LGD, exposure and the macroeconomic state.

U.S. GAAP credit losses and statutory differences

The U.S. GAAP current expected credit loss approach under FASB Topic 326 generally evaluates expected credit losses on relevant financial assets measured at amortized cost over their contractual lives using appropriate expected-loss information; there are separate rules for available-for-sale debt securities. The Federal Reserve's CECL explanatory guidance cites FASB Accounting Standards Update 2016-13 and explains the assets within its general scope, including certain receivables and reinsurance recoverables. As a bank-regulator FAQ, that document explains the financial-asset accounting framework and not the complete insurer-specific implementation of every U.S. GAAP rule.

An important comparison concerns initial recognition and credit-risk change. Under IFRS 9's general approach, many performing assets begin with a 12-month ECL allowance and move to lifetime ECL after a significant increase in credit risk. Under CECL for relevant U.S. GAAP amortized-cost financial assets, the general allowance estimates lifetime expected loss from the start. It is therefore inappropriate to import IFRS 9 staging labels directly into a CECL financial statement or to call both allowances equivalent merely because both employ forecasts. Each accounting system has its own scope, elections, threshold provisions and presentation.

United States statutory insurance accounting adds a further layer. The NAIC's explanation of statutory rules emphasizes solvency, admissibility, conservatism, and the effects of prescribed and permitted practices. Those objectives can require measurement and recognition outcomes different from U.S. GAAP, including the treatment of invested assets and impairment. The exact applicable SSAP and insurance entity are needed before stating the recognition trigger for a particular security. Treating a CECL allowance as an automatic statutory-accounting entry for all insurers would disregard the independent regulatory accounting framework.

The comparative finding is that credit-loss conventions cannot be ranked simply by which one produces the larger allowance in an illustrative period. Their timing, scope, classification, discounting and presentation differ. For a sound comparison, the same underlying borrower cash-flow scenarios should be applied to each system and then translated by the system's recognition rules. Differences should be decomposed into accounting perimeter, horizon, probability methodology, discounting, and presentation. This reveals what a reader can infer from the reported charge and where the economic risk must be examined separately.

Model sensitivity, staging, and verification

Goodwill and investment-property impairment tests provide distinct measurements from financial-asset ECL; the IAS 36 standard overview makes this boundary explicit. For an insurer with both a loan portfolio and a directly held office building, separate model inventories and triggers are appropriate. A single enterprise-wide ratio labeled impairment divided by total assets risks conflating incomparable measurements.

For financial assets, auditability begins with source data: contract files, cash-flow schedules, purchase yields, credit grades, borrower financials, collateral values, modification records, guarantees and payment histories. Forward-looking overlays require documented judgments, scenario weights and evidence of how macroeconomic information changes actual debt recoverability. A valuation committee should verify whether the reported allowance is consistent with independently evaluated borrower files, model validation results, observed restructurings, later recoveries and the portfolio's risk concentrations.

A useful sensitivity exercise keeps one factor fixed at a time before testing joint effects. Holding PD fixed while raising LGD can identify collateral sensitivity; holding LGD fixed while changing PD isolates borrower-default assumptions; changing discount rates shows the present-value mechanism. But one-factor tests must be supplemented with coherent joint scenarios because PD, recovery and collateral values are usually correlated in severe macroeconomic states. Independence assumptions should be justified by the asset economics and evidence, not adopted for computational convenience.

Model validation should distinguish discrimination, calibration and economic relevance. A model might rank loans correctly from strongest to weakest while systematically underestimating defaults. It might predict defaults well but produce recovery estimates with inappropriate time horizons. It might perform adequately for public investment-grade bonds yet misrepresent the behavior of private credit. Backtesting can be limited by the scarcity of defaults in certain portfolios, so limitations of experience data must be transparent. The analytical conclusion is that a single ECL number is a conditional estimate embedded within a hierarchy of evidence and judgments.

Illiquid Asset Valuation

The investment case for private assets

Private loans, private placements and other illiquid instruments can help insurers pursue contractual yield, duration matching, diversification and exposure to risks less accessible through public markets. These potential benefits are conditional on underwriting discipline, contractual protection, servicing, and the ability to finance long-term holdings against the insurer's liability profile. The same assets may have sparse secondary trading, limited public financial information, heterogeneous covenants and material reliance on external administrators or ratings. An insurer's ability to hold an asset through time does not eliminate the need to measure its contemporary economic value.

