Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Equity and Alternative Investment Risk for Insurers

Abstract. Insurers hold equity and alternative investments to obtain total return, diversify sources of cash flow, gain exposure to real assets, and finance liabilities with differing horizons. These allocations also introduce market losses, contingent funding commitments, opaque valuation, governance conflicts, and the risk of forced transactions during insurance-specific liquidity stress. This monograph investigates listed equity, direct real estate, infrastructure equity, private-equity partnerships, and other primarily noncredit alternative exposures in insurance general accounts. It distinguishes economic exposure from fund form, investment management arrangements from ownership of insurers, observable prices from modeled asset values, and expected long-run return from cash available when obligations fall due. The inquiry combines primary international supervisory standards, selected European and U.S. regulatory rules, 2025–2026 supervisory data, peer-reviewed financial research, and reproducible illustrative asset–liability calculations. Its central analytical result is that the suitability of an alternative investment is jointly determined by its underlying cash flows, liquidity under stress, valuation reliability, risk capital, correlated liability demands, and the insurer's ability to control and understand the full exposure. Apparent diversification based on stale returns is an unreliable substitute for economic stress analysis. Long-term liabilities can support investment patience, but do not by themselves ensure short-term cash availability. The regulatory comparison differentiates EU Solvency II investment and capital principles, U.S. statutory portfolio reporting, and international IAIS supervision while preserving jurisdiction-specific differences. The paper establishes practical interpretive conclusions within existing finance and actuarial scholarship, without proposing a new investment theory or claiming novel empirical discovery.

Keywords: Insurance investment risk; listed equity; private equity; real assets; infrastructure; real estate; alternative investments; asset-liability management; liquidity risk; valuation smoothing; Solvency II; prudent person principle; capital adequacy; insurance general accounts; stress testing; look-through; governance.

Introduction

Object, decision setting, and scope

The investment function of an insurer exists within an institution that promises cash payments contingent on insured events and contractual rights. An insurer can own assets whose ultimate value is substantial and still encounter difficulty meeting an immediately payable claim, collateral call, surrender, or margin settlement. Conversely, assets whose prices move substantially may be supportable when cash commitments, guarantees, capital and liquidity buffers are structured to absorb that movement. The relevant unit of analysis is therefore the asset–liability balance sheet through time, including commitments and contingent demands beyond the reported portfolio snapshot.

This Dossier examines five overlapping but analytically distinguishable forms: publicly traded common equity; holdings in operating businesses through unlisted shares and funds; commercial and other income-producing real estate; infrastructure-related equity, including concession and regulated-business ownership; and residual alternative investment structures such as hedge funds and private partnership interests. These forms include direct positions, equity fund units, joint ventures and look-through exposures in affiliates. The subject is the insurer's exposure to equity-like and primarily noncredit risks. Private credit, loan default, bond spreads and credit-specific securitization belong to adjacent research, but their interactions are relevant when the same sponsor, sector, financing facility or insurance group creates correlated risk.

A methodological distinction runs throughout. An insurer holding a private-equity fund is an investor in a private asset. An insurer owned by a private-equity firm is a corporate entity with a particular control and business structure. An insurer using a private-equity-affiliated manager is in an agency relationship. These roles can coincide; none establishes the size or performance of the others. The NAIC's investment analysis of PE-owned insurers refers to insurers identified by ownership, whereas the asset categories in the industry-wide 2025 investment report refer to holdings. Conflating the two produces invalid market statistics and misleading capital comparisons.

The international reference point is the IAIS Insurance Core Principles, especially ICP 15 on investment and ICP 16 on enterprise risk management. The EU's prudent person principle requires insurers to understand and manage risks, and to consider security, quality, liquidity, and profitability of the portfolio and liability characteristics. The principle concerns the whole institution rather than a nominal security label. National investment eligibility, supervision, accounting and capital formulae still determine legal implementation in each jurisdiction.

Research questions and evidence method

The study asks which risk sources equity and real assets contribute; when apparent diversification survives realistic liquidity and valuation treatment; how fund cash-flow measurement changes risk inference; what can be learned from recent supervisor-reported portfolio data; how capital, accounting and liability valuation alter the decision; and what governance tests make the conclusions operationally meaningful. It applies documentary analysis of law and standards, financial-economic interpretation of published research, comparison of disclosure denominators, balance-sheet identities and explicitly hypothetical stress arithmetic. Source dates refer to publication or data vintage, not inferred current positions. Regulatory rules scheduled to apply in January 2027 are not represented as effective in November 2026.

The source evidence is heterogeneous. The NAIC 2025 asset-mix study reports U.S. statutory book/adjusted carrying values and includes affiliated and unaffiliated assets. The EIOPA October 2026 risk dashboard reports European distribution indicators with defined exclusions. Peer-reviewed fund-performance studies use proprietary or selected investment samples. These evidence populations are compared for institutional meaning, not pooled into a supposedly homogeneous global allocation statistic. All numeric portfolios constructed below are pedagogical counterfactuals, fully identified as such; they are not measured insurer outcomes.

The economic logic is developed in successive stages: exposure and return formation; underlying real-asset cash flows; private-fund measurement; accounting, valuation and exit capacity; asset–liability stress interaction; and decision governance. Related areas of the approved science architecture are Insurance Finance and Asset-Liability Management, Solvency and Capital, Enterprise Risk Management in Insurance, and Annuities and Retirement Insurance. This relationship records shared scholarly scope without multiplying the Dossier's single canonical identity.

Equity Asset Returns — Concepts, History, and Scientific Object

Equity claims and their distinct distribution of outcomes

Common equity is a residual claim. Dividends and liquidation proceeds depend on contractual priorities, operating performance, financing costs, taxation and governance decisions. A simple one-period shareholder return can be written as

Rt+1=Pt+1−Pt+Dt+1Pt,R_{t+1}=\frac{P_{t+1}-P_t+D_{t+1}}{P_t},

where prices must be measured on a consistent basis and dividends or other distributions treated appropriately. A total-return index and a dividend-yield series are distinct data products; neither is a fixed schedule of payments equivalent to a bond coupon or a matched insurance liability. Corporate leverage magnifies the variability borne by the residual claimant. The relationship of firm value to equity value matters especially in infrastructure projects and leveraged buyouts, where modest asset-value changes can imply large equity losses.

