RESEARCH DOSSIER
Embedded Value and Value of New Business in Life Insurance
Abstract. Embedded value (EV) measures an insurer's shareholder interest in a specified existing book of long-duration insurance business under an explicit valuation methodology; value of new business (VNB) measures the contribution associated with policies originated during a specified reporting interval. The connection between these quantities is economically significant and analytically delicate. EV is a dated stock of modeled value; VNB is a flow contribution whose measurement depends on the timing of sales, new-business boundary, acquisition costs, discounting, contract options, taxes, risk allowances and required capital. This monograph examines the European Embedded Value (EEV) and Market Consistent Embedded Value (MCEV) approaches, comparative traditional EV approaches, and the distinct objectives of IFRS 17 and Solvency II. It develops a reproducible discounted-cash-flow illustration, a value-movement reconciliation, and an interpretive analysis of issuer disclosures in China and the international life market. It establishes that a VNB growth rate alone does not identify a change in underlying franchise value unless bases, business mix, and assumption revisions are made comparable. In the documentary cases studied, disclosed changes in discount-rate conventions and assumptions can be large enough to alter conclusions drawn from nominal growth. The contribution is a source-grounded analytical synthesis and a transparent worked example; no new actuarial theory or empirical population estimate is claimed.
Keywords: Embedded value; European Embedded Value; Market Consistent Embedded Value; traditional embedded value; value of new business; new business profit; life insurance; value of in-force business; required capital; frictional cost of capital; financial options and guarantees; non-hedgeable risk; persistency; mortality; actuarial valuation; economic assumptions; IFRS 17; Solvency II; sensitivity analysis; disclosure comparability.
Introduction
A life insurer accepts long-duration commitments whose premium receipts, benefit payments, operating expenditures, investment outcomes and claims emerge over many years. The economic significance of a policy portfolio therefore cannot be read directly from the premiums written in a single reporting year. Distributable earnings occur across time and depend on what the insurer must retain to support its liabilities. Embedded-value reporting addresses this temporal and capital structure by estimating the shareholders' interest associated with covered existing business. New-business value reporting asks how much an interval of sales contributes under a specified set of assumptions.
The European Embedded Value Principles (CFO Forum, 2016) define EV from the shareholder's perspective and require a declared boundary of covered business. The later market-consistent specification, reflected in the MCEV Principles and Guidance (CFO Forum, 2016), makes the economic treatment of financial risk, options, capital and residual non-hedgeable exposures explicit. The frameworks are principally supplementary performance and valuation reporting conventions; their application must be identified at the reporting-company level.
Three questions direct this investigation. First, what monetary interest is being measured, at what valuation date, over which contracts, and subject to which assumed future cash flows? Second, how should VNB enter the movement from one EV balance to another without being counted twice or confused with sales volume or accounting profit? Third, what can be concluded from sensitivity disclosures and comparisons across companies using different valuation bases? These questions are addressed through primary professional principles, regulatory and accounting instruments, dated public-company disclosures, and selected peer-reviewed actuarial and financial-reporting research. The analysis does not rely on unobserved policy-level datasets, invented company calibrations, or universal causal claims.
A distinction governs the entire monograph: the projected economic value accruing to owners of existing contracts, the measurement of insurance liabilities for prudential supervision, and recognition of contractual service earnings in financial statements answer different institutional questions. Their data and modeling methods can overlap. Their resulting figures have different purposes, boundaries, sign conventions and recognition schedules.
Measurement Object, Sources, and Valuation Boundaries
Historical development and documentary hierarchy
Embedded-value practice emerged in long-term insurance reporting before the modern harmonization of insurance accounting. The American Academy of Actuaries' MCEV practice note (2011) traces prominent UK disclosure practice to the 1980s and notes the development of achieved-profits reporting. The European Insurance CFO Forum issued its original EEV principles in 2004, MCEV principles in 2008, amendments in 2009, and amended EEV/MCEV guidance in April 2016. The CFO Forum's 2016 MCEV Basis for Conclusions documents amendments permitting selective alignment with Solvency II after that regime began in 2016. The 2016 date belongs to the amended principles, not to the first invention of embedded value.
Source hierarchy matters. The CFO Forum's principles state what adherents should measure and disclose; an individual insurer's annual report states what that insurer actually chose and reported; accounting standards prescribe recognition within their scope; regulatory instruments determine prudential obligations; empirical scholarship tests specified hypotheses in defined samples. In this dossier, source-specific conclusions are kept within those boundaries. The Academy's practice note is explanatory practitioner guidance and explicitly does not itself constitute a binding actuarial standard. A jurisdiction's enforceable rules take precedence over analogies drawn from international supplementary reporting.
