RESEARCH DOSSIER
Deferred Taxes and Loss-Absorbing Capacity in Solvency
Abstract. Deferred taxation links the measurement of an insurer's assets and liabilities, the tax consequences of adverse outcomes, and the credibility of regulatory capital. A capital loss may diminish taxable income, reverse a deferred tax liability, or create a deferred tax asset. Only a tax effect that can be substantiated under the applicable legal and supervisory regime can provide recognized loss absorption. This monograph examines the legal architecture, economic mechanisms, projected taxable-profit tests, actuarial assumptions, supervisory governance, and cross-jurisdictional comparability of deferred taxes in insurance solvency. Its central finding is that tax loss absorption is conditional on a jointly consistent post-stress balance sheet and enforceable tax-utilization pathway. A nominal tax rate applied mechanically to a capital shock measures a hypothetical ceiling, rather than the value of a demonstrated resource. Close examination of Solvency II Articles 15, 205–207, the 2019 amendments, EIOPA guidance, the United Kingdom's PRA practice, IAS 12, and the IAIS Insurance Capital Standard reveals distinct approaches to the same recoverability problem. Fully disclosed hypothetical calculations show how loss allocation, reversals, projection horizons, and taxable-profit evidence change the Solvency Capital Requirement and reported coverage. The analysis develops reproducible tests for evidence, consistency, sensitivity, and governance without proposing a new regulatory theory or treating illustrative results as observations of actual insurers.
Keywords: Solvency II; deferred tax assets; deferred tax liabilities; loss-absorbing capacity; LAC DT; solvency capital requirement; taxable profit; insurance taxation; IAS 12; prudential valuation; recovery; stress testing; insurance capital standard; actuarial governance; regulatory comparability.
Introduction
A tax effect becomes relevant to insurer solvency when the realization of a loss changes an enforceable future tax obligation or creates a tax benefit that can actually be used. The recognition question is therefore simultaneously legal, financial, actuarial, and evidentiary. The legal system determines which transactions generate deductions, what can be carried backward or forward, which entities can consolidate, and when tax claims expire. Accounting and prudential frameworks determine which differences are recognized at the reporting date. Actuarial and financial projections determine whether expected future taxable income remains credible after the modeled shock. Supervisors determine whether an asserted tax absorption effect meets the prudential standard of proof.
Under the Solvency II loss-absorption rule, unexpected losses may be compensated through changes in technical provisions, deferred taxes, or both. That legislative connection carries a material measurement implication: the before-stress tax balance sheet and the after-stress tax balance sheet must correspond to the same underlying economic events. The Solvency II capital calibration concerns an adverse one-year event at a 99.5% confidence level. Recognizing a tax benefit derived from that event demands evidence that the tax benefit remains usable in circumstances in which the insurer has suffered the event itself.
This Research Dossier studies the independent prudential object of deferred taxes and their loss-absorbing capacity. Its boundary excludes comprehensive corporate tax advice, jurisdiction-specific filing instructions, and the general accounting of deferred income taxes as a standalone subject. It treats IAS 12 as an essential accounting reference and makes clear where prudential recognition adds a stress-specific condition. Its geographical scope covers Solvency II in the European Union, the related but separately maintained United Kingdom framework, and internationally active insurance groups under the IAIS Insurance Capital Standard (ICS). Other jurisdictions are discussed as comparability problems without inferring identical legal requirements.
The method combines version-specific documentary interpretation, accounting-to-solvency reconciliation, dimensionally explicit algebra, constructed examples, sensitivity analysis, and critical interpretation of supervisory publications. Regulatory propositions are attributed to authentic legal instruments or official supervisory rulebooks; supervisory guidance is treated according to its legal status; IFRS material is used for financial-reporting concepts; the ICS describes a separate group standard. Calculations are hypothetical, transparently specified, and reproducible. The literature review is targeted rather than a claim of exhaustive systematic coverage. The primary analytical contribution is an integrated explanation of when legal tax capacity, projected taxable income, and recognized capital relief coincide, and why their divergence matters for capital adequacy.
Deferred Tax Positions and Capital
Tax bases, temporary differences, and recoverability
A deferred tax liability (DTL) generally represents the future tax consequence of a taxable temporary difference. A deferred tax asset (DTA) can arise from a deductible temporary difference, an unused tax loss, or an unused tax credit, subject to recognition criteria. The IAS 12 standard overview explains temporary differences by comparing the carrying amount of an asset or liability with its tax base. This distinction matters because a mark-to-market loss may affect an insurer's economic capital without creating a deduction of the same amount, in the same tax period, or even under the same tax regime.
For illustration, suppose an investment has an economic carrying value of 140, a tax base of 100, and a legislated tax rate of 25%. If its recovery is taxable and no exception or offset applies, the 40 taxable temporary difference corresponds to a DTL of 10. If a market shock reduces the carrying amount to 110 while the tax base remains 100, the DTL becomes 2.5. The change of 7.5 is an economically intelligible tax relief, but its legal characterization and prudential recognition must be confirmed by the relevant rules. A different investment with a tax-exempt gain, an exempt loss, or a tax base that moves with the carrying amount would produce a different tax outcome. The example demonstrates a tax-base reconciliation, not a universal formula for admissible capital.