Official supervisory evidence indicates why this sector deserves careful examination. The NAIC's year-end 2025 analysis estimates $1.21 trillion of broadly defined U.S. insurer private credit, while specifying the narrower segments and the methodological uncertainty of a category without a universal definition. Approximately $544 billion of the total lies in privately rated bonds, business development companies and private credit funds emphasized in the NAIC's discussion of structural complexity. These are year-end 2025 U.S. reporting estimates, not universal percentages of private credit risk. A portfolio's vulnerability depends on the underlying claims, cash-flow seniority, leverage and liquidity, not on the private-credit label alone.

The IAIS Global Insurance Market Report excerpt highlights inconsistent international definitions of insurer private credit, limitations in supervisory exposure data and characteristics such as bespoke structuring, illiquidity and diminished public transparency. It notes that some jurisdictions' estimates were below five percent of total insurance sector assets, a figure that varies with definition and jurisdiction. The apparent discrepancy with the NAIC's 13% broad U.S. estimate is therefore not itself evidence of conflicting data: the scope, population and taxonomy differ. A comparative study must put these denominators and inclusion criteria side by side.

Valuation techniques and evidence quality

The valuation starting point under IFRS 13 and Solvency II is an exchange-oriented measure reflecting market-participant assumptions, with technique selection adjusted for the quality of the available market data. The Solvency II Article 10 hierarchy explicitly calls for identical active-market prices when available, then comparable quoted prices with adjustments, then alternative approaches using the greatest feasible share of observable information. Where assumptions become unobservable, the user must establish why those assumptions would be adopted by market participants.

The market approach uses observable transactions in the same or similar exposures, adjusted for credit quality, seniority, contractual rights, collateral and liquidity. The income approach discounts estimated cash flows using assumptions consistent with their risk and timing. A current-replacement-cost technique may be suitable for certain nonfinancial assets. These approaches are not arbitrary competing answers; they correspond to different characteristics of the asset and different degrees of observation. Reconciliation across two appropriately designed techniques can serve as evidence of sensitivity rather than as proof that their numerical average is the true price.

For a private loan, a cash-flow valuation might begin with contractual principal and coupon schedules, then evaluate plausible early refinancing, defaults, restructurings and recoveries. The discount rate should be consistent with whether forecast cash flows are promised, risk-adjusted or risk-neutral. An adjustment for transfer restrictions or low market liquidity must represent market-participant pricing of the instrument. A shareholder's desire to avoid recognizing a loss is not a market-participant input. An illiquid-asset model that assumes the original origination yield indefinitely will tend to miss changes in prevailing credit spreads and liquidity preferences.

When comparable trades are rare, the evidence trail is important. A transaction from six months ago may be relevant but require explicit adjustment for subsequent changes in rates, borrower results, covenants and financing markets. A broker indication can be informative but may represent a nonbinding quote rather than an executable bid. A sponsor's valuation may be a useful input yet requires independent review when the sponsor has economic incentives connected with fees, borrowing-base tests or capital raising. Price-services data need scrutiny for stale observations, smoothing and common-model dependencies. The integrity of the valuation is partly a question about the provenance and independence of its inputs.

Private credit, structured claims, and fair-value controls

Structured securities distribute losses and cash flows among contractual tranches. Their valuation requires knowledge of collateral performance, allocation priorities, triggers, servicer practices, correlations and embedded optionality. The research by Coval, Jurek, and Stafford (2009) explains how securitization can create securities whose ratings or perceived safety are sensitive to assumptions about systematic risks and underlying collateral. That historical analysis supports caution about tail dependence; it does not prove that all present private credit or collateralized loan obligations are incorrectly priced.

A generic discounted-cash-flow model treating a senior structured tranche as an ordinary homogeneous loan can miss attachment-point dynamics. Small changes in collateral default correlation may alter the timing and scale of losses reaching senior tranches. Reinvestment rules, overcollateralization tests and payment waterfalls can redirect cash flows between debt classes. Scenario valuation should therefore follow the legal structure of the claim. The value of a fund interest also depends on governance rights, fees, restrictions on redemptions, and the valuation of underlying holdings rather than simply its reported net asset value.