The valued economic interest
Let the valuation date be time 0 and let C denote a declared population of covered policies and related allocated assets, obligations and capital. In an intentionally simplified traditional EV representation,
where ANW is adjusted net worth allocable to covered business and VIF is the present value of shareholders' distributable interest arising from business in force at time 0 after the relevant allowances. EV is a value estimate as of a particular date, expressed in a specified currency and under particular valuation assumptions. It is not the face amount insured, the total premiums paid by policyholders, the accounting equity of an entire conglomerate, or the market capitalization of the issuing company.
The 2016 EEV definition of covered-business value decomposes EV into free surplus, required capital net of the cost of holding it, and the present value of future shareholder cash flows from covered in-force business. The precise presentation varies with the basis adopted. Some companies start from adjusted net worth and add a risk-adjusted discounted value in force; others separately report free surplus, required capital, financial guarantee costs and non-hedgeable risks. Both approaches require attention to what has already been deducted in the projected distributable earnings.
A contract boundary is fundamental. The in-force set generally consists of contracts already written as of the valuation date, with their contractual renewals or future premiums treated according to the adopted contract-boundary rules. Projected sales of policies that do not yet exist are ordinarily excluded from an EV of current covered business. VNB arises when newly sold contracts enter covered business. A full appraisal of a growing insurer might also assign value to future sales, distribution arrangements or franchise opportunities, but that is an additional appraisal question and cannot be silently added to the EV of in-force business.
Cash-flow definition and actuarial population
For a policy cohort j, suppose P(j,t) is expected premium inflow attributable to the shareholder valuation, B(j,t) is expected insured benefit and claim outflow, E(j,t) is acquisition and maintenance expense, T(j,t) is tax and K(j,t) denotes other distributable cash-flow adjustments under the selected basis. A schematic distributable profit sequence is
This is an accounting identity for a carefully defined projection ledger, not an authorization to subtract insurance liabilities twice or to treat investment principal as profit. For a life policy, each component depends on survival, death, lapse, paid-up conversion, premium continuity, benefit indexation, reinsurance and other contractual transitions. Projection cells require a consistent unit of exposure, issue age, duration, gender or other permissible rating attributes, policy form, currency, underwriting class, policy status and valuation period. Heterogeneous cohorts are aggregated only after all components have been translated into a consistent shareholder-value basis.
Mortality and persistency are conditional processes. For a simple term-life cohort with survival probability p and surrender probability l, next-period premium-paying exposure depends on both insured survival and continuation. A surrender can eliminate future expected claims and future expected premiums simultaneously. Thus the sign of an EV sensitivity to lapse rates can differ between a profitable protection book and a profitable savings book, and can reverse across policy durations. Deterministic average assumptions are useful for transparent baseline calculations; nonlinear contractual options and market-linked guarantees frequently require stochastic modeling.
Interpretation, evidence, and methodological limits
The object of inference determines what evidence is adequate. Annual embedded-value disclosures reveal the company's reported assumptions and modeled outputs; they do not reveal its full policy-level distributions or independently identify causal effects of management decisions. Peer-reviewed tests of market value relevance evaluate investor response and measurement association in a particular sample; they cannot establish that embedded value is a uniquely correct sale price. The published valuation-effects study by El-Gazzar, Jacob, and McGregor (2013) reports evidence of valuation relevance and information effects for embedded-value disclosure, but its abstract and publisher metadata do not supply a contemporary cross-country replication or policy-level validation.
For all later calculations, nominal and real cash flows must be consistently discounted, currency-specific yield curves must be used where applicable, and acquisition expenditures cannot be omitted merely because the year of policy issuance precedes some future profit projections. Comparing values means comparing valuation dates, scope, currency, economic bases, policy cohorts and assumptions together.
Embedded-Value Methodologies and the Capital Architecture
Traditional, EEV, and market-consistent approaches
Traditional embedded value commonly discounts projected shareholder profits using a risk-adjusted discount rate; the choice of rate implicitly compensates for aspects of investment uncertainty and other risks. The EEV framework elaborated risk, capital, options, shareholder interests and disclosures while permitting a broader range of economic valuation implementations. Market-consistent embedded value aims to calibrate financial risk consistently with observable markets and to identify separate adjustments for features that cannot be replicated or hedged fully. These are descriptions of common methodological families, and a firm's own reported compliance statement is decisive for its implementation.
Under the 2016 MCEV definition, a common decomposition is
where FS means free surplus, RC required capital, PVFP the present value of future profits before separately identified adjustments, TVFOG the time value of financial options and guarantees, FCRC frictional costs of required capital, and CNHR the cost of residual non-hedgeable risks. The formula is illustrative of the additive methodology: amounts and signs depend on consistent definitions and permitted alignments, and required capital appears as an allocated asset component while its frictional cost enters as a deduction. Replacing this with a discounted future-profit figure that already contains all costs and subsequently deducting them again would understate value.