Deferred taxes exhibit timing and jurisdictional specificity. Current tax receivables, refundable prior-period taxes, deductions carried forward to future periods, and reductions of future DTLs have different pathways to economic realization. A nominal tax asset is not cash on the shock date. A potential carryback depends on assessed historical profits and refundability rules; a carryforward depends on future taxable income within a permitted period; a DTL reversal depends on whether the relevant taxable temporary difference genuinely disappears. Group relief requires tax law and enforceable arrangements. These distinctions should appear in the evidence supporting any prudential credit.
Solvency balance sheets and accounting reconciliation
A solvency balance sheet uses its own recognition and valuation principles. The Solvency II deferred-tax valuation rule, Article 15 of Delegated Regulation 2015/35, requires deferred taxes to be recognized and valued in connection with assets and liabilities valued for solvency or tax purposes; it describes valuation from the differences between prudential amounts and their tax values. Article 15(3) confines a positive DTA value to circumstances in which future taxable profit is probable, taking account of legal and regulatory carryforward time limits. Consequently, a solvency DTA may differ from the amount in published financial statements even when both are prepared carefully.
Accounting and prudential figures should be reconciled asset by asset or by appropriately homogeneous classes of temporary differences. The analyst begins with accounting carrying amounts and their tax bases, identifies Solvency II valuation adjustments, maps their tax consequences, determines existing recognized DTA and DTL, and explains any divergences in offsetting or recognition. A source of confusion is the temptation to treat the entire deferred-tax line on a financial-statement balance sheet as already eligible for Solvency II stress absorption. Balance-sheet recognition answers a reporting-date question. Loss-absorbing capacity answers a separate conditional question about the effect of a specified severe loss.
The valuation bridge also requires a distinction between two mechanisms. The first is ordinary DTA/DTL recognition on the current solvency balance sheet, affecting excess assets over liabilities and therefore own funds. The second is a hypothetical change in deferred taxes under the prescribed shock, affecting the capital requirement through the LAC DT adjustment. A single economic tax event must not be included once as an already recognized asset and again as an incremental reduction in the SCR. The general adjustment provision, Article 205, separately identifies the technical-provisions and deferred-tax adjustments.
Recognition before and after stress
The essential accounting object can be written as DT(0), the recognized net deferred tax position at the valuation date, and DT(s), the comparable recognized position after the prescribed shock s. If DT represents DTA less DTL, an increase in DT corresponds to an economic tax gain. Regulatory LAC DT is conventionally expressed as a nonpositive adjustment in the SCR expression; it therefore has the opposite sign to a positive improvement in the net tax position. This convention must be recorded in model documentation because insurers, reporting systems, and analytic papers sometimes use alternative signs.
The post-stress calculation requires remeasurement of taxable and deductible temporary differences, revaluation of DTL and DTA, and separate tests of recoverability. A change in interest rates can simultaneously affect asset valuations, insurance liabilities, risk-free projections, and the tax consequences of future taxable gains. A catastrophic underwriting loss may generate deductible claims, but the deductibility and timing of booked insurance reserves are matters for local tax law. A lapse shock may change the projection of premium-related profit, not only current asset values. A transfer between legal entities may change which taxpayer possesses the carryforward. Such pathways show why a single proportional haircut applied to a pre-stress DTA is rarely a complete model.
Stress and Tax Loss Absorption
Statutory architecture of the adjustment
The legal definition sits in Article 207 of Delegated Regulation 2015/35. Under the standard formula, the instantaneous loss used to measure the tax adjustment corresponds to the Basic Solvency Capital Requirement (BSCR), plus the adjustment for loss-absorbing capacity of technical provisions, plus the operational-risk capital requirement. The technical-provisions adjustment is usually nonpositive. Hence, if BSCR = B, technical-provisions adjustment = T ≤ 0, and operational requirement = O, the pre-tax loss relevant to Article 207 can be expressed, within the standard-formula representation, as L = B + T + O. The resulting SCR is B + T + O + D, where D is the deferred-tax adjustment, conventionally D ≤ 0.
Article 207(3) says a reduction of a DTL or increase of a DTA generates a negative adjustment, while Article 207(4) sets an otherwise positive adjustment to zero. The capital formula thus recognizes qualifying tax mitigation, subject to the recovery constraints, without rewarding an adverse tax consequence by permitting a negative reduction to be reversed into a separate tax capital charge through the adjustment. A tax effect unrelated to the legally prescribed loss cannot be simply appended to D.
The 2019 amendment to Article 207 inserted a more explicit evidentiary discipline into the regulation. Paragraph 2a requires proof satisfactory to the supervisory authority that future taxable profit is probable for an increase in DTA resulting from the loss; it explicitly addresses carryforward limits, the scale and consequences of the stress, and increasing forecast uncertainty. Paragraphs 2b and 2c impose consistency constraints and restrictions on assumptions for new business, investment returns, and projection horizons. This amendment is central to the assessment of older commentary because the earlier text did not express all these supervisory tests in the same level of detail.
Interaction with discretionary benefits
Technical-provisions loss absorption and deferred-tax loss absorption arise from different balance-sheet channels. For participating life business, expected future discretionary benefits may be reduced under adverse conditions, subject to contractual, legal, and managerial limits. The technical-provisions adjustment rule, Article 206, expresses the adjustment through a comparison between BSCR and net BSCR, limited by future discretionary benefits. Tax relief is evaluated after this adjustment for the standard formula. EIOPA's reporting clarification on the ordering confirms that LAC TP precedes LAC DT and describes LAC TP as pretax in this structure.