Stocks offer a share of economic growth and, in some cases, longer-run sensitivity to inflation through nominal revenues and replacement costs. These possible characteristics depend on pricing power, margins, regulation, production costs and valuation. An inflation increase can simultaneously raise certain revenues and the discount rates applied to those revenues, leaving share values lower. Treating equity as a mechanical inflation hedge is therefore unsound. A genuine hedge must be evaluated against the specific insurer liability cash flows, including inflation-linked claims or guaranteed policy payments.

The financial literature on portfolio diversification establishes that an asset's standalone return distribution does not determine its contribution to portfolio risk. The covariance of assets and the conditions under which that covariance is observed also matter. The market-liquidity and funding-liquidity analysis by Brunnermeier and Pedersen (2009) demonstrates how liquidity can contract together under stress. This is relevant to insurers even when most transactions originate outside the insurance sector: correlated sales, margin constraints and reduced dealer or fund liquidity transmit market dislocation to the ability to liquidate invested assets.

Insurance general accounts, linked accounts, and risk transfer to policyholders

A general-account equity loss typically affects the insurer's asset value and, subject to its contracts and regulatory accounting, its capital capacity. Linked or separate-account arrangements can pass some investment risk to policyholders, although minimum guarantees, expenses, surrender options, contractual pooling, operational obligations and investment-related guarantees may leave residual risk with the insurer. A shareholding that backs a unit-linked liability cannot be evaluated in the same way as an identical holding supporting guaranteed annuity benefits. Economic ownership, accounting presentation, legal segregation and residual insurer guarantees must be reconciled rather than inferred from product branding.

Insurance liabilities generate a second source of asymmetry. A long-lived life or annuity policy may extend for decades, while surrender behavior, derivative collateral, financing maturities, operating costs and claims can create near-term cash needs. Nonlife insurers may face concentrated catastrophe payouts and litigation settlements even when expected average liability duration is moderate. A long expected liability term is an insufficient proxy for the earliest plausible high-volume outflow. This remains true whether the alternative asset itself is high-quality or has a long average holding period.

Observed U.S. allocation evidence and its denominator

The NAIC Capital Markets Bureau reported about $9.58 trillion in U.S. insurer cash and invested assets at year-end 2025, on a book/adjusted carrying value basis. Bonds represented 59.8%, common stocks 13.6%, mortgage loans 9.1%, and other long-term investments reported in Schedule BA 6.7%. Those categories are distinct; Schedule BA is a statutory reporting basket that cannot be equated entirely with private-equity exposure. The 2025 insurer asset-class table further attributes about 77% of industry common-stock investments to property/casualty insurers, with roughly half of their approximately 32% common-stock portfolio share reported as affiliated holdings. By contrast, common stocks were about 4% of life insurers' portfolios according to that report.

These disclosures undermine any singular claim that “insurers allocate” one global share to equity. An industry aggregate combines business models and affiliated holdings; reporting classifications reflect accounting and statutory practice; book or adjusted carrying amounts differ from timely executable sale prices. Affiliated equity often embodies a group investment that cannot be liquidated independently without restructuring the operating group. Accordingly, portfolio analysis should publish at least insurer segment, direct-versus-affiliated status, valuation basis, fund look-through and relevant liabilities. The numbers here describe reported U.S. holdings at the specified date, not present-day market values or a causal investment recommendation.

The EIOPA October 2026 dashboard reports a median 6% equity and participations share in Q2 2026 for its relevant insurer sample, excluding exposures through collective investment undertakings; it separately reports median property investment of 2.8% and liquid assets of 46%. A median undertaking, a pooled U.S. aggregate, differing asset categories and different observation periods cannot be compared as if they estimate the same population parameter. The useful comparative observation is the importance of insurer type and exposure classification, not a spurious transatlantic league table.

Return expectations, valuation, and required capital

Expected return is the probability-weighted mean of future economic gains, net of costs, fees and taxes when those are relevant. It is distinct from a contractual yield and from realized historical performance. A security with higher expected return may demand enough capital, liquidity and management effort to produce a lower benefit to the insurer's policyholders and owners at the enterprise level. For example, a risky asset offering an illustrative expected gross return of 9% and charging 2% annual management costs does not deliver an identified 7% risk-adjusted margin: loss uncertainty, incentive fees, financial leverage, valuation errors, capital cost and illiquidity remain unpriced in that subtraction.

Regulatory capital is also not an expected loss. Under the EU standard formula, ordinary type 1 and type 2 equity exposures are subject to different prescribed stress charges, adjusted as specified. Article 169 of the delegated regulation gives, for ordinary exposures, 39% plus the symmetric adjustment for type 1 and 49% plus that adjustment for type 2. Qualifying long-term and strategic positions can follow separate treatment and eligibility rules. The symmetric-adjustment provision constrains the adjustment within prescribed limits. These are formulaic regulatory shocks, not empirical forecasts that a given share will lose precisely 39% or 49%. The resulting solvency capital requirement also depends on aggregation with other market and underwriting risks, and on applicable adjustments; a stand-alone equity shock cannot be presented as the insurer's final total capital requirement.

The independent analytical question is whether the additional equity return compensates the economic and regulatory burden for the actual balance sheet. Answering it requires an investment-level cash-flow projection, scenario-dependent liability comparison, valuation scrutiny and an explicit account of the liquidity and capital constraints the investment consumes. A return target on its own is not an asset–liability strategy.