The MCEV Basis for Conclusions explains that April 2016 amendments allow, without compelling, alignment of selected assumptions and projection methods with Solvency II. Permitted alignment must be disclosed; it is therefore unsafe to assume that every item called MCEV follows the same model as the insurer's regulatory balance sheet. Market consistency in asset pricing does not eliminate the need for explicit decisions on taxes, expenses, contract boundaries, policyholder behavior, residual risks or surplus allocation.
Economic scenarios and options
A guaranteed crediting rate, participating bonus, minimum death benefit or surrender guarantee can have a nonlinear payoff. If G is a guarantee payoff depending on future market state X and Q is an appropriately calibrated risk-neutral pricing measure for hedgeable market exposures, the market-consistent value is schematically
with extensions for intermediate exercise, dynamic bonuses, mortality transitions and state-dependent policyholder behavior. In a purely deterministic projection, a central market scenario may yield a small guarantee cost even where the probability-weighted cost across severe market states is material. TVFOG captures the additional value attributable to optionality and convexity, taking into account the actual product's design and assumptions. Bauer, Kiesel, Kling, and Ruß (2006) use risk-neutral valuation of participating contracts under German-style bonus and guarantee rules. Zaglauer and Bauer (2008) extend participating-contract valuation to stochastic interest-rate environments. These papers demonstrate why rate dynamics and bonus mechanics matter; they do not determine any particular insurer's unobserved TVFOG.
A market-consistent implementation needs documented financial model calibration, asset returns appropriate to projection conventions, arbitrage-consistent reference prices for hedgeable exposures, interest-rate dynamics, simulation error assessment and disciplined treatment of model extrapolation. A single discount-rate stress is useful disclosure but is not a substitute for distributions of states when the contract contains significant nonlinear choices.
Required capital and its opportunity costs
The shareholder cannot necessarily distribute every asset that supports a life portfolio at the valuation date. Required capital may be tied up by regulatory requirements, internal capital policy, rating objectives or contractual restrictions. Capital also earns investment returns and may eventually be released. The valuation problem therefore concerns the incremental friction of retention, including investment expenses, tax drag, and the modeled cost of risks not otherwise allowed for, rather than the physical disappearance of the entire capital balance.
Required capital is not identical to the Solvency Capital Requirement (SCR) in every framework or market. Even where the same projected capital path is used for operational efficiency, the economic assumptions and accounting definitions require reconciliation. The EIOPA presentation of Solvency II Article 77 states that prudential technical provisions are composed of a best estimate and a risk margin, and requires appropriately modeled probability-weighted liability cash flows. The corresponding risk-margin calculation in Delegated Regulation (EU) 2015/35 Article 37 references a cost-of-capital charge on projected future SCR associated with transferred obligations. That risk margin serves a prescribed prudential liability objective. It must not be relabeled automatically as an MCEV frictional cost of required capital or as the cost of all residual non-hedgeable risks.
Shareholder value and financial-reporting measurement
The IFRS 17 standard overview identifies fulfilment cash flows, a risk adjustment for non-financial risk, and a contractual service margin (CSM) for unearned profit. The standard generally recognizes the profit from profitable insurance contract groups as services are provided, while identifying losses on onerous groups according to its requirements. IFRS 17 has applied for annual periods beginning on or after January 1, 2023, subject to applicable adoption arrangements. The CSM represents a financial-reporting mechanism for timing unearned insurance-service profit, not a direct valuation of corporate shareholder franchise or the EV of a business segment.
The IFRS Foundation's IFRS 17 key terms distinguish insurance risk, fulfilment cash flows and the compensation required for bearing non-financial risk. A reconciliation from CSM to VNB would require changes for contract groupings, discounting, acquisition cash flows, shareholder capital, financial guarantees, risk adjustments, coverage periods, taxes and products outside the covered EV perimeter. A high initial CSM may accompany substantial capital consumption; a positive VNB can coexist with initially low statutory earnings. These relationships follow from differing measurement functions and cannot be reduced to a universal conversion multiplier.
Value of New Business: Cohort Economics and Measurement
Definition, issue cohorts, and date conventions
Value of new business is the value contributed by life policies sold or written in a specified period, measured using a declared valuation convention. A standard economic representation for a cohort issued during the interval is
Under a fully consistent source ledger, acquisition outlays at sale must appear once, either as a time-zero outflow in the cohort cash-flow projection or in the separately modeled cost terms. Some presentations measure VNB at the individual point of sale; others aggregate new business and revalue it at the reporting date. The American Academy's VNB practice discussion explicitly recognizes different reporting-date conventions and the need to disclose the valuation assumptions. The nominal value will differ across conventions even when the policy economics are otherwise identical.