This sequencing has economic consequences. If discretionary policyholder benefits absorb part of a market shock, the remaining loss borne by shareholder capital decreases. The relevant tax loss can therefore be smaller than BSCR alone. Applying the tax rate to the gross BSCR while also deducting full LAC TP would assert deductions on a loss already transferred to policyholders. Moreover, the effect of a reduction in future discretionary benefits on insurer tax profits is not identical for every contract or tax jurisdiction. Taxable profit and the timing of deductibility must be reconstructed rather than assumed.
An insurer without material future discretionary benefits may have no corresponding LAC TP, but the DTA recognition constraint remains. Conversely, a life insurer with substantial participating benefits may show a significant LAC TP and little LAC DT if future tax capacity after the event is weak. This combination can be entirely coherent: the two mechanisms are controlled by different legal and economic resources.
The instantaneous-loss reconstruction
A robust computational sequence starts with the prudential valuation date and an explicit definition of the prescribed loss. It then maps the loss into risk drivers and changes in asset and liability values, including management actions allowed by the framework. Next it calculates resulting tax bases and differences, determines taxes payable or refundable, evaluates DTL reversals and additional DTA, and finally applies the recoverability test to those incremental positions. The final adjustment is recorded together with its risk allocation, supporting evidence, and assumptions.
Where allocation to causes is necessary, Article 207(5) requires consistency with contributions of standard-formula modules and submodules to BSCR, and it supplies a rule for partial internal models outside the model scope. This prevents a taxpayer from assigning an aggregate loss entirely to the risk driver generating the most convenient tax deduction if that would contradict the regulatory structure. The allocation should retain tax-relevant characteristics, including whether a loss changes the tax base, qualifies for deduction, enters ordinary versus capital tax categories, or has timing differences. Purely financial allocation and legally correct tax allocation must be reconciled.
Tax asymmetries and double-counting controls
Tax law frequently treats income and losses asymmetrically. A taxable gain may increase tax liabilities immediately, while a comparable economic loss creates a deduction usable only after future profits arise. Limitations on interest deductions, capital-loss offsets, cross-border loss relief, and group consolidation further complicate equivalence. These restrictions matter disproportionately in catastrophe and financial-crisis scenarios precisely because losses may become widespread across subsidiaries and market participants. A company can experience a real economic loss while the resulting tax deduction is legally deferred, ring-fenced, or unusable.
Three particular double-counting hazards deserve scrutiny. First, a DTA already recognized in own funds cannot automatically be counted again as a newly created stressed DTA. Second, the same future taxable profit cannot support several independent claims on a limited tax capacity, including recognition of a starting DTA, a new stressed DTA, and other tax attributes, without a combined utilization schedule. Third, a future management action or reinsurance recovery cannot both restore earnings in the profit projection and be added as a separate tax capacity if the calculation already reflects it. The analyst should reconcile a single tax ledger across base and shocked states and carry forward all constraints.
Assumptions on Future Taxable Profit
Legal utilization, expiration, and reversal schedules
The future taxable-profit test begins with an inventory of legal tax attributes and an entity-by-entity schedule. For each deduction or credit it should identify the taxpayer, jurisdiction, tax class, expiry, eligible income type, timing of expected reversal, restrictions on group relief, and availability of carryback. A tax benefit that depends on a foreign subsidiary generating profit cannot be asserted at the parent level unless the applicable law and agreements permit its use. A projected DTL reversal can support utilization only when the relevant reversal is taxable in the appropriate entity, within the relevant period, and not itself dependent on the stressed deduction being counted twice.
An illustrative timing case makes the point. Assume a new deductible tax loss of 120, taxed at 25%, with potential DTA of 30. If the law permits utilization only against ordinary taxable profit in the next three years, projected taxable amounts of 20, 30, and 10, net of existing deductions and other claims, support no more than 60 of utilization, giving a DTA benefit of 15. An additional expected profit of 100 in year six cannot rescue an expiring loss. If the jurisdiction instead permits indefinite carryforward subject to a percentage annual taxable-income ceiling, a different schedule follows. The model must implement actual law, not a universal three-year rule; these figures are purely constructed.
The legal proof should be traceable to enacted or otherwise applicable rules as of the valuation date, including changes in relevant rates. A tax rate is not necessarily one scalar applicable to every transaction or across the full projection horizon. A change in taxable income category can affect both rate and availability of offsets. The IAS 12 accounting framework provides a basis for understanding recognized DTAs and DTLs, while the Solvency II prudential recognition test adds the post-shock condition.
Post-stress business projections
Forecasts supporting an incremental stressed DTA must start from the business that remains after the event. Insurer management plans prepared for normal conditions may overstate premium growth, investment income, persistency, distribution capacity, or access to reinsurance after a major loss. Article 207(2a) expressly directs attention to business pricing, demand, market profitability, reinsurance coverage, and macroeconomic conditions. A plausible profit projection should be causally connected to the particular shock, rather than copied from an unstressed budget with a single year of loss inserted.
Article 207(2b) prohibits assumptions more favorable than those underlying ordinary DTA valuation under Article 15. The Article 207 projection constraints include a bar on assuming new sales beyond business plans and beyond a five-year maximum; forward-rate-derived post-shock investment returns unless stronger evidence supports higher yields; and a finite horizon with increasing haircuts where relevant new-business projections exceed the business-planning horizon. These are carefully delimited rules, not permission to assume five years of uninterrupted positive profits. Other, already-in-force business and legally permissible reversals still require their own reasonable treatment.