Real Asset Exposures — Financial and Actuarial Mechanisms

What qualifies as a real-asset exposure

“Real asset” describes the underlying source of value, while “equity,” “loan” and “fund unit” describe financial claims. A commercial building may be owned outright by an insurer, held through a property company, financed by a mortgage, or represented within a multiasset partnership. The property itself has rental and terminal-sale cash flows, but a mortgage loan secured on that property remains primarily a credit claim. This Dossier examines ownership and equity-like exposure while treating debt service and secured financing as determinants of its residual value.

For an income-producing real estate asset, an unlevered discounted cash-flow representation is

V0=∑t=1TNOIt−CapExt(1+kt)t+TVT(1+kT)T,V_0=\sum_{t=1}^{T}\frac{NOI_t-CapEx_t}{(1+k_t)^t}+\frac{TV_T}{(1+k_T)^T},

where net operating income is income after relevant operating expenses, capital expenditure is separately identified to avoid double counting, and the terminal value reflects assumptions about remaining operating cash flows and market pricing. Consistent discounting requires a coherent risk-adjusted discount-rate term structure or equivalent certainty-equivalent method. It is analytically invalid to take a conservatively stressed cash-flow forecast and arbitrarily add an unrelated discount-rate penalty without identifying the intended stress scenario and double-counting risk.

The relevant exposures include occupancy, tenant failure, renegotiation, rent regulation, insurance costs, maintenance, obsolescence, property taxation, legal title, environmental liabilities and refinancing. Many costs rise with inflation, while rent adjustment is delayed or restricted by contract. Unobserved transaction costs and time-to-sale also matter. Leverage converts property downside into amplified equity impairment and refinancing dependence. Environmental and physical climate exposures may impair the asset even when insurance protection exists, particularly through deductibles, exclusions, availability changes or adaptation costs; specific climate loss quantification belongs to an appropriately scoped hazard model.

Infrastructure income as a collection of contractual states

Infrastructure equity includes interests in transportation, utilities, energy networks, digital infrastructure, water and social facilities. Its income may arise from availability payments, regulated tariffs, concession fees, volume-dependent charges or merchant prices. These are different economic structures. An availability-based infrastructure contract can reduce traffic-volume exposure but may introduce public-counterparty, contract-performance, maintenance and political risk. A merchant power project may have strong nominal revenue opportunity yet remain highly exposed to energy prices and construction costs. Long concession terms can coincide with substantial leverage, refinancing maturity walls, counterparty concentration and periodic capital expenditure.

The OECD infrastructure-investment review documented life-insurer participation in infrastructure and the predominance of unlisted fund and direct-equity routes in its studied sample. That study describes a particular data collection and period rather than a universal current insurance share. Its relevant lesson is structural: direct and fund routes give different control, information and exit rights, while liability horizon helps explain investment demand.

An insurer considering infrastructure as a liability-matching asset must map the timing and enforceability of its distributions. An expected annual dividend is a managerial forecast; it is not a legally identical cash obligation to the insurer's benefit-payment schedule. Contract indexation may be partial, subject to caps or lagged, and can be interrupted by public policy changes or disputes. Revenue exposure to an apparently distinct industry may also correlate with insurers' own claim demands—for example, when a severe regional hazard disrupts infrastructure income and simultaneously causes insured losses.

Risk budgets and equity leverage in real assets

For a project with asset value A and senior net debt D, simplified equity value is E = A – D, subject to claim priorities and omitted contingent items. If illustrative A is 100 and D is 70, E is 30. A 15% fall in A to 85 reduces E to 15, a 50% equity loss, assuming the debt claim remains at 70. The arithmetic is a sensitivity identity, not a prediction about infrastructure volatility. Additional lender covenants, mandatory amortization and equity injections can exacerbate the economic consequences before the asset's long-run operating value is realized.

For an insurer, the equity instrument's valuation should be reconciled with the underlying asset cash flows, lender terms and contingent commitments. Two funds each described as “infrastructure” may have radically different debt covenants, counterparty portfolios, payout discretion and construction exposure. Risk aggregation requires a look-through by geography, sector, counterparties, manager, refinancing date and correlated hazard, not only by the marketing category. Infrastructure's physical persistence does not ensure financial-claim durability.

Distinguishing economic hedges from narrative correlations

Inflation-sensitive assets are sometimes proposed as matching instruments for inflation-sensitive claims. A justified hedge requires cash-flow sensitivity to the same inflation measures, timing, contractual indexation terms and currency as the liabilities. Property income may adjust slowly, while wage-related bodily-injury claims can rise rapidly. A utility's indexed tariffs may adjust annually and be subject to regulatory reset while annuity payments are monthly. These differences can turn a long-run correlation into a short-horizon funding mismatch.

Let an insurer's relevant balance-sheet surplus be S = A – L. Under an inflation or rate state x, the local sensitivity is ∂S/∂x=∂A/∂x−∂L/∂x\partial S/\partial x=\partial A/\partial x-\partial L/\partial x. An asset is economically protective if the joint change in surplus and cash availability improves in the adverse states that matter. Standalone positive inflation beta is insufficient. Valuation should distinguish total surplus sensitivity, the schedule of realized distributions, and any derivative or reinsurance hedges already in place.

Valuation and the real-asset holding problem

Appraisal and discounted-cash-flow estimates can be informative while remaining insensitive to rapidly changing transaction conditions. IFRS 13's fair-value framework defines an exit price using market-participant assumptions at the measurement date. The intention to hold indefinitely does not redefine a fair-value measurement. This distinction directly matters when the economic purpose is long-term cash generation: an insurer may rationally plan not to sell, but its risk committee must still estimate the price and time available if that plan is disrupted.

A governance-grade real-asset file should therefore retain a cash-flow forecast, market assumptions, debt terms, valuation methodology, independent challenge, legal and environmental risks, downside scenarios, and expected transaction-time and discount analysis. The key inference is that “real” describes the source of economic activity, not an exemption from valuation, liquidity or solvency risk.