Independent pricing should investigate whether two apparent market quotations derive from the same data vendor or model. Multiple identical outputs do not constitute independent corroboration when they share inputs. For a privately rated bond, the rating must be checked against available borrower and structural evidence; external classification can inform but cannot substitute for an actual valuation model. The NAIC's private-credit commentary specifically identifies complexity and underlying alternative-asset exposure as reasons for additional scrutiny in privately rated and fund-related segments.

A well-supported valuation file identifies the security, reporting date, cash-flow convention, source of underlying data, selected method, competing methods considered, scenario assumptions, model ownership, control checks, and approval history. It records whether a comparable price was independently transacted, merely indicated, or inferred. It also distinguishes a credit-driven price change from a liquidity-driven spread movement. Such a file is essential for insurer economic-capital reporting because otherwise the apparent precision of a point estimate can exceed the information used to construct it.

Liquidity, collateral, encumbrance, and stale marks

Market liquidity describes the capacity to trade an asset at a relevant price and size within a relevant time. Funding liquidity describes the ability of an institution or market participant to obtain cash or financing. Brunnermeier and Pedersen's analysis explains the feedback between these mechanisms and the possibility of reinforcing liquidity pressures. For an insurer, the connection is especially important when catastrophe claims, margin calls, derivative collateral demands or policyholder options accelerate cash requirements at the same time as investment markets weaken.

The analysis should separate at least three prices: a reasoned orderly market-participant fair value; a modelled stressed fair value under an adverse but orderly market state; and cash proceeds obtainable from rapid sales under firm-specific time and volume constraints. They need not coincide. A liquidity stress must also include which assets are pledged or otherwise encumbered and the sequencing of payments. High-quality collateral that secures another transaction may have less immediate availability for paying claims than an unencumbered asset with a similar mark.

The absence of a new price does not establish the absence of a new loss. Private assets can display infrequent valuation changes because the input process updates slowly. An insurer relying on quarterly sponsor marks while calculating daily economic exposures should identify the interpolation and adjustment method, including the timing lag between borrower information and portfolio recognition. A stable series produced by stale prices tends to understate short-horizon volatility and cross-asset correlation. Before inferring diversification from measured covariances, the analyst should examine the frequency and independence of actual observations.

At the same time, equating illiquidity with an automatic impairment charge would confuse distinct effects. An asset can be difficult to sell while contractual collections remain robust; market participants may nonetheless demand a liquidity discount. An asset can be frequently traded while its borrower becomes impaired. The correct accounting effect depends on the governing measurement classification, loss-recognition standard, and economic facts. Liquidity is an explanatory dimension of price and funding capacity, not a universal substitute for a credit-loss trigger.

Supervisory disclosure and governance

The EU delegated regulation's Article 263 controls require documentation of alternative valuation methods, their assumptions and valuation uncertainty, and regular comparison with experience. This is a direct supervisory basis for evaluating the credibility of illiquid-market marks. It makes model uncertainty an explicit governance subject rather than a residual number left implicit in a valuation spreadsheet.

At the international level, the IAIS ICP valuation material underscores the importance of appropriate valuation and supervisory transparency. In the U.S. context, the NAIC's statutory accounting explanation makes state-level prudential reporting and policyholder protection central. A comparison should therefore focus on the substantive governance tasks common across different regimes: clear responsibility for assumptions, documented data sources, reproducible valuation processes, independent challenge, meaningful sensitivity disclosures, and escalation when reliable evidence is unavailable.

The appropriate disclosures differ by reporting perimeter and instrument, but several analytical questions remain useful. What percentage of the portfolio rests on unobservable material inputs? How far is the model valuation from a transaction-based benchmark? How often are those benchmarks refreshed? What changes if borrower financials deteriorate or a loan's maturity must be extended? Are affiliated funds or related parties involved in price sourcing? What is the cash consequence if the portfolio must be liquidated to meet a claim shock? These questions convert valuation quality into observable research and control tasks.

Asset Value Stress

Stress architecture and economic mechanisms

A stress exercise makes the dependence of reported financial strength on valuation assumptions explicit. Its objective is to construct coherent adverse states across market prices, recoverable cash flows, asset liquidity and insurance liabilities. A stress scenario is neither a forecast nor a proof of the probability of its occurrence unless a justified probabilistic structure accompanies it. A useful test describes shocks, transmission channels, accounting treatment, and the economically relevant results.