Profits after stress should be measured on a taxable basis, not inferred from total economic cash generation or gross premiums. Reinsurance payments may improve liquidity while affecting taxable profit at a different date. The release of insurance reserves may generate accounting profits whose tax recognition differs. Profits may be consumed by carried losses created before the modeled catastrophe. Dividend restrictions and policyholder participation can alter economic cash flows, while tax law determines which corresponding items affect taxable earnings. A balance-sheet projection cannot replace a taxpayer-specific tax forecast.
Financial-market and actuarial assumptions
Tax recovery projections interact with actuarial assumptions in ways that can reinforce the original adverse event. A long-tail non-life insurer may experience claims inflation, adverse reserve development, and investment losses together. A life insurer may face declining reinvestment yields, guarantees, policyholder behavior, and changing mortality or morbidity trends. Health insurers may encounter persistently elevated claims costs after a systemic shock. Assuming rapid return to historical profitability without modeling those mechanisms exaggerates DTA recoverability in the exact states for which solvency capital is intended.
Investment income is especially sensitive because an asset-market shock changes both the current valuation and forward investment opportunity set. Under Article 207(2c), a post-loss return exceeding forward-rate-implied risk-free returns requires credible evidence. A portfolio's historical average yield does not itself establish a post-stress achievable yield. Projected risk premiums must also be consistent with the risks and capital consequences of the assumed investment strategy. A credible strategic response might require asset sales, liquidity retention, or reinsurance restructuring; all can alter expected taxable profits.
Demographic, lapse, and catastrophe projections should use consistent scenario definitions. If a capital model posits a mass lapse event, the taxation forecast cannot retain the pre-stress policyholder base without explanation. If a catastrophe model increases retained claims due to exhausted reinsurance limits, the projected underwriting margin should reflect those losses and potentially more expensive renewal coverage. A profit model that independently resets every adverse driver to its pre-stress mean before the first forecast year will frequently understate persistence and correlation.
Group transfers and supervisory evidence
Tax consolidation, fiscal unity, and contractual transfers can change the location of tax-utilization capacity. Recognition nevertheless depends on the actual legal arrangement and on the economic ability of the transferee or transferor to use the claim. EIOPA's Guideline 9 interpretation expressly rejects the reading that an insurer with no demonstrable current tax liabilities or future taxable profit can obtain LAC DT merely because another group entity pays for transferred tax losses. The stated limit is connected to what the transferring undertaking could use itself under the guideline's treatment.
This point protects against an attractive but unstable form of group-capital creation. A contract promising to buy a tax loss may be valuable in some legal arrangements, but tax relief recognized under the particular Solvency II guidance cannot be inferred from the contract alone. The group must consider legal enforceability, contingent capacity, counterparty exposure, applicable caps, tax authority treatment, transfer pricing, and whether the projected income remains available after the simultaneous group stress. The importance of group structure also demonstrates why a high consolidated profit cannot establish taxable-profit capacity in every legal entity.
The documentary record should include signed tax-group agreements, entity-level forecasts, audited or reconciled tax bases, board-approved business plans, tax-law memoranda, detailed stress allocations, and an explicit schedule of recognized and unrecognized losses. Auditors and supervisors may evaluate reasonableness differently, but the essential scientific criterion is replicability: a reviewer should be able to follow each numerical benefit from legal eligibility through forecast utilization to the reported capital figure.
Regulatory Capital Adjustment
Reproducible hypothetical capital case
Consider an invented insurer reporting in monetary units, with BSCR of 180, technical-provisions adjustment of -30, and operational-risk capital of 20. Its pre-DTA-adjustment instantaneous loss is L = 180 - 30 + 20 = 170. Assume a single relevant tax jurisdiction with a 25% rate, fully deductible loss components, and no separate capital-loss categories, so the maximum arithmetic tax effect before recoverability limits is 170 × 0.25 = 42.5. This multiplication is an upper bound within the example's tax assumptions, rather than evidence of an allowed deduction.
Suppose a properly documented DTL reversal associated with the stressed balance sheet yields 12.5 of tax relief, equivalent to 50 of relevant taxable temporary difference at 25%. The residual maximum potential relief is 30, or 120 of deductions. Assume that approved taxpayer-level forecasts and available offset capacity show that all 120 can probably be utilized within legal deadlines after the shock. In that deliberately favorable case, LAC DT is -42.5 and SCR = 180 - 30 + 20 - 42.5 = 127.5. The example treats the 12.5 DTL reversal and the 30 new DTA as nonoverlapping sources; it does not assert this allocation for real firms or infer that an accounting balance necessarily has an identical capital effect.
Now suppose the same hypothetical insurer has eligible own funds of 190, held constant across the scenario comparisons solely to isolate the denominator's effect. The pre-tax-adjustment coverage is 190 / 170 = 111.76%. If the full tax effect satisfies the recognition test, coverage is 190 / 127.5 = 149.02%. The difference of 37.25 percentage points reflects the constructed recognized adjustment and ratio arithmetic, not an increase in cash reserves, a fresh capital injection, or a change in economic loss itself. In practice, current-balance-sheet deferred tax recognition also affects eligible own funds, and a complete comparison must recompute numerator and denominator consistently.