Private Equity and Alternatives — Quantitative Methods and Assumptions

Private partnership interests as sequences of contingent cash flows

An interest in a private-equity partnership represents rights and obligations governed by the fund's constitutional and contractual documents. The limited partner typically commits a capital amount, receives drawdown notices over an investment period, pays fees or expenses under stated terms, and receives distributions after realization events. The general partner or manager controls timing and many operating decisions. The insurer's risk therefore depends on commitment size, outstanding callable capital, portfolio company leverage, the partnership waterfall, governance terms, related-party transactions, borrowing facilities and restrictions on transfer. An unfunded commitment is a potential future cash outflow and belongs in the same liquidity analysis as the insurer's other contingent obligations.

The cash-flow structure makes performance harder to compare with a continuously traded index. An initial commitment of 100 is not necessarily an immediate investment of 100: the manager may draw 20 now, 30 later and 50 in subsequent years, while the insurer must reserve or plan the capacity to fund future calls. Fund net asset value (NAV) represents an estimate at a reporting date; distributions are actual realized proceeds. Reported NAV, the investor's share of underlying enterprise value and executable secondary-market proceeds need not coincide.

For period-end convention, a common aggregate private-fund metric is

TVPI=Cumulative distributions+Remaining NAVPaid-in capital;DPI=Cumulative distributionsPaid-in capital.TVPI=\frac{\text{Cumulative distributions}+\text{Remaining NAV}}{\text{Paid-in capital}}; \qquad DPI=\frac{\text{Cumulative distributions}}{\text{Paid-in capital}}.

These multiples reveal different facts: total modeled plus realized value, and distributions actually returned. They do not by themselves measure the timing of contributions or the market risk of the residual portfolio. An internal rate of return (IRR) is the discount rate solving ∑t=0TCFt/(1+r)t=0\sum_{t=0}^{T} CF_t/(1+r)^t=0 for consistently timed net cash flows. For irregular cash dates an XIRR-style day-count convention is appropriate. IRR may be nonunique when net cash-flow signs change more than once, may reward early distributions disproportionately, and does not necessarily estimate the gain obtainable on a repeatable invested-dollar strategy. The investment committee should examine full dated cash flows as well as the quoted summary statistic.

Kaplan and Schoar's (2005) private-equity performance and persistence study found substantial heterogeneity and relationship-level persistence in its historical sample; its return comparisons depended on the period and the data used. Harris, Jenkinson and Kaplan (2014), in Private Equity Performance: What Do We Know?, documented stronger buyout results in a later dataset while distinguishing buyout from venture outcomes. Different samples, vintage selection, net-fee treatment, benchmark definitions, incomplete fund liquidation and treatment of remaining NAV explain why these studies do not provide a universal forward return premium that an insurer can book as an assumption.

Public-market equivalents and counterfactual comparators

A public-market-equivalent (PME) measure asks how the fund's cash calls and realizations compare with investing on the same dates in a selected public benchmark. A simplified terminal-wealth-style ratio can be written

PME=∑tDt IT/It+NAVT∑tCt IT/It,PME=\frac{\displaystyle\sum_t D_t\,I_T/I_t+NAV_T}{\displaystyle\sum_t C_t\,I_T/I_t},

where DtD_t and CtC_t are dated distributions and contributions, ItI_t is a total-return benchmark index and TT is the comparison endpoint. This illustrative convention must not be confused with all PME variants, which differ in reinvestment mechanics and terminal-value treatment. A value above one can indicate outperformance against that chosen index under the stated assumptions, but cannot isolate skill from systematic risk, financial leverage, selection bias or favorable vintage timing.

An insurer needs an additional comparator: the financing and solvency consequences of the alternative investment relative to an eligible portfolio that would otherwise back the same liabilities. A private fund that outperforms public equity after fees can still produce inferior outcomes for an insurer if its drawdowns occur during large claims, its secondary-sale discounts are severe, or its stressed capital demand exceeds the benefits of return. The insurer should report both investment performance and enterprise-level surplus and cash-flow effects rather than present a PME as an asset–liability result.

The difference between return volatility and reported NAV volatility

Many private assets trade infrequently; independent transaction prices are sporadic and fund valuations are updated periodically. An observed quarterly return can be a mixture of current and lagged economic information. Getmansky, Lo and Makarov (2004) analyze the connection between serial correlation and illiquidity, demonstrating how smoothing can understate volatility and distort measured risk-adjusted performance in hedge-fund settings. This analysis does not imply that every private-equity NAV series follows the same statistical mechanism. It identifies a general measurement hazard: a slowly revised estimate is not equivalent to a stable liquid market value.

A deliberately simple illustration separates true economic return rt∗r_t^* from reported return rtr_t:

rt=θrt∗+(1−θ)rt−1∗,0<θ≤1.r_t=\theta r_t^*+(1-\theta)r_{t-1}^*,\qquad 0<\theta\le 1.

If economic returns are independent with variance σ∗2\sigma_*^2, the reported variance under this illustrative two-period filter is [θ2+(1−θ)2]σ∗2[\theta^2+(1-\theta)^2]\sigma_*^2. At θ=0.5\theta=0.5, reported variance is half the underlying variance and observed one-period volatility is approximately 70.7% of underlying volatility. This is arithmetic under a specified filter, not an estimated empirical smoothing coefficient for insurer investments. The filter also creates serial dependence and changes observed covariance with assets marked more promptly. Risk estimates should inspect appraisal frequency, lagged response, cash realizations and possible changing economic dependence.

Alternatives beyond private equity

Hedge funds, commodities, commodity-linked exposures and nontraditional derivatives are often grouped as alternatives even though their payoff shapes differ. A hedge fund using short selling, relative-value leverage, prime brokerage, derivatives or illiquid credit can have substantial gross exposures and contingent collateral requirements despite a small reported net position. A commodity holding or commodity-linked derivative may provide sensitivity to input prices but can have negative roll returns, margin calls and basis risk. A fund of funds adds intermediate expenses, potential hidden concentration and restricted redemption schedules. A complex investment structure can distribute risk to more than one investment vehicle without dispersing ultimate economic exposure.