New-business volume and new-business value answer different questions. Annual premium equivalent (APE) commonly weights regular first-year premiums at 100% and single premiums at 10%, as explained in Prudential's 2024 annual filing. APE is a volume convention rather than a discounted shareholder profit. Present value of new business premiums (PVNBP) uses a different denominator, aggregating discounted expected new-business premium inflows subject to the issuer's stated conventions. Accordingly,
The two margins need not be equal and should always be reported with their denominator and period. Negative-margin volumes can increase premium production while reducing shareholder value. A positive-margin business can consume near-term cash and required capital because acquisition costs and initial capital support occur before much of the future profit arrives.
A reproducible new-business example
Consider a hypothetical cohort of newly issued life policies, measured in millions of an unspecified currency. Assume that acquisition and initial setup consume 1.80 at issue; expected after-tax shareholder profits available at the ends of the next five years, excluding separately identified capital and guarantee adjustments, equal 1.00, 1.50, 1.50, 1.40, and 1.20. Assume a flat annual effective discount rate of 4.5% solely to illustrate the present-value arithmetic. The initial actuarial profit calculation is
Suppose separate, non-overlapping model outputs—assumed rather than inferred from data—are TVFOG = 0.22, FCRC = 0.30 and CNHR = 0.30. The example's VNB is then 3.161905 million. For an assumed APE of 20.00 million, the new-business margin on APE is 15.8095%. If instead the reporting denominator were 80.00 million of PVNBP, the VNB-to-PVNBP margin would be 3.9524%. Neither figure is a general market benchmark; their different percentages arise entirely from denominator choice and the deliberately specified cash flows.
| Hypothetical new-business component | Million currency units |
|---|---|
| Issue-date acquisition and setup | -1.800000 |
| Discounted future shareholder profits before separate adjustments | +5.781905 |
| Present-value future profit net of issue-date cost | +3.981905 |
| Time value of options and guarantees | -0.220000 |
| Frictional capital cost | -0.300000 |
| Residual non-hedgeable risk allowance | -0.300000 |
| Value of new business | +3.161905 |
The calculation is deliberately parsimonious: it demonstrates discounting and the absence of double counting, while abstracting from multiple issue dates, capital releases, investment market calibration, mortality modeling, tax jurisdictions and stochastic guarantee exercise. It is a reproducible teaching example, not a calibrated insurer valuation. If the future profits already incorporate the time value of financial guarantees and capital friction, the separate deductions must instead be set to zero or redefined.
Growth quality and business mix
VNB can change because the insurer writes more policies, changes products or channels, reduces acquisition costs, improves persistency, alters reinsurance, modifies policyholder benefits, changes capital requirements or revises its assumptions. A useful decomposition starts from a constant-basis value per unit of comparable new-business volume and then assesses mix and modeling changes. One may write an interpretive algebraic identity for two dates as
where Q is a consistently defined volume measure and M = VNB/Q is the corresponding margin. This identity separates nominal volume and average margin changes when Q and M are comparably defined. It does not identify why the margin moved. Where the valuation basis or premium denominator changes, a restatement or disclosed bridge is needed before either term can be read as an operating contribution.
Management should also consider the distribution of profitability across cohorts rather than a single positive average. A product with attractive central-scenario VNB may have substantial downside under adverse lapses, interest rates, longevity, claims concentration or capital stress. The survival of a nominally profitable business model depends on distributable cash and continued solvency, not only on the central estimate of future gains.
From sales to closing embedded value
VNB is a bridge component between opening and closing EV, conditional on measurement conventions. In a conceptually consistent value movement analysis,
where ER represents the expected return or unwinding on the opening EV, OV operating experience variance, EVAR economic experience and assumption remeasurement where separately defined, AC remaining assumption and model changes, CF external capital flows and distributions, FX currency effects, and OTH any disclosed residual items. This is an intentionally generic reconciliation, with line labels varying by issuer. It is not an IFRS statement of comprehensive income. VNB measures value attributed to new business; it is not added again as an independent asset outside the closing in-force and net-worth figures.
If assumptions are set at point of sale, some post-sale experience on newly written contracts may be reported in variance lines; an end-of-period VNB convention can incorporate more of these changes directly into VNB. Cross-company comparisons therefore require careful reading of the movement-analysis definitions, including the placement of expense overruns, interest unwind and economic variances.
Sensitivities, Disclosures, and Model Governance
The scope of an informative sensitivity
Embedded-value numbers derive from models and assumptions that are consequential precisely because the contracts are long-lived. Discount curves affect the current value of future shareholder cash flows; mortality and morbidity affect benefit incidence; lapse affects both premium continuation and benefit outflows; expenses determine maintenance and acquisition economics; tax and reinsurance affect available distributable cash; capital rules can change the timing and cost of remittances. The EEV disclosure and sensitivity guidance identifies these as relevant topics for explaining reported value and its uncertainty.
A sensitivity must state what was stressed, by how much, against which baseline, over what projection horizon, and which other variables were held fixed. Financial options can change value nonlinearly; simultaneous interest-rate and lapse movements may differ materially from the sum of independent one-variable stresses. A sensitivity is a conditional model output, not a confidence interval and not an observed probability distribution unless the statistical structure establishes that interpretation.