Sensitivities and capital-ratio interpretation
Retain the constructed DTL relief of 12.5 but change the amount of supportable newly created DTA. With an available taxable-profit base supporting only 40 of deductions at 25%, the incremental DTA support is 10, total recognized tax relief becomes 22.5, and the SCR is 147.5. The fixed-numerator ratio is 190 / 147.5 = 128.81%. With no supportable new DTA, relief is 12.5, SCR is 157.5, and the ratio is 190 / 157.5 = 120.63%. If the assumed DTL reversal itself does not survive inspection of taxable temporary differences, and no DTA is recoverable, the example returns to SCR 170 and ratio 111.76%.
| Hypothetical utilization case | DTL relief | New DTA support | LAC DT | SCR | Coverage, own funds fixed at 190 |
|---|---|---|---|---|---|
| Full support | 12.5 | 30.0 | -42.5 | 127.5 | 149.02% |
| Restricted taxable profit | 12.5 | 10.0 | -22.5 | 147.5 | 128.81% |
| No new DTA recoverability | 12.5 | 0.0 | -12.5 | 157.5 | 120.63% |
| No recognized tax absorption | 0.0 | 0.0 | 0.0 | 170.0 | 111.76% |
The ratio moves nonlinearly because the relief changes the denominator. If C = F/(L - A), where F is own funds, L is the capital requirement before tax relief, and A is the accepted positive amount of relief, then ∂C/∂A = F/(L-A)^2, provided L > A. This derivative means a unit of added recognized tax relief has a larger proportional effect when the denominator is already small. Capital-ratio comparisons that show only the final percentage can conceal material variation in stressed tax evidence. An analyst should therefore report recognized relief in currency, relative to the pre-tax capital amount, and alongside both the initial and adjusted solvency ratios.
A second sensitivity concerns the taxable-profit utilization horizon. The illustrative 120 of prospective deductions might be supported by pre-stress business forecasts but fall to 40 when policyholder lapses, higher reinsurance costs, and new-business restrictions are incorporated. The result is not a correction to the tax rate: it is a correction to the amount of usable taxable capacity. A third concerns partial deductibility. If only 80 of the 170 shock is tax deductible under relevant law, the maximum nominal effect becomes 20 at 25%, before considering any DTL reversals already captured and any allocation of those effects. Such sensitivities should be assessed separately to avoid changing several assumptions without traceability.
European Union, United Kingdom, and IAIS comparison
The EU Solvency II approach computes the adjustment from a prescribed instantaneous loss and applies detailed evidence requirements for newly created stressed DTAs. The EIOPA LAC guidelines address consistent calculation and supervisory implementation, while Regulation 2015/35 remains central to the binding computation. EIOPA's guidance and answers should be read alongside the actual legal text and local implementation rather than treated as replacements for it.
The United Kingdom maintains a separately governed Solvency UK regime. The Prudential Regulation Authority's current recognition-of-deferred-tax statement, SS2/14, is published in its November 2024 version, effective December 31, 2024. It emphasizes the credibility of profit projections for both balance-sheet DTA recognition and a stressed tax effect. The PRA's loss-absorption policy index separately identifies permissions and modifications by consent related to LAC DT. This demonstrates that a reader cannot apply EU Article 207 automatically to every present-day UK insurer: the current PRA Rulebook and granted permissions determine the specific UK treatment.
The IAIS adopted its ICS for internationally active insurance groups in 2024. According to the official Insurance Capital Standard, ICS is a global group-capital reference, implemented through supervisors rather than a universal corporate tax statute. In its ICS tax provisions, paragraphs L1-144–149 and L2-339–353, the group-level effective tax rate derives from specified historical insurance-related earnings, and the standard recognizes tax effects from both ICS balance-sheet adjustments and the capital requirement. The default usable tax effect on the ICS capital requirement under L2-348 is 80% of the notional tax effect. When required by the group-wide supervisor, L2-349 instead uses a capacity-sensitive limit accounting for carrybacks, post-stress taxable income, and net DTA/DTL positions. These are different methods from Solvency II's case-specific Article 207 DTA test; they share an economic subject while differing in regulatory object, aggregation unit, and operative formula.
The ICS includes a detailed illustrative group example that allocates a notional tax effect among entities and derives usable capacity from carryback and post-stress income components. Its importance is methodological: a consolidated headline tax rate does not eliminate differences in legal utilization. The ICS definition of a group and the tax subgroup or entity allocation also affect recognition. Using the ICS's default 80% multiplication as though it were the EU Solvency II methodology, or treating an EU LAC DT as already an ICS-compliant amount, would manufacture comparability that the texts do not establish.
International measurement boundaries
A cross-jurisdictional solvency comparison must specify valuation basis, risk horizon, confidence level or prescribed stress, legal entity versus group scope, capital numerator, capital denominator, loss-absorption ordering, and tax accounting basis. Even where two jurisdictions use market-consistent balance sheets, one may restrict recognized tax relief through standardized formulas and another through institution-specific supervision. The absence of an analogous adjustment is also meaningful; comparisons should state whether the metric is pretax or post-tax instead of interpreting all reported coverage ratios as the same quantitative object.
A prudential result is not globally transferable merely because the insurer and its corporate group operate internationally. The tax authority and taxable-income unit may differ from the supervisory group. Ring-fenced businesses, funds with restricted transferability, branches, and material non-insurance subsidiaries may prevent free migration of tax benefits. The scientific object remains the verified tax effect available to the relevant regulated capital calculation, not a theoretically realizable benefit somewhere in the consolidated organization.