The relevant identification questions are therefore: What is the underlying economic factor or contractual payoff? Who bears leverage and financing risk? How frequently can value be verified externally? How much committed capital can be demanded? Who may suspend or gate redemptions? What is the worst plausible overlap with insurance liabilities? These questions apply to both “alternative” and apparently conventional instruments. Reclassification from an exchange-listed security to a fund unit does not eliminate the underlying equity exposure.

Private-equity-owned insurers versus insurer-owned private equity

Private-equity ownership is a separate governance and systemic-risk subject with direct connections to this Dossier. In its year-end 2024 PE-owned insurer study, the NAIC identified 137 PE-owned U.S. insurers with approximately $704.3 billion in cash and invested assets. The reported portfolio was dominated by bonds and mortgages, at approximately 61% and 20% respectively. Ownership does not imply that the $704.3 billion was invested in PE partnership equity. The NAIC emphasizes its evolving identification process, and the count must not be extrapolated uncritically to an international insurer population.

The BIS examination of life insurance and financial stability explains the interaction of private-asset origination, insurer ownership, asset-intensive reinsurance and cross-border risk transfer. It identifies channels through which investment and fee incentives, related-party arrangements and balance-sheet complexity can reinforce each other. That institutional interaction does not establish that any particular ownership form is inherently imprudent; it makes control of conflicts, transparency of transactions, intragroup exposures and claims-paying security empirically relevant. The owner, the asset manager, the originator, the reinsurer and the ultimate fund vehicle may be legally separate yet economically connected.

Fee, leverage and vintage sensitivity

Private investment comparisons require the full economics of management fees, monitoring fees, performance allocations, carried interest, expenses, financing fees, subscription credit lines, tax structure and transaction costs. A reported gross asset IRR is not the net limited-partner return. A subscription line may postpone the date of a recorded investor cash call and increase a cash-flow-weighted IRR without changing the economic acquisition return of the underlying investment. The same financing facility may add refinancing exposure, covenant obligations and a dependence on the manager's lending relationships.

Vintage-year risk is also substantial. Two equally skilled managers can buy economically similar assets at different multiples and interest-rate environments. Portfolio companies acquired after a prolonged expansion in valuations can be sensitive to exit-market repricing. A governing board should distinguish diversification across managers from diversification across transaction vintages, business-cycle stages, sectors, financing sources and ultimate holdings. A roster of fund names is a weak proxy for independent risk if several funds buy from or sell to each other, borrow from the same intermediaries, or depend on the same macroeconomic conditions.

The analytic conclusion of this chapter is methodological: private assets must be assessed through dated economic cash flows, look-through holdings, contingent commitments and realistic sale conditions. A NAV-based return series is one evidentiary input, not the risk distribution of the underlying insurer investment.

Measurement and Liquidity Challenges — Insurance Applications and Comparative Evidence

Three distinct values for one position

A position can have a carrying amount in regulatory or financial statements, a model-based economic value at the reporting date, and a cash amount realizable within a specified sale horizon. Those numbers answer different questions. Differences may be lawful and informative because the reporting frameworks, underlying measurements, bid–ask costs and timing differ. Problems arise when a portfolio committee silently substitutes one concept for another. A commitment to hold an investment for ten years is relevant to business strategy; it is not an accounting reason to ignore a market-participant exit price where fair value is prescribed.

IFRS 13 Fair Value Measurement establishes an exit-price objective and market-participant assumptions. Its hierarchy distinguishes unadjusted active-market quoted prices for identical items (Level 1), other observable inputs (Level 2) and significant unobservable inputs (Level 3). Classification identifies the observability of inputs, not a guaranteed return ranking or a binary test of investment quality. Private stakes, real estate and infrastructure can require substantial Level 3 estimates. Valuation techniques can be rigorous, but uncertainty and possible manager incentives demand documentation, calibration, consistency and challenge.

IFRS 9 Financial Instruments governs recognition and classification of financial assets within its scope; it distinguishes measurement categories by business model and contractual cash-flow features, with relevant equity measurement and presentation options. The accounting category of an asset does not erase its economic drawdown, settlement or market risk. An insurance group reporting under IFRS and a U.S. insurer filing statutory investment schedules may publish different balance-sheet figures for related economic holdings. Comparisons must identify which reporting convention and level of consolidation each datum uses.

The U.S. NAIC investment reporting series is especially useful for asset-class distribution, but Schedule BA assets comprise a heterogeneous category. Reconstructing private-equity holdings requires looking beyond the Schedule BA aggregate, including relevant statutory descriptions and vehicle characteristics. Conversely, listed common stock held in an affiliate is both a stock for reporting purposes and a group-related investment for governance; its ordinary public-market liquidity may not be available to the insurer in isolation.

Appraisal and model risk

Valuing an illiquid business or property often requires estimated future income, discount rates, capitalization rates, leverage and terminal values. Every such input can be correlated with market conditions and insurance liabilities. The effect of a lower discount rate on estimated NAV can be much larger for distant terminal value than for near-term operating cash. A higher terminal exit multiple may also raise today's modeled value while increasing dependence on future capital-market liquidity. Professional valuation should disclose this sensitivity rather than rely on a single central estimate.

A robust review examines alternative market inputs, independent transaction evidence, subsequent realizations, changes in comparable spreads or multiples and the time elapsed since the last truly independent price. It also distinguishes fair value from conservative solvency adjustment: once a valuation uses a market-consistent risk premium, a second arbitrary “prudence” deduction requires a specifically identified purpose. A separate liquidity contingency or capital stress is legitimate when labeled as such, but it must not be concealed in an internally inconsistent carrying value.