Discount sensitivity of the illustrative portfolio
Consider a separate illustrative in-force portfolio with nominal end-year shareholder profits of 3.20, 3.60, 3.40, 3.00 and 2.60 million over five years, valued with a flat effective rate of 4.5%. Its PVFP, before separate market-option and capital adjustments, is 13.940294 million. With the same profit amounts and discount rates of 3.5% and 5.5%, PVFP becomes 14.322488 and 13.574090 million respectively. Thus a 100-basis-point increase in this narrow discount convention reduces PVFP by approximately 0.366204 million, or 2.63% of the 4.5% baseline, while a 100-basis-point decrease increases it by approximately 0.382194 million, or 2.74% of baseline.
| Effective discount rate | Illustrative PVFP, million | Difference from 4.5% baseline |
|---|---|---|
| 3.5% | 14.322488 | +0.382194 |
| 4.5% | 13.940294 | 0.000000 |
| 5.5% | 13.574090 | -0.366204 |
These calculations isolate pure present-value discounting. They do not simultaneously change future investment returns, credited bonuses, reserve adequacy, liability behavior, capital needs or the stochastic value of guarantees. Those interactions are a principal reason why a full insurer sensitivity can differ in both size and sign from a simple discounted deterministic profit stream.
Capital, liquidity and non-hedgeable risk
Sensitivity to required capital depends on whether capital is externally binding, internally targeted, liquid, transferable between subsidiaries or subject to local dividend restrictions. A dollar of locked-up capital can be represented in EV as a component of adjusted net worth while its modeled friction reduces future distributable value. The expense of keeping capital in place should not be confused with its face amount. Where a valuation aligns selected capital assumptions with Solvency II, the methodological choice should be distinguished from supervisory compliance. EIOPA's Article 37 risk-margin rule and the CFO Forum's MCEV adjustments address different measurement objects even if both involve projected capital and cost rates.
Residual non-hedgeable risk includes exposures for which relevant losses cannot be precisely replicated or hedged using available market instruments. Mortality and persistency have portfolio diversification and underwriting dimensions; catastrophic shifts, model error, parameter uncertainty and operational exposure can remain. Their allowance must be documented so that the same risk is not recognized independently in an inflated risk discount rate, a conservative cash-flow assumption and a separate cost term without an explicit reconciliation.
Governance and auditability
A robust EV process documents the valuation perimeter, methodology version, data valuation date, product and model mapping, assumption-setting committee, experience studies, economic scenario calibration, tax and regulatory changes, scenario convergence, reconciliations to opening values and movements to closing values. Independent review arrangements and disclosure of material methodology exceptions strengthen the integrity of an inherently assumption-dependent estimate. The professional CFO Forum principles distinguish their formal principles from implementation guidance and insist on disclosure of departures; the MCEV Basis for Conclusions describes that structure.
An actuarial appraisal standard may also govern work performed by particular professionals in a specific jurisdiction and engagement. The U.S. Actuarial Standards Board's ASOP No. 19 sets professional expectations for appraisals of insurance businesses; an appraisal of an entire enterprise is a different assignment from a CFO Forum EV disclosure. The Academy's 2011 practice note identifies its own nonbinding status. The legal force of accounting, prudential and professional requirements should be assessed independently in each country.
A disclosure is especially informative when it enables readers to trace three bridges: from prior-period EV to current EV; from gross or pre-capital projected profits to reported value; and from the preceding year's VNB to current VNB on comparable assumptions. Without those bridges, confidence in a number can exceed the evidence that generated it.
Comparative Evidence and Jurisdictional Implementation
New China Life: disclosed assumptions and arithmetic
The listed insurer New China Life disclosed in its 2024 annual report that its embedded value estimate followed the China Association of Actuaries' 2016 appraisal standards. Its report used deterministic discounted cash flow with a risk-adjusted discount rate, with an allowance for risks embedded in that rate. It defined embedded value as adjusted net worth plus the value of in-force business after the cost of required capital. For December 31, 2024, the company reported adjusted net worth of RMB 189,233 million, value of in-force business before required-capital cost of RMB 104,223 million, and capital cost of RMB 35,008 million. Their difference produces in-force value of RMB 69,215 million, and the sum with adjusted net worth produces reported EV of RMB 258,448 million. The reported VNB after required-capital cost for the year was RMB 6,253 million.
This example makes two points observable rather than hypothetical. First, capital cost was a separately visible deduction from projected in-force value. Second, the new-business figure was a contribution to value, not the entire year-end EV and not a sales-premium volume measure. The same report states that its valuations are assumption-dependent and should not substitute for financial results under other reporting bases. Its risk-discount rate for the 2024 calculation was 8.5%, which must not be confused with the risk-free discount curve or treated as a globally mandated parameter.