For legal version control, the EU Solvency II amending Directive 2025/2 entered into force in January 2025, but Member States must transpose by January 29, 2027 and apply the pertinent transposing measures from January 30, 2027. This monograph therefore distinguishes the regulatory position researched as of November 14, 2026 from reforms scheduled for later application; it does not mechanically backdate the future application of an amendment to current solvency filings. Country-by-country tax rules and future changes in delegated regulations require fresh checks for subsequent valuation dates.
Evidence and Limits
Source hierarchy and empirical interpretation
The primary legal foundation comprises the Solvency II Directive and Delegated Regulation as amended, supplemented by official EIOPA rulebook pages and the relevant European amending act. The IFRS Foundation supplies the accounting-standard overview; EIOPA guidelines and final Q&A supply official supervisory interpretations; the Bank of England publishes current PRA guidance and policy navigation; the IAIS publishes the ICS. These materials are directly authoritative for the particular propositions attributed to them. A regulatory implementation webpage or Q&A is not the same kind of source as an enacted legal instrument, and the monograph distinguishes their status.
The European Insurance Overview 2024 includes charts of LAC DT impact by country. This is evidence that supervisors consider the measure material and internationally variable. The chart presentation alone does not identify a causal effect of tax policy, isolate differences in business mix, or demonstrate the reliability of every insurer's forecast. Cross-country distributions require attention to reporting definitions, sample selection, date, and the fact that the reported tax adjustment is already shaped by supervisory recognition. The numerical cases in this work are constructed, not calibrated from those charts.
A historical supervisory source can still be valuable when its version is respected. The 2019 regulation changed Article 207's explicit constraints; the PRA updated SS2/14 in November 2024; the ICS was adopted in 2024. Quotations or formulas lifted from earlier periods without dates may misdescribe current rules. The absence of primary local tax legislation for each jurisdiction in this dossier means that the examples deliberately use abstract tax conditions. A particular insurer requires a current tax-law opinion and actual legal-entity data before the model can support an operational capital filing.
Independent research and supervisory assessments
The academic and supervisory literature provides independent reasons to regard deferred-tax adjustments as an object of empirical scrutiny. The research summarized in Loisel and colleagues’ stress analysis studies how recalibration of tail risk and loss-absorbing capacities affects post-shock solvency ratios, drawing on supervisory data. It demonstrates the general importance of re-estimating capital needs after an event; it does not establish a universal recovery ratio for a specific insurer's tax loss. That distinction strengthens the methodological position of this dossier: the pre-event recognized adjustment and post-event realizable relief require separate evidence.
The IMF's 2025 euro-area insurance assessment observes that recognition of LAC DT allows significant interpretation and that material cross-country variation warrants heightened supervisory attention. This observation supports a particular supervisory research question: do differences in reported tax loss absorption reflect genuine differences in legal utilization and business economics, or different degrees of permitted projection optimism? Public aggregates alone cannot identify the answer. Examination of tax-return histories, model assumptions, and supervisory adjustments would be required.
In the United States, the working paper Regulatory Capital and Deferred Tax Assets by Eastman, Ehinger, and Meegan (2026 revision) reports associations between admitted DTAs and measures of insurer risk and capital issuance. The available publisher-hosted abstract is the level of evidence examined here. The claims are associations in the reported research design, not identified universal causal effects of deferred taxation. Its focus on U.S. admitted DTAs also differs conceptually from Europe's separate LAC DT adjustment. It nevertheless highlights why balance-sheet admissibility and solvency outcomes should be studied jointly, with attention to selection, regulatory changes, and risk taking.
The Netherlands' supervisor discusses its domestic review in DNB's deferred-tax good practices. This source explicitly places DTA, DTL, LAC DT, risk management, and capital management within a common supervisory review. Its institutional significance is practical rather than a substitute for the EU regulation: evidentiary quality emerges from reconciling tax, capital, and governance information within the regulated undertaking.
Model risk, governance, and audit trail
The most material model risks arise from scenario inconsistency, tax-base classification, forecast optimism, undocumented legal transfers, estimation of reversals, and feedback between recovery actions and solvency. A forecasting engine can be internally consistent yet institutionally invalid if it assumes a tax offset barred by law. Conversely, a legally eligible carryforward may have little prudential value if the stress removes the profitable business needed to exploit it. Independent scrutiny must therefore address both mathematical mechanics and premise validity.
At minimum, the audit trail should record the reporting-date legal texts, tax rates, tax bases, recognized DTA/DTL by taxpayer, planned reversals, mapping from capital risk to taxable event, post-stress balance-sheet changes, forecast taxable profits, carryback/carryforward utilization, management actions, and specific supervisory approvals where relevant. A reviewer should be able to recompute the tax adjustment from these records without relying on a narrative estimate. Source references should match exact assertions and temporal scope, including the distinction between an EU directive, delegated rule, EIOPA interpretive answer, PRA guidance, and ICS group specification.
Governance also affects projection independence. Front-office or commercial management may be incentivized to retain optimistic premium growth after stress; tax staff may focus on legal deductibility without modeling insurer capital; actuarial teams may project aggregate earnings without tax class and taxpayer detail. A controlled process reconciles the tax, finance, actuarial, risk, and capital models and escalates contradictory assumptions. Board oversight and supervisory challenge are particularly important where recognized tax relief makes the difference between a breached and a compliant solvency ratio.