Interest-rate increases lower the present value of many fixed-rate assets, but their effect on economic surplus depends on simultaneous liability movements. Credit-spread widening can lower market prices with varying contributions from future default loss, risk premium and liquidity. Default and recovery shocks affect expected cash flows, impairment allowances, statutory admissibility or credit charges. Equity and real-estate shocks affect assets with different accounting bases. Currency movements affect both foreign investments and relevant liabilities. A catastrophe loss may create a liquidity drain with relatively little direct bond-credit deterioration. Correlated scenarios are essential where financial market and insurance stress occur together.

The construction of a consistent stress begins with an inventory that separately identifies starting book value, current market value, contractual exposure, asset category, liquidability, and the applicable liability cash-flow profile. The analyst selects a valuation convention before applying each shock. A percentage market-price haircut is an expedient teaching approximation, not a full pricing method or regulatory capital factor. In operational use, instrument-level scenario curves, borrower credit paths, collateral valuation, and contractual cash-flow models would normally be required.

Reproducible investment portfolio scenario

Consider a hypothetical insurer with 1,000 monetary units of assets at an initial economic measurement date. For illustration only, allocate 80 to cash, 300 to sovereign bonds, 240 to corporate bonds, 230 to private credit, 80 to equities, and 70 to investment property. The numbers are chosen to demonstrate arithmetic, not to describe any real market, issuer, or representative portfolio. The hypothetical stress applies current-market value shocks of 0%, -6%, -12%, -18%, -25%, and -15%, respectively.

Asset category Initial economic value Assumed price shock Value reduction Stressed economic value
Cash 80.0 0% 0.0 80.0
Sovereign bonds 300.0 -6% 18.0 282.0
Corporate bonds 240.0 -12% 28.8 211.2
Private credit 230.0 -18% 41.4 188.6
Equities 80.0 -25% 20.0 60.0
Investment property 70.0 -15% 10.5 59.5
Total 1,000.0 — 118.7 881.3

The arithmetic is a direct weighted sum of initial values multiplied by the selected price changes: 18.0 + 28.8 + 41.4 + 20.0 + 10.5 = 118.7. The portfolio loses 11.87% in this particular stylized scenario. The contributions of sovereign bonds, corporate bonds, private credit, equities and property to the 118.7 reduction are approximately 15.2%, 24.3%, 34.9%, 16.8% and 8.8%. The private-credit allocation is therefore the largest contributor in these chosen assumptions, an outcome generated jointly by its weight and assumed shock rather than by an empirical claim that private credit is necessarily the riskiest category.

This portfolio is deliberately measured on one hypothetical economic basis. The table must not be directly substituted for an IFRS profit-and-loss statement: amortized-cost instruments, FVOCI debt and FVTPL investments would generate different recognized accounting pathways. Nor is it a calculation of a Solvency II solvency capital requirement or U.S. risk-based capital ratio. Those regulatory results require different specified methodologies, calibration and eligible-own-funds treatments. The stress demonstrates cross-asset economic value sensitivity as a precursor to those separate calculations.

Coupled asset-liability scenario

Assume initial insurance and other liabilities have a hypothetical economic value of 880, leaving initial economic surplus of 120. Under the adverse market scenario above, assets fall to 881.3. Suppose a simultaneous increase in discount rates reduces the present value of liability cash flows by 50, while adverse insurance cash-flow developments increase their value by 20. The stressed liability is 880 - 50 + 20 = 850, and the resulting simplified economic surplus is 881.3 - 850 = 31.3.

The arithmetic surplus decline is 88.7, equal to the 118.7 asset-value loss less the net 30 liability-value reduction. The example illustrates an asset-liability offset that materially changes the interpretation of asset losses. The choice of 50 and 20 is hypothetical. In a real life-insurance liability model, policyholder options, guarantees, cash-flow management, contract boundaries and the risk-adjustment treatment can change the direction and magnitude of the offset. For catastrophe-exposed nonlife insurance, an immediate claim payment creates a different cash-flow and liquidity structure from an interest-rate-only repricing.

A second analytical layer considers forced asset monetization. If the insurer had to sell all 230 initial units of private credit rapidly and an incremental forced-sale liquidity concession were hypothesized at 8% of that initial amount, the additional cash-proceeds loss would be 18.4. That 18.4 is a separate hypothetical transaction-cost and urgency scenario, not an additional IFRS 13 fair-value markdown to be applied automatically to the earlier stressed mark. The two scenarios refer to different sale conditions; the model must specify whether the 8% applies to the original position, the stressed mark, or the actually liquidated quantity before applying it operationally.