The report's disclosed EV movement from RMB 250,510 million at the beginning of 2024 to RMB 258,448 million at year end includes VNB +6,253, expected return +16,514, operating experience variances +4,674, economic experience variances +20,357, operating assumption changes -2,220, economic assumption changes -35,458, capital/dividend flows -4,336, other items +211, and non-life business value change +1,945, all in RMB millions. These component signs and the reconciliation are company-specific rather than generic economic laws. The net increase of RMB 7,938 million is far smaller than the sum of positive operating and investment contributors because large negative economic assumption revisions offset them. The sum of the individually rounded component amounts listed here is RMB 258,450 million, RMB 2 million above the company's reported closing figure; the discrepancy is consistent with the report's warning that displayed figures may not add due to rounding. This is direct evidence that a rising EV or VNB requires a movement analysis before being attributed to operations alone.
New China Life's 2025 annual report subsequently disclosed EV of RMB 287,840 million and one-year VNB of RMB 9,842 million for December 31, 2025, compared with RMB 258,448 million and RMB 6,253 million for 2024. The arithmetical year-over-year changes are approximately +11.37% for EV and +57.40% for VNB. They are observed disclosed aggregate changes; the two percentages do not establish how much arose from constant-assumption productivity, revised financial assumptions, channel mix, product changes or shifts in required capital. A separate matched-basis bridge would be necessary to make those causal allocations.
Prudential: methodology transition as a comparability test
The international insurer Prudential stated in its 2024 Form 20-F that it intended to replace its EEV reporting basis with traditional embedded value (TEV) reporting beginning in 2025. The stated goals included greater comparability with important peers and less economic volatility in published EV measures. The company identified differences relating to discount rates and treatment of central head-office expenditure while retaining many operating assumptions. Its plan illustrates that changes in reported EV or VNB may arise from a reporting-basis transition even without an equivalent change in the economics of each policy sold.
In its 2025 full-year results, filed in March 2026, Prudential reported TEV-basis new-business profit of USD 2,782 million and a new-business margin of 42% for full-year 2025, describing 12% growth in profit on its stated comparison basis. The figure should be attributed to the company's TEV framework and its own growth comparatives rather than placed unadjusted next to a Chinese deterministic EV or a hypothetical MCEV valuation. Prudential's 2025 interim reporting also defines TEV return and new-business profit over embedded value using explicitly defined equity and intangible exclusions. These denominators are distinct from APE and PVNBP margins.
International comparability and the limits of aggregate numbers
The New China Life and Prudential cases provide a particularly useful contrast. One issuer's disclosed EV relied on a deterministic risk-discount-rate approach under local professional conventions; the other changed its own disclosed method from EEV to TEV, while continuing to report separate IFRS and capital measures. The figures differ by reporting currency, entity scope, contract portfolio, issue-date conventions, interest rate structure, required capital basis and macroeconomic conditions. A numerical ranking using their headline VNB amounts would therefore measure size and currency choice at least as much as profitability. A comparison of their VNB margins would still depend on denominator and basis matching.
The CFO Forum's April 2016 MCEV amendments expressly recognize latitude in alignment with Solvency II, while expecting explanation of assumptions and sensitivities. Such latitude improves reporting feasibility across national capital regimes but leaves a continuing comparability task for analysts. The appropriate analytical response is a transparent comparability table: valuation date; scope of covered business; currency; economic basis; discount curve or rate; treatment of options and guarantees; mortality and lapse assumptions; capital definition; acquisition and maintenance expenses; taxes; and new-business denominator. An absent item belongs in an explicit 'not disclosed in materials reviewed' category rather than being filled through inference.
Empirical evidence and its limits
The peer-reviewed embedded-value disclosure study (El-Gazzar, Jacob, and McGregor, 2013) reports evidence that disclosed EV has valuation relevance and information effects for equity investors. It supports the proposition that market participants may find the measure informative alongside traditional statements. It does not establish that every EV report is accurate, internationally comparable, immune to management assumption choices or a sufficient basis for company valuation. Similarly, the option-valuation research of Bauer and colleagues (2006) and Zaglauer and Bauer (2008) supports the importance of explicit model structure for guarantees without supplying an empirical estimate of the companies examined here.
The documentary comparison in this dossier is not a statistical sample of the global life insurance market. It is a comparative reading of identified primary reports. It yields sound claims about definitions, reported amounts, model choices and the conditions for interpretation, but cannot support estimates of global insurer profitability, policyholder welfare, or causation in company stock prices.
Integrated Scientific Findings and Reproducible Analysis
The stock-flow interpretation
The first finding follows from the algebra of valuation dates. EV is a stock evaluated over a defined in-force population, and VNB is the value attributed to issuance in a period. The opening EV is remeasured under the passage of time and updated experience. The newly written book moves into the closing in-force population. Therefore the flow VNB must be traced through an EV movement account and must not be multiplied indiscriminately by a growth rate or added to closing EV as though the new policies were not already included. This conclusion is an analytical reading of established CFO Forum definitions, demonstrated by New China Life's reported EV roll-forward; it is not a proposed new theoretical framework.