Data limitations and falsification tests
A useful scientific model specifies evidence capable of falsifying its favorable result. A DTL reversal claim fails if the assumed temporary difference remains unchanged after stress or relates to a legally exempt gain. A stressed DTA forecast fails if tax losses expire before sufficient qualifying taxable profit exists. A group claim fails if the relevant transfer cannot be executed under tax law or the supported recipient profits are already committed. A projection fails if its investment yields or new-business assumptions violate the applicable prudential constraint. A tax-effect computation fails if the stress allocation differs materially from the regulatory risk contributions without reconciliation.
Sensitivity analysis should vary taxable profits, horizon, investment return, underwriting margin, claims inflation, tax rate, loss characterization, DTL reversal rate, and legal group offsets independently where feasible. The question is how much recognized relief survives each plausible adverse variation, with the combination of variations used to test dependence and concentration. Statistical intervals are meaningful only where grounded in data and modeling assumptions. This work does not provide an empirical confidence distribution, a taxpayer-specific legal opinion, or evidence for the actual recoverability of any named undertaking's DTA.
Accounting-Prudential Reconciliation
Three reconciliations across reporting layers
The integrated analysis can be organized into three reconciliations using established accounting and prudential concepts. First is the valuation reconciliation: accounting carrying amounts, tax bases, Solvency II values, and recognized deferred taxes at the reporting date. Second is the shock reconciliation: movement from the prescribed before-stress regulatory capital loss to the corresponding after-stress asset, liability, and tax positions, including proper sequencing of LAC TP and LAC DT. Third is the utilization reconciliation: legally usable offsets, taxable-profit projections, existing DTA competition, DTL reversals, and the increment recognized as capital relief. These are transparent analytical stages of the existing regulatory problem, not a new named doctrine or capital model.
The reconciliations are complementary. The first protects the numerator of a solvency ratio against misvaluation. The second protects the denominator against inconsistent stress assumptions and double counting. The third establishes whether a nominal tax effect has the economic prospect of utilization required by the governing rule. The distinction makes sense of cases where balance-sheet DTA falls while the ability to absorb a new loss also falls; where a DTL reverses but no future earnings arise; and where a firm has strong group earnings yet little usable capacity in the legally relevant taxpayer.
Consider a firm with a strong balance sheet but concentration in one tax jurisdiction. Its aggregate insurance losses could be diversified across geography, while taxable profit is concentrated in a different subsidiary. The regulatory gross capital model may recognize insurance diversification, but the tax system may prohibit the corresponding transfer of deductions. The proper result combines favorable capital diversification with restricted LAC DT. The reverse is also possible: a firm with stable taxable income and enforceable carrybacks may support some loss absorption despite comparatively volatile underwriting earnings. Scientific comparison must preserve this institutional heterogeneity rather than normalize it away.
Scientific synthesis and practical implications
The central analytical result is a conditional relationship between tax law, stress-consistent expected income, and allowable prudential credit. The arithmetic tax shield generated by an economic loss is an upper bound defined by assumed tax deductibility and rates. Recognized loss-absorbing capacity is the portion supported by actual temporary-difference changes and satisfactory evidence of utilization under the chosen framework. The difference between the two is material economic information: it measures limitations of the tax system, remaining profitability, timing, and supervisory recognition in severe states.
This relationship has five direct implications. Capital planning should report tax relief transparently next to the unadjusted capital measure; stress testing should link future profits to the same shock generating the loss; tax and actuarial forecasting should share entity-level exposure and assumption registers; group reporting should preserve the legal taxpayer boundaries within consolidated analysis; and international comparisons should identify when capital ratios contain fundamentally different tax adjustments. In each case, the objective is a traceable capital measure rather than the largest arithmetically possible adjustment.
A research program examining actual firms could compare reported LAC DT in currency and as a fraction of pre-adjustment capital, documenting tax jurisdictions, life/non-life mix, legal carryforward restrictions, starting DTL, stressed profit assumptions, and supervisory review outcomes. That would require high-quality nonpublic tax schedules and careful treatment of selection bias. Available public solvency ratios and country-level charts provide useful descriptions but cannot independently validate the private recovery forecasts underlying each recognized DTA. Accordingly, this monograph advances a documentary and mathematical synthesis; it makes no claim of an empirically discovered causal effect or new original tax-capital theory.
Conclusion
Deferred taxes affect insurance solvency through an already recognized balance-sheet position and through the potential adjustment for loss-absorbing capacity under a prescribed adverse event. Both effects demand consistent valuations and legal tax characterization. Under Solvency II, Article 207 measures the tax consequence of an instantaneous loss built from BSCR, technical-provisions absorption, and operational risk, and strictly conditions the use of an increased stressed DTA on probable future taxable profit and its supporting assumptions. The 2019 regulatory amendments strengthened the specificity of that evidence; EIOPA interpretation, current PRA guidance, and the IAIS ICS establish related but institutionally distinct approaches.
The worked example demonstrates that an insurer with a nominal 42.5 tax benefit can have a recognized adjustment of 42.5, 22.5, 12.5, or zero depending on valid DTL reversals and supported future taxable income. A capital ratio derived from the accepted adjustment can change markedly even while the illustrative starting own-funds amount remains fixed. The calculation clarifies why an observed improvement in solvency coverage should be accompanied by disclosure of the recognized tax benefit and its basis. A complete practical calculation must additionally recompute the capital numerator and denominator under mutually consistent recognition decisions.