An insurer's urgent liquidity need is also a question of available assets and payment timing. Even if market-value surplus is positive, an entity could face a short-term cash mismatch if much of the portfolio is encumbered, illiquid or dependent on settlement periods longer than expected claims disbursement. Stress analysis should therefore report surplus and liquidity requirements separately. The market-funding liquidity interaction gives a theoretical reason to examine their feedback under market pressure.

Nonlinearities, concentrations, and model risk

Simple percentage shocks conceal important nonlinearities. A callable bond can shorten or lengthen its expected life as rates change. A mortgage security can respond to both prepayment incentives and borrower deterioration. Structured securities can suffer a sharp jump in senior-tranche expected loss once collateral losses cross an attachment point. A private loan may become impaired when a refinancing window closes even before a scheduled principal payment is missed. A derivative portfolio's collateral calls may accelerate funding pressure during the same market stress that raises measured hedging gains.

Concentration risk also extends beyond legal issuers. Several bonds may share exposure to the same commercial property market, commodity prices or infrastructure revenue. Private credit funds and directly held private loans may ultimately finance overlapping borrowers. Ratings issued by a small number of vendors may rely on similar models and collateral assumptions. A nominally diversified portfolio can contain correlated exposures invisible to asset-class percentages. Scenario analysis should therefore supplement accounting classes with issuer, sector, geography, underlying collateral, counterparty and structural dependencies.

Macro scenarios need timing coherence. A one-day interest-rate market shock and a three-year recession credit-loss distribution cannot simply be added without clarifying the measurement date and valuation methodology. An accounting ECL estimate may incorporate a range of future economic paths, while a one-day fair-value stress reprices current expected returns and premia. An insurer's expected claims cash flows may respond to inflation with lags. Careful temporal alignment allows the analyst to reconcile differences instead of inventing precise net losses from incompatible horizons.

The IAIS's 2025 private-credit review makes data-definition and transparency concerns especially relevant to cross-border stress comparisons. Two samples of insurers with the same stated private-credit share can differ greatly if one includes mortgage lending and private placements while another limits the definition to direct lending and privately structured exposures. A sample-level model should document its denominator and composition before claiming a sector-wide sensitivity.

Sensitivity interpretation

An effective stress report should display a baseline, each isolated risk shock, coherent combined scenarios, and the limitations of the valuation method. For each material asset, it should identify how much of the result comes from contractual cash-flow changes, discount-rate changes, market liquidity, model assumptions and forced-sale constraints. Where these are not separately identifiable from observed data, the uncertainty should be described instead of inventing a precise attribution.

The illustrative portfolio demonstrates an important accounting-to-economic distinction. A nominal 11.87% economic asset-value decline is not automatically an 11.87% profit-and-loss charge. Recognition depends on classification, impairment, current value and reporting regime. The same portfolio's 88.7 economic surplus loss depends on simultaneous liability assumptions and the starting liability value. Liquidity remains a separate requirement. An insurer's resilience is therefore more accurately studied through linked, explicit measurement schedules than by any isolated percentage of invested assets.

Accounting and Economic Reconciliation

The reconciliation question

A reconciliation explains why different legitimate measurement systems produce different asset, income, equity and solvency figures for one economic portfolio. Its object is to preserve the distinctions among accounting recognition, fair-value measurement, credit-loss estimation, prudential eligibility, liquidity, and future contractual cash receipts. A reconciliation is successful when the reader can trace each material difference to a clearly identified standard, factual input or valuation assumption.

Take a qualifying hold-to-collect bond subject to IFRS 9 expected-credit-loss recognition. Its IFRS carrying value may be based on gross amortized cost less loss allowance. Its IFRS fair-value disclosure may incorporate a current credit and liquidity premium. The insurer's Solvency II economic asset value may follow market-consistent valuation methods, while a U.S. statutory amount would be computed under a separate statutory framework if the insurer reports on that basis. These values are not additive and cannot be converted by a universal fixed discount. They are answers generated by different recognition and measurement rules.

The IFRS 9 source standard, IFRS 13 fair-value definition, and EU Article 75 principle supply separate reference points for the comparison. An actual entity may report under only some of these regimes. The researcher's international comparison does not establish that a U.S. statutory insurer is required to adopt IFRS 9 or that every EU insurer presents IFRS consolidated statements.