The second finding concerns valuation boundaries. Suppose two otherwise identical insurers choose different definitions of covered contracts, treat head-office expenses differently or adopt different capital restrictions. Their embedded values can diverge even before market conditions are considered. A higher reported VNB in one entity is not decisive evidence of more economic value generated per unit of capital. The comparison needs a common definition of risk-bearing population and shareholder-distributable cash.
Reproducible in-force illustration
Take an illustrative 5-year portfolio, independent from the new-business example above. At valuation, free surplus equals 24.00 million and required capital equals 12.00 million. The end-year shareholder-profit stream is 3.20, 3.60, 3.40, 3.00 and 2.60 million, with PVFP at 4.5% of 13.940294 million. Separately identified, non-overlapping valuation allowances are TVFOG 1.10 million, FCRC 0.90 million and CNHR 0.85 million. The illustrative MCEV is
The value is an internally consistent classroom arithmetic example, not an estimate of capital or franchise value for any named insurer. Under the fixed-cash-flow sensitivity in the prior chapter, the 5.5% discount assumption would lower the hypothetical EV to 46.724090 million, holding every other amount fixed. The illustration shows why the capital asset component and its cost must be represented separately, while also showing how little a one-variable sensitivity says about the full market-consistent value if economic assumptions interact.
Assumption changes and interpretation of growth
The third finding concerns reported VNB growth. A nominal increase can be decomposed arithmetically into comparable volume and comparable margin, but cannot be attributed causally without matching the economic and operating assumption bases. The 2024 movement analysis of New China Life reports economic assumption changes of minus RMB 35,458 million alongside VNB contribution of plus RMB 6,253 million. The 2025 headline growth in its VNB, though substantial, cannot by itself identify stable-economics profitability improvement. Prudential's voluntary transition from EEV to TEV supplies an independent demonstration that method choice is a live disclosure issue rather than a purely theoretical concern.
The meaning of market consistency
The fourth finding is that market consistency is a specified valuation discipline for relevant financial risks, not a claim that any single EV number equals an observable market transaction price. The methods address shareholder economic value under assumptions; stock market price additionally reflects the market's beliefs about future sales, management, capital distribution, taxes, liabilities not in covered business, control, governance and market liquidity. The literature on financial options supports sophisticated stochastic valuation where embedded guarantees are nonlinear; deterministic comparisons remain meaningful for limited purposes if labeled and interpreted accordingly.
Reporting, prudential protection, and decision use
The fifth finding concerns the institutional location of each number. EV and VNB measure supplementary shareholder-interest estimates; the IFRS 17 fulfilment cash flows and CSM organize recognized insurance obligations and service-profit timing; Solvency II best estimates and risk margins address prudential liability measurement and capital protection. The difference is substantial for interpretation of dividends and long-term guarantees. Positive EV and positive VNB may exist alongside near-term funding pressure; shareholders' projected returns depend on continuing compliance with enforceable capital and policyholder-obligation constraints. Neither accounting CSM nor regulatory technical provisions can be mechanically substituted for VNB.
These five findings arise from tested definitions, public disclosures and transparent arithmetic. They are original in their comparative arrangement and explanatory synthesis, while remaining within established valuation mathematics. No novel doctrine, proprietary empirical discovery or universally optimal EV formula is asserted.
Conclusions
Embedded value expresses the estimated shareholder interest associated with a declared portfolio of existing long-duration insurance business. Value of new business quantifies the modeled economic contribution from contracts newly written in a defined interval. Their connection is operationally expressed through the closing-value roll-forward, which must distinguish new-business contribution, expected emergence of earnings, experience variances, economic and operating assumption changes, capital transactions and currency effects. This connection supplies the correct frame for interpreting growth and profit margins.
The EEV and MCEV principles establish a detailed vocabulary for free surplus, required capital, in-force earnings, financial options, frictional capital costs, non-hedgeable risks and disclosure. Traditional risk-discount-rate approaches follow a different allocation of risks within discounting and projected cash flows. A valuation basis becomes intelligible through its declared modeling boundaries and assumptions. For long-term guarantees, stochastic models are frequently necessary to measure nonlinear financial exposure, while transparent deterministic examples remain useful as limited illustrations.
The investigated public-company reports establish the practical consequences. New China Life's reported 2024 EV, capital-cost deduction and movement analysis show that assumption changes can materially offset the contribution of new business. Its 2025 disclosures demonstrate changes in reported aggregate EV and VNB, while leaving causal attribution dependent on a fuller comparable-basis bridge. Prudential's transition from EEV to TEV confirms the importance of explicitly identifying methodology when studying trends or comparing firms. Published empirical research indicates that EV disclosures can be information-relevant, subject to the limits of the study and the quality of the underlying estimates.