The statistical consequence of smoothed reporting is important. Observed low correlation between quarterly private-fund NAV changes and contemporaneous public equity may partly reflect delay in valuation recognition. The Getmansky–Lo–Makarov methodology illustrates why estimated risk-adjusted performance should be sensitive to serial dependence, stale marks and liquidity. An insurer can assess this by comparing risk estimates from reported NAVs, independently appraised underlying holdings, comparable public price shocks and actual secondary transactions. None of these is automatically the single “true” measure; agreement and disagreement between them are evidence about model uncertainty.

Liquidity as an event- and horizon-dependent property

Liquidity is the ability to convert a particular asset into usable cash at a relevant time and cost, subject to enforceable restrictions. A fund might permit quarterly redemptions with long advance notice, limited windows, discretionary gates and suspension clauses. A property may be saleable in normal markets but require months of diligence and finance. A private partnership may have a secondary market where transfer consent is required and discounts widen precisely when the insurer most needs liquidity. The distinction between liquidity in normal trading and liquidity in stress is fundamental.

In an insurance context, short-horizon liquidity demand can arise from claims, catastrophic events, surrender behavior, derivatives, reinsurance settlements, asset-intensive reinsurance collateral, capital calls on investments, operating expenses and funding-line covenants. Stressed funding needs can coexist with lower investment prices. Brunnermeier and Pedersen's (2009) funding-liquidity mechanism explains how market participants' funding constraints can magnify the very discounts faced by forced sellers. Insurers with stable liabilities may be less subject to daily investor redemption pressure than open-ended funds, but they can nevertheless be connected to the same market liquidity spiral through their liabilities and financing contracts.

The BIS analysis of asset-intensive reinsurance adds a specific insurance channel: investment assets can be pledged or transferred within reinsurance arrangements while collateral maintenance depends on contract terms and market values. A claim that “liabilities are long-term” should therefore be followed by a cash-flow map including collateral provisions, surrender terms, potential ratings triggers and cross-defaults. Those provisions can accelerate cash needs even if promised ultimate benefits mature later.

Supervisory evidence for Europe and the United States

The EIOPA June 2026 Financial Stability Report observed continuing demand for alternative and less liquid assets, while identifying valuation opacity, market liquidity, structural complexity and contagion as relevant vulnerabilities. Its Q4 2025 breakdown describes a significant illiquid-asset share, with life general-account illiquid exposures reported near 30.4% under that report's classification, including mortgages, unlisted equity, illiquid funds, nontraded corporate bonds and property. This estimate is wider than the noncredit equity and real-asset subject of the present Dossier. It must therefore not be relabeled a 30.4% private-equity share. Holdings identified as “illiquid” also depend on a supervisory classification that is different from the NAIC's Schedule BA definition.

The NAIC's 2025 investment mix provides a different but complementary observation: a growing common-stock share and Schedule BA share alongside the predominance of bonds. The NAIC's private-equity ownership monitoring shows that supervisory interest also concerns governance, affiliated transactions and business-model changes. Neither source offers a clean causal test of whether alternatives improved insurers' resilience. Such a claim would require matched insurer histories, consistent liability and capital measures, realized stressed returns and credible treatment of selection, ownership and jurisdictional differences.

The October 2026 EIOPA risk dashboard further notes that market risk remained high with an increasing outlook and provides a median liquid-asset share near 46% in Q2 2026. A median liquid-asset ratio does not measure cash accessible within a specified shock horizon. Its analytical use is context: supervisors monitor assets and liquidity separately because each measure addresses a different solvency channel. A self-reported industry average similarly cannot resolve the adequacy of an individual insurer's cash-flow ladder.

Look-through and the danger of the “fund wrapper”

Article 84 of the EU delegated regulation provides a look-through approach for fund investments in calculating the solvency capital requirement, together with conditions and alternatives when full look-through cannot be applied. Article 168 provides equity risk classifications that encompass specified indirect or alternative exposures. These rules counter the temptation to classify a fund solely by its name or legal wrapper. The exact treatment depends on qualifying conditions, asset types, information available and the rules applicable at the valuation date.

The same principle has economic force beyond regulation. Several funds may each report diversified portfolios while shareholding exposures ultimately depend on the same listed technology companies, regional real estate tenants or bank credit lines. An insurer needs a consolidation view that maps direct investment, fund look-through, affiliates, collateral and economic derivatives without double counting. If full look-through is unavailable, the residual unknown exposure should be identified and assessed conservatively as model uncertainty; assuming it is diversified is not a scientifically justified substitute for data.

Measurement findings

The strongest cross-source conclusion is that valuation reliability and liquidity should be treated as independent dimensions. An appraisal can estimate an economically defensible long-term cash-flow value while offering little information about the cash attainable next week. Conversely, a liquid listed equity provides frequent price discovery but can suffer a severe immediate market decline. The insurer's enterprise policy should measure both and connect them to liability timing. For private assets, the same reporting package should disclose NAV method, valuation age, observability, sensitivity, realizations, unfunded commitments, contractual exit restrictions and secondary-sale evidence. A final reported portfolio number without those attributes obscures rather than summarizes risk.

ALM Interactions — Stress Testing, Model Limits, and Verification

Asset–liability management as a joint stochastic problem

Asset–liability management (ALM) concerns the joint evolution of invested assets and insurance obligations. Its basic accounting identity is surplus St=At−LtS_t=A_t-L_t, with valuation definitions determined by the intended analysis. The economic risk of an investment allocation is a distribution of future surplus and available cash rather than a vector of asset returns alone. Cash liability estimates include benefits, claim settlements, expenses, policyholder options and relevant guarantees. They are conditional on mortality, morbidity, catastrophe events, lapses, interest rates, inflation, operational continuity and legal obligations. Contractual and behavioral uncertainty can interact with investment stress.