The decisive scientific conclusion is that credible tax loss absorption is determined by the intersection of a legally recognizable tax effect, an economically coherent post-stress state, and an auditable utilization pathway. Each can be tested. Where the evidence supports all three, the prudential adjustment conveys a genuine conditional mitigation of losses. Where the assumptions fail, the nominal tax saving remains an arithmetic possibility rather than a recognized capital resource. Preserving that distinction improves the interpretation of solvency ratios, the quality of supervisory scrutiny, and the comparative study of insurance capital systems.
References
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Directive 2009/138/EC, Article 101, Solvency Capital Requirement. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2188_en
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Directive 2009/138/EC, Article 108, Loss-absorbing capacity of technical provisions and deferred taxes. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2194_en
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Delegated Regulation (EU) 2015/35, Article 15, Deferred taxes. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2428_en
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Delegated Regulation (EU) 2015/35, Article 205, General provisions. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-5858_en
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Delegated Regulation (EU) 2015/35, Article 206, Adjustment for loss-absorbing capacity of technical provisions. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-5859_en
European Insurance and Occupational Pensions Authority (EIOPA). (n.d., accessed November 14, 2026). Solvency II Single Rulebook: Delegated Regulation (EU) 2015/35, Article 207, Adjustment for loss-absorbing capacity of deferred taxes. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-5860_en
European Insurance and Occupational Pensions Authority (EIOPA). (2015, guidance edition available in 2023). Guidelines on the loss-absorbing capacity of technical provisions and deferred taxes, EIOPA-BoS-14/177. https://www.eiopa.europa.eu/publications/guidelines-loss-absorbing-capacity-technical-provisions-and-deferred-taxes_en
European Insurance and Occupational Pensions Authority (EIOPA). (2015–2026 access). Q&A 364: Interpretation of Guideline 9 and tax-loss transfer arrangements. https://www.eiopa.europa.eu/qa-regulation/questions-and-answers-database/364_en
European Insurance and Occupational Pensions Authority (EIOPA). (2023–2026 access). Q&A 2565: Reporting LAC TP before tax and sequencing with LAC DT. https://www.eiopa.europa.eu/qa-regulation/questions-and-answers-database/2565_en
European Insurance and Occupational Pensions Authority (EIOPA). (2024). European Insurance Overview 2024, including figures on LAC DT impact by country. https://www.eiopa.europa.eu/assets/insurance-overview-report/European_Insurance_Overview_2024.html
European Union. (2019). Commission Delegated Regulation (EU) 2019/981 of March 8, 2019, amending Delegated Regulation (EU) 2015/35. Official Journal of the European Union. https://eur-lex.europa.eu/eli/reg_del/2019/981/oj/eng
European Union. (2025). Directive (EU) 2025/2 of November 27, 2024, amending Directive 2009/138/EC. Official Journal of the European Union, 2025/2. https://eur-lex.europa.eu/legal-content/en/ALL/?uri=CELEX%3A32025L0002
IFRS Foundation. (n.d., accessed November 14, 2026). IAS 12 Income Taxes: Issued Standards overview and standard history. https://www.ifrs.org/issued-standards/list-of-standards/ias-12-income-taxes/
Bank of England, Prudential Regulation Authority. (2024). Supervisory Statement SS2/14: Solvency II — Recognition of Deferred Tax, November 2024 version, effective December 31, 2024. https://www.bankofengland.co.uk/prudential-regulation/publication/2014/solvency2-recognition-of-deferred-tax-ss
Bank of England, Prudential Regulation Authority. (2024–2026 access). Operational Risk and Loss Absorbing Capacity: Policy and supervisory resource index. https://www.bankofengland.co.uk/prudential-regulation/prudential-and-resolution-policy-index/insurance/operational-risk-and-loss-absorbing-capacity
International Association of Insurance Supervisors (IAIS). (2024). Insurance Capital Standard: Adoption, scope, and implementation background. https://www.iais.org/activities-topics/standard-setting/insurance-capital-standard/
International Association of Insurance Supervisors (IAIS). (2024; accessed November 14, 2026). Insurance Capital Standard Online Tool, Level 1 and Level 2 texts, Section 6 (Tax), notably L1-144–149 and L2-339–353. https://www.iais.org/ics-online-tool/
European Insurance and Occupational Pensions Authority (EIOPA). (2017). Re-evaluation of the Capital Charge in Insurance After a Large Shock: Empirical and Theoretical Views, thematic research article, Financial Stability Report, June 2017. https://www.eiopa.europa.eu/publications/re-evaluation-capital-charge-insurance-after-large-shock_en
International Monetary Fund. (2025). Euro Area: Financial Sector Assessment Program — Technical Note on Insurance Micro- and Macroprudential Supervision, IMF Country Report, discussion of LACDT. https://www.elibrary.imf.org/view/journals/002/2025/208/article-A001-en.xml
Eastman, E., Ehinger, A., & Meegan, C. (2026, October 31; revised November 3). Regulatory Capital and Deferred Tax Assets, working paper, SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3461346
De Nederlandsche Bank (DNB). (2020, December 8; accessed November 14, 2026). Q&A and Good Practices on the Role of Deferred Taxes (DTA, DTL and LAC DT) in Solvency II. https://www.dnb.nl/en/sector-information/open-book-supervision/open-book-supervision-sectors/insurers/balance-and-capital-pillar-1/qa-and-good-practices-on-the-role-of-deferred-taxes-dta-dtl-and-lac-dt-in-solvency-ii/