A systematic balance-sheet bridge

Begin with a position-level inventory and the legal definition of each investment. Establish the reporting entity and consolidation perimeter before taking accounting values from financial statements. Map the applicable classification and impairment standard. Then obtain the best supported current market-participant value with documented input quality, without silently replacing the recognized carrying measure. Calculate any differences attributable to ECL, interest-rate repricing, credit spreads, liquidity, optionality and nonfinancial valuation. Finally, aggregate the aligned assets with liabilities measured under each respective reporting convention.

The reporting bridge should preserve several fields for each position: IFRS gross carrying amount; recognized IFRS allowance where relevant; IFRS net carrying amount; IFRS fair value and hierarchy level; U.S. GAAP value where actually applicable; statutory carrying amount and admitted value where relevant; prudential economic value; cash-flow forecast; contractual maturity; market liquidity indicator; encumbrance; and valuation uncertainty. Empty fields should reflect genuinely inapplicable regimes rather than be filled with fabricated equivalences. If management claims to use a modelled market value that differs from IFRS fair value, the difference needs a stated economic or methodological justification.

One simple mathematical bridge is economic surplus = fair-value assets minus appropriately measured economic liabilities. A comparable accounting equity identity is reported assets minus reported liabilities. The difference between the two is the sum of differences across assets and liabilities plus perimeter and admissibility effects. Because changes on both sides may interact through asset-liability strategies, the result should be decomposed, at minimum, into market remeasurement, credit allowances, liability valuation, accounting location and other recognition adjustments. The decomposition is an application of familiar accounting identities and financial analysis rather than a new theoretical system.

A numerical bridge for the hypothetical bond at the end of year one begins with gross amortized cost of 85.4804. An illustrative credit allowance of 0.5769 would yield a net accounting value of 84.9035 if the instrument and assumptions fell within the relevant amortized-cost impairment rules. The scenario market value of 79.2094 differs by approximately 5.6941 from that net value. The difference should be investigated as market yield and other assumptions rather than misreported as a separate realized loss or as evidence that the allowance was mathematically incorrect. The toy ECL example used EAD 100, not the bond's actual end-year gross carrying amount, so the figures have intentionally illustrative cross-example assumptions and must not be presented as one fully internally calibrated security valuation. A live calculation would use one reconciled EAD, exposure horizon, effective interest convention and asset cash-flow schedule across the entire bridge.

That last qualification illustrates a research control principle. Numerical examples can be arithmetically sound individually while becoming inconsistent when linked through different assumed exposures. A scholarly presentation should identify such deliberate simplifications and refuse to make claims that demand a complete instrument-level model. The purpose of the example is to distinguish the economic interpretations of gross contractual accretion, credit allowance, and current market price.

Recognition, profit, other comprehensive income, and own funds

Reporting location changes the interpretation of financial performance. Under the applicable IFRS 9 category, changes in asset fair value can affect profit or loss or OCI; certain eligible debt investments carried at amortized cost follow interest and impairment accounting rather than full recurring remeasurement through earnings. IFRS 17 adds the insurance-liability accounting choices and requirements that govern how insurance finance income and expenses are presented. The EIOPA report on IFRS 17 implementation identifies different valuation objectives and important differences between IFRS 17 and Solvency II. A comparison of asset and liability effects must therefore identify the full accounting presentation instead of treating net income alone as the comprehensive statement of economic surplus.

Under the EU prudential regime, the valuation hierarchy for assets and liabilities follows the Article 10 supervisory methodology. Eligibility and classification of own funds, required capital and the risk profile add other prudential layers beyond the asset valuation itself. An analyst cannot obtain an insurer's solvency ratio by dividing financial-statement book equity by an arbitrary asset risk percentage. A solvency balance sheet, eligible own funds and an applicable capital calculation must be reconstructed under the relevant rules.

Under U.S. statutory insurance reporting, the NAIC SAP framework emphasizes admissible assets, conservative recognition and regulators' oversight. Its focus explains why an asset may receive a statutory treatment that differs from investor-oriented financial reporting. State-prescribed and regulator-permitted practices must be examined to establish the actual insurer-specific reporting basis. Accounting figures need explicit provenance to support comparisons between domestic and international insurers.