For scholarly interpretation, an embedded-value assertion is complete only when it states the valuation date, covered population, currency, cash-flow rules, discount and risk assumptions, capital basis, options treatment, new-business boundary, profit-margin denominator and sensitivity context. Its documentary support should distinguish professional principles, legal requirements, accounting standards, company disclosures and empirical tests. This is the methodological basis for evaluating embedded value and VNB as meaningful, reproducible and properly bounded measures within international life-insurance research.
References
Actuarial Standards Board. (2005; revised 2011). Actuarial Standard of Practice No. 19: Appraisals of Casualty, Health, and Life Insurance Businesses. https://www.actuarialstandardsboard.org/asops/appraisals-casualty-health-life-insurance-businesses/
American Academy of Actuaries, Life Financial Reporting Committee, MCEV Subgroup. (2011). Practice Note on Market Consistent Embedded Values (March 16). https://actuary.org/wp-content/uploads/2025/05/MCEV%20Practice%20Note%20Final%20WEB%20031611.pdf
Bauer, D., Kiesel, R., Kling, A., & Ruß, J. (2006). Risk-neutral valuation of participating life insurance contracts. Insurance: Mathematics and Economics, 39(2), 171–183. https://www.sciencedirect.com/science/article/pii/S016766870600045X
European Insurance CFO Forum. (2016a). European Embedded Value Principles and Guidance (April). https://www.cfoforum.eu/mediaitem/587a904d-c334-4de7-b2ca-c09dc2304814/CFO-Forum_EEV_Principles_and_Guidance_April_2016.pdf
European Insurance CFO Forum. (2016b). Market Consistent Embedded Value Principles and Guidance (April). https://www.cfoforum.eu/mediaitem/3b8d66a9-7752-497e-8f71-2d0804ae9e73/CFO-Forum_MCEV_Principles_and_Guidance_April_2016.pdf
European Insurance CFO Forum. (2016c). Market Consistent Embedded Value Basis for Conclusions (April). https://www.cfoforum.eu/mediaitem/ff342697-93db-4fff-b321-00581612fae9/CFO-Forum_MCEV_Basis_for_Conclusions_April_2016.pdf
European Insurance and Occupational Pensions Authority. (2021). Solvency II Single Rulebook: Directive 2009/138/EC, Article 77 — Calculation of Technical Provisions. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2160_en
European Insurance and Occupational Pensions Authority. (2021). Solvency II Single Rulebook: Commission Delegated Regulation (EU) 2015/35, Article 37 — Calculation of the Risk Margin. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2450_en
El-Gazzar, S. M., Jacob, R. A., & McGregor, S. (2013). The valuation effects of embedded value disclosure by life insurers. International Journal of Economics and Accounting, 4(1), 26–53. https://www.inderscience.com/info/inarticle.php?artid=50592
IFRS Foundation. (2023; webpage updated subsequently). IFRS 17 Insurance Contracts: Standard Overview, Application and History. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-17-insurance-contracts/
IFRS Foundation. (n.d.). IFRS 17 Insurance Contracts: Key Terms. https://www.ifrs.org/supporting-implementation/supporting-materials-by-ifrs-standards/ifrs-17/key-terms/
New China Life Insurance Company Ltd. (2025). Annual Report 2024, Embedded Value and Independent Actuary's Review, pp. 48–55. Hong Kong Exchanges and Clearing Limited filing. https://www1.hkexnews.hk/listedco/listconews/sehk/2025/0410/2025041000287.pdf
New China Life Insurance Company Ltd. (2026). Annual Report 2025. Hong Kong Exchanges and Clearing Limited filing. https://www.hkexnews.hk/listedco/listconews/sehk/2026/0409/2026040900722.pdf
Prudential plc. (2025a). Annual Report 2024, Form 20-F, Embedded Value and New Business Reporting. U.S. Securities and Exchange Commission filing. https://www.sec.gov/Archives/edgar/data/1116578/000110465925027998/puk-20241231x20f.htm
Prudential plc. (2025b). Half Year 2025 Results, Form 6-K (August 27), Traditional Embedded Value Metrics. U.S. Securities and Exchange Commission filing. https://www.sec.gov/Archives/edgar/data/1116578/000110465925083706/tm2523773d1_6k.htm
Prudential plc. (2026). Full Year 2025 Results, Form 6-K (March 18). U.S. Securities and Exchange Commission filing. https://www.sec.gov/Archives/edgar/data/1116578/000165495426002350/a0358x.htm
Zaglauer, K., & Bauer, D. (2008). Risk-neutral valuation of participating life insurance contracts in a stochastic interest rate environment. Insurance: Mathematics and Economics, 43(1), 29–40. https://www.sciencedirect.com/science/article/pii/S0167668707001047