At a horizon TT, a minimal balance-sheet scenario records AT(s)A_T(s), LT(s)L_T(s), and ST(s)S_T(s) for scenario ss. A liquidity assessment adds the usable cash immediately available by each required date, contractual cash inflows, committed capital calls and financing or asset-sale constraints. When benefits are paid, cash assets and outstanding obligations may both decline; this is not automatically an additional economic loss. A sale discount relative to a contemporaneous fair value, transaction fees or financing penalty can create a separate loss. Keeping these flows separate prevents the stress test from double counting the same claim payment and its financing.

The IAIS supervisory framework connects investment risk to capital adequacy and integrated enterprise risk management. It recognizes the interdependence between the characteristics of investments and insurance liabilities. The EU Article 132 investment obligations likewise refer to risk identification, liquidity and the nature and duration of liabilities. An insurer's risk governance must translate those broad requirements into time-resolved business evidence.

Why asset duration cannot resolve equity uncertainty

Duration approximates price sensitivity to discount-rate changes for appropriately defined fixed-income or valued liability cash flows. Common equity has a residual uncertain cash-flow profile, and a private-equity fund has discretionary distribution and contribution timing. A duration estimate based on a modeled asset's distant terminal value is useful as a valuation sensitivity, but it is not a promise of coupon or principal that can be matched to a scheduled benefit. It may also be unstable when the business model, discount rate and forecast change together. This explains why an insurer can have an apparently long-duration portfolio and still bear a large mismatch in near-term cash.

Suppose an annuity insurer owes long-term fixed benefits and adds equity in place of matched bonds. The equity may produce favorable long-run surplus under some growth states. Yet it also increases sensitivity to a recession that reduces stock values while long-duration guaranteed liabilities remain payable. Even where falling yields raise bond and liability present values together, equity prices and lapses can react differently from those matched cash flows. The insurer must measure the conditional distribution of surplus and cash under joint economic paths, not rely on historical average returns to replace a contractual cash-flow hedge.

Transparent example: portfolio loss and liability interaction

The following stylized calculation is intentionally hypothetical, uses arbitrary monetary units denominated as billions and represents no named insurer. Assume assets of 100, liabilities of 92 and starting own funds of 8. The asset allocation comprises 60 conventional fixed-income assets, 13 liquid reserves and short investments, 8 listed equities, 6 private-equity interests, 8 income-producing real estate equity and 5 infrastructure equity. The six weights sum to 100. For one scenario, assume price changes of −6%, 0%, −30%, −35%, −25% and −20%, respectively. The individual asset losses are 3.6, 0, 2.4, 2.1, 2.0 and 1.0, which sum to 11.1. Assets decline to 88.9.

Suppose in the same scenario the remeasured economic value of liabilities declines by 4 because the joint yield curve and liability cash-flow assumptions imply that change. Then liabilities are 88 and remaining economic surplus is 0.9. The initial surplus of 8 falls by 7.1, the difference between 11.1 of asset losses and 4 of liability relief. The liability reduction is stipulated for the example and is not derived from a universal duration assumption. If liabilities instead remained at 92, the same asset losses would produce a deficit of 3.1. If a liability guarantee increased the insurer's obligations during a market decline, the result would worsen further. The choice of liability scenario is therefore as decisive as the investment shock.

Position Starting value Illustrative market change Value loss
Conventional fixed income 60 −6% 3.60
Liquid reserves and short investments 13 0% 0.00
Listed equity 8 −30% 2.40
Private equity 6 −35% 2.10
Real estate equity 8 −25% 2.00
Infrastructure equity 5 −20% 1.00
Total 100 Scenario-specific 11.10

The table's shocks are exogenous assumptions, not observed crisis losses, calibrated regulatory capital charges or estimates of asset-class worst cases. The private-equity shock is imposed on contemporaneous economic value; a reporting-date NAV may recognize it later. No standalone stress of this kind can claim an exact probability without calibrated distributions and dependence assumptions. The example demonstrates sensitivity of economic own funds to the balance of assets and liabilities.

Liquidity overlay and the cost of immediate cash

Continue the illustration separately. Assume total near-term required cash of 9, including claims and collateral, but only 4 of the 13 liquid reserves and short investments is contractually or operationally available within the required window; the rest settles later or is restricted for the assumed period. This produces a date-specific cash gap of 5. If the insurer must sell another stressed asset at a 15% discount to its immediately preceding stressed market valuation, raising 5 requires disposing of assets with a preceding value of 5/0.855/0.85, approximately 5.882. The incremental sale discount is about 0.882 before additional costs. Starting from post-market-shock surplus 0.9, that incremental loss would leave approximately 0.018, assuming no other profit, taxes, liability changes or capital movements.

Paying the claims already included in the liabilities is a simultaneous settlement of assets and obligations; it is not charged a second time to surplus. The incremental discount is separately recognized because the asset is sold below its assumed contemporaneous value. If instead the portfolio's recorded “market value” had already been marked to executable bid prices including this discount, subtracting it again would be wrong. Conversely, if a private-asset valuation remains stale, an apparent absence of sale loss could conceal prior economic impairment. The example therefore identifies an audit principle: distinguish prices, valuations, timing restrictions and realized cash losses explicitly.

The available liquidity of 4 is a scenario assumption, not a claim about the NAIC or EIOPA median cash position. The example illustrates why even an insurer with 13% nominally liquid or short-term assets may be forced to sell if asset accessibility and liability settlement dates are mismatched. The more useful operational quantity is cash usable by each specified deadline under a defined joint scenario, after eligibility constraints and margin conditions.

False diversification from outdated marks

Consider a hypothetical two-bucket investment program: 70% publicly valued assets with annualized volatility of 5% and 30% alternatives reported at 14%. Assuming a return correlation of 0.2 yields portfolio variance

σp2=(0.70×0.05)2+(0.30×0.14)2+2(0.70)(0.30)(0.05)(0.14)(0.2)=0.003577,\sigma_p^2=(0.70\times 0.05)^2+(0.30\times 0.14)^2+2(0.70)(0.30)(0.05)(0.14)(0.2)=0.003577,

or portfolio volatility near 5.98%. If the true correlation in the relevant stressed economic state is 0.8 with otherwise unchanged inputs, the variance is 0.005341 and volatility about 7.31%. If corrected volatility for alternatives is instead 24% and the correlation is 0.8, the same formula yields variance 0.010441 and volatility about 10.22%. These are constructed sensitivity cases, not estimates derived from market data. All calculations use the standard two-asset variance identity and are mechanically reproducible.

The inference is more important than the particular numbers. A measured low correlation can arise from true diversification, lagged appraisal, selective marks, a benign sample or the use of different reporting horizons. Under joint funding stress the effective correlation and marginal liquidity cost can be much higher. The serial-correlation study of illiquid funds warns against treating smoothed reported returns as ordinary market-traded observations. A credible insurer study should estimate and disclose the effects of lag adjustments, valuation errors, fat-tailed events and time-varying dependence, while acknowledging the identification limits of those adjustments.

Stochastic and deterministic tests

A general stochastic implementation simulates joint equity returns, real asset revenues, interest-rate and inflation paths, credit and financing conditions, liability cash flows, lapse or surrender behavior, insurance losses, policyholder options and investment-fund capital calls. The distribution of available surplus can then be measured at multiple horizons. Dependence assumptions need expert justification; a model that fixes correlations near benign historical averages may exaggerate diversification during precisely the events that threaten policyholder security. Parameter estimation error and limited observations for severe events should be documented rather than obscured within a single confidence number.

Deterministic stresses remain indispensable because a historical sample may contain few comparable insurance and asset-market crises. Stress scenarios can combine a property-price decline with a catastrophe-driven claims surge; a recessionary equity drawdown with rising annuity withdrawals; a credit and equity selloff with fund capital calls; an interest-rate jump with collateral needs; a delayed infrastructure-construction completion with unavailable dividends; and a cyber or operational event that interrupts asset settlement and claims processing. Each scenario specifies the initiating event, time path, legal and operational assumptions, asset-price changes, liability effects, financing and asset-sales constraints, and management actions. Simultaneous effects should be consistent: for example, a project-finance impairment and sponsor loan default that represent the same underlying cash-flow loss cannot both be applied to one exposed asset as independent full write-downs.

Insurance-specific ALM segmentation

Life and annuity general accounts commonly have substantial predictable long-term obligations, but guarantees, mortality, longevity, policyholder behavior and collateral can alter cash needs. Property/casualty portfolios can have shorter claims horizons, catastrophe aggregation and legal-reserve uncertainty; some have material long-tail claims. Health and disability lines differ in payment frequency, trend exposure and operational liquidity demands. Reinsurance contracts may add premium and claims settlement timing, collateral requirements and counterparty risk. These distinctions matter more to asset suitability than a broad label of “insurance company.”

An insurer should segment assets supporting guaranteed obligations, linked liabilities, discretionary surplus, required capital and collateral arrangements; then identify whether assets are legally available across segments. For example, a subsidiary's listed stocks may appear in group consolidated statements but cannot necessarily be mobilized to pay another subsidiary's policyholders without approvals or capital consequences. Similarly, an asset that economically generates reliable cash flow may be pledged and unavailable to general creditors. The ALM analysis needs both the consolidated exposure and the legal-entity cash accessibility map.

Model validation and limitations

The first model-validation question is whether data actually measure the risk being claimed. Manager-supplied NAVs, regulatory book values, listed closing prices and executable bids represent different inputs. Historical stress tests should preserve observation time stamps, vintage composition and reclassification history. Backtesting should compare previously reported investment values with later independent transactions and realized distributions, while accounting for market movement and deal selection between those dates. Errors should be explained by valuation assumptions, not simply absorbed into a generalized reserve.

Second, uncertainty in the model is different from ordinary modeled market risk. A 99.5% capital measure, an internal economic capital estimate and a liquidity stress can all be calculated correctly yet omit hazards not identified in their input data or model structure. Governance should challenge expert assumptions about extreme dependence, fund gates, construction delays, legal settlement, capital calls and pricing of correlated positions. This Dossier's stress calculations demonstrate arithmetic consistency and conceptual sensitivity; they are not a calibrated insurer model, a regulatory own-risk assessment, or an empirical validation of any probability distribution.

Stress and Governance — Integrated Interpretation and Implications

International supervisory principles and the current EU perimeter

The IAIS Insurance Core Principles provide the international supervisory baseline. ICP 15 addresses investment requirements and the relationship of investment quality to solvency; ICP 16 addresses integrated enterprise risk management, including ALM, liquidity, risk appetite and own-risk evaluation. These are globally relevant supervisory standards whose national implementation differs. They support a substantive institutional question: can an insurer identify, measure, monitor, manage, control and report the full risks it has accepted for the sake of a portfolio return?

Within the EU, Article 132 of Solvency II applies the prudent person principle to all assets and requires investment in a manner appropriate to the nature and duration of insurance liabilities. The detailed capital regime categorizes equity risks and applies specified stress treatment. Article 168 of the delegated regulation defines relevant equity risk groupings, including alternative-investment treatment in specified situations. Article 169's equity stresses distinguish ordinary type 1 and type 2 positions, while separate qualifications address strategic holdings and certain long-term equity positions. These provisions must be read together with fund look-through, infrastructure qualification, symmetric adjustment, aggregate SCR and applicable transitional or amended rules.

Article 171a on long-term equity makes special treatment conditional on demonstrable criteria concerning identified investments, assignment to identified obligations, management and holding capacity. Its existence does not establish that every unlisted share held by a life insurer earns a preferential charge. Nor is a long investment intention sufficient evidence that liabilities and cash needs will allow the intended holding period in all relevant states.

Temporal accuracy is essential. EIOPA identifies revisions taking effect on January 30, 2027, including changes under Directive (EU) 2025/2 and the related delegated framework. That future application date is after the November 15, 2026 research cutoff. This Dossier discusses currently accessible