Governance, disclosure, and reproducibility

A reproducible valuation system records both numbers and their conditions of validity. It should state the reporting date, accounting regime, model version, data vintage, market source, approval owner, material unobservable assumptions, historical backtesting, scenario definition, limitation of use, and reconciliation of competing estimates. Changes in model methodology should be separated from changes in market information. A valuation control that cannot reproduce a prior material figure cannot demonstrate whether subsequent differences arise from true market developments or revised algorithms.

The Solvency II valuation-control requirements call for identification of assets subject to alternative methods, justification, documented assumptions, valuation uncertainty and comparison with experience. In international supervisory comparisons, the IAIS's valuation principles similarly frame appropriate measurement as part of reliable risk oversight. Financial reporting also depends on meaningful disclosure under IFRS 7. Together, these sources support a research-oriented governance conclusion: a price estimate should be assessed in relation to its source, method, uncertainty and balance-sheet purpose.

Uncertainty reporting should convey more than the outputs of a model sensitivity table. Unobservable inputs can exhibit identification uncertainty when several combinations produce similar prices, estimation uncertainty arising from limited data, and regime uncertainty when past information has weak relevance to prospective borrower behavior. The report should make clear which judgments dominate the mark and whether secondary methods genuinely cross-check the price. Market transactions, cash collections and subsequent impairments provide useful evidence for retrospective validation, but they do not magically reveal the unique correct fair value at an earlier date.

For external readers, effective disclosures connect investment concentration and asset classification to the insurer's claims-paying obligations. A long-duration bond portfolio with low current liquidity may be compatible with long-duration liabilities yet vulnerable to surrender or collateral demands. A portfolio with little accounting impairment may still have material risk-free-rate sensitivity. A portfolio with significant FVOCI remeasurement may also carry corresponding interest-rate sensitivities in insurance liabilities. The underlying economic exposures must be located before using profit, equity, statutory surplus or market-value figures as indicators of strength.

Conclusions from the comparative evidence

The first comparative finding is that asset valuation is regime-dependent because measurement objectives differ. IFRS 9, IFRS 13, Solvency II and NAIC statutory principles do not ask identical questions. A reliable account must preserve those differences before calculating a reconciliation. The second finding is that current market-value deterioration and accounting expected credit loss are related but separate phenomena. An observed price decline can reflect changes in rates, premia and liquidity in addition to expected losses; impairment requires application of the relevant recognition rule.

The third finding is that valuation reliability decreases when assumptions substitute for observable evidence without adequate challenge. This is especially important in private credit and structured finance, where delayed price observations, complex contractual waterfalls and correlated collateral shocks can affect modeled value. The fourth finding is that insurers' solvency cannot be understood from asset movements in isolation. Interest-rate sensitivities, guarantees, claims, collateral calls and liquidity restrictions connect the same asset shock to different outcomes for different insurers.

The fifth finding is methodological: the most useful scientific comparison is an explicit bridge among the relevant measurements, with instrument-level inputs, time alignment and reproducible scenarios. This approach does not invent a universal value. It identifies the origin of each number, the evidence that constrains it and the conditions under which it informs policyholder protection, investor reporting or managerial choice. The conclusion remains valid across diverse insurance markets while the legal and reporting details remain jurisdiction-specific.

Conclusion

Asset valuation in insurer balance sheets is the measurement of financial and nonfinancial claims held to support uncertain insurance liabilities under specified accounting, economic and prudential objectives. The central scientific requirement is an auditable connection between asset characteristics, governing measurement basis, available evidence and the purpose for which a reported number will be used. Amortized cost communicates one trajectory of contractual lending returns; fair value estimates a market-participant exchange price; expected credit loss translates prospective shortfalls into a recognized accounting estimate; stress valuation examines the consequences of adverse states; and statutory or prudential valuation interprets the asset side in a framework designed to protect policyholders.

The study's comparative analysis establishes that these measures can diverge without inconsistency in the underlying cash-flow facts. Market interest-rate changes can alter exit prices without an equivalent change in expected default. Credit deterioration can raise allowances while market prices respond to additional liquidity and risk premia. Illiquid holdings can exhibit stable reported marks even when meaningful market inputs have changed. A fall in asset values can be amplified or partly offset by valuation movements in insurance liabilities. Aggregate investment amounts and headline private-credit exposure percentages cannot resolve these differences without common definitions and aligned measurement dates.

The numerical exercises demonstrate these relationships directly. A hypothetical fixed-rate zero-coupon bond's gross amortized-cost accretion differs from its new market price under a yield shock. Simplified ECL estimates change when default probabilities, recovery assumptions and macroeconomic scenario weights change. A hypothetical multi-asset insurer loses 118.7 monetary units of market asset value in the selected stress; an assumed simultaneous 30-unit net liability-value reduction produces an economic surplus decline of 88.7. These results describe the specified constructions only. They should not be interpreted as predictions, regulatory capital charges or empirical estimates of the insurance industry.

For scientific and supervisory work, the strongest practice is to retain position-level contractual evidence, apply the correct jurisdictional accounting classification, quantify expected credit losses with coherent time and probability definitions, validate market-participant prices from defensible inputs, examine forced-sale liquidity separately, and reconcile asset and liability movements on compatible bases. The IFRS financial instruments standard, the fair-value measurement framework, the Solvency II valuation hierarchy, and NAIC statutory principles together illustrate the range of questions that such a reconciliation must answer.

The next step for any insurer-specific empirical application is to obtain a dated asset-level sample, corresponding liability cash flows and actual reporting policies; independently verify market and borrower data; and execute the model under the entity's legally applicable regime. The present monograph completes the conceptual, comparative and illustrative research object. It claims original analytical synthesis and does not claim new theoretical doctrine, invented regulation, an empirical insurer cohort, or independent external peer review.

References

  • Board of Governors of the Federal Reserve System (2019–2024 updated FAQ). Frequently Asked Questions on the New Accounting Standard on Financial Instruments—Credit Losses. Source record.

  • Brunnermeier, Markus K., and Lasse Heje Pedersen (2009). Market Liquidity and Funding Liquidity. Source record.

  • Coval, Joshua, Jakub Jurek, and Erik Stafford (2009). The Economics of Structured Finance. Source record.

  • Duffie, Darrell, and Kenneth J. Singleton (1999). Modeling Term Structures of Defaultable Bonds. Source record.

  • European Commission / EIOPA (2015; consolidated rulebook). Commission Delegated Regulation (EU) 2015/35, Article 10: Valuation Methodology—Valuation Hierarchy. Source record.

  • European Commission / EIOPA (2015; consolidated rulebook). Commission Delegated Regulation (EU) 2015/35, Article 263: Alternative Methods for Valuation. Source record.

  • European Insurance and Occupational Pensions Authority (2024). Report on the Implementation of IFRS 17—Insurance Contracts. Source record.

  • European Insurance and Occupational Pensions Authority (2026). Solvency II: Legislative Framework and 2027 Reform Timeline. Source record.

  • European Parliament and Council / EIOPA (2009; consolidated rulebook). Directive 2009/138/EC, Article 75: Valuation of Assets and Liabilities. Source record.

  • IFRS Foundation (2004, updated through 2026). IAS 36 Impairment of Assets. Source record.

  • IFRS Foundation (2003, updated through 2026). IAS 40 Investment Property. Source record.

  • IFRS Foundation (2011, updated through 2026). IFRS 13 Fair Value Measurement. Source record.

  • IFRS Foundation (2017, effective 2023). IFRS 17 Insurance Contracts. Source record.

  • IFRS Foundation (2005, updated through 2026). IFRS 7 Financial Instruments: Disclosures. Source record.

  • IFRS Foundation (2014, updated through 2026). IFRS 9 Financial Instruments. Source record.

  • IFRS Interpretations Committee (2022). IFRIC Update, March 2022: Lessor Forgiveness of Lease Payments (IFRS 9 and IFRS 16). Source record.

  • International Association of Insurance Supervisors (2025). Global Insurance Market Report 2025: Insurers’ Increasing Investments in Private Credit, Section 3.2. Source record.

  • International Association of Insurance Supervisors (2024/2025; current online tool). Insurance Core Principles and Common Framework: ICP 14 Valuation of Assets and Liabilities. Source record.

  • Longstaff, Francis A., Sanjay Mithal, and Eric Neis (2005). Corporate Yield Spreads: Default Risk or Liquidity? New Evidence from the Credit Default Swap Market. Source record.

  • Merton, Robert C. (1974). On the Pricing of Corporate Debt: The Risk Structure of Interest Rates. Source record.

  • National Association of Insurance Commissioners (2026). Statutory Accounting Principles: Insurance Topics Overview. Source record.

  • National Association of Insurance Commissioners, Capital Markets Bureau (2026; year-end 2025 data). Private Credit and Insurance: A Data-Driven View from State Regulators. Source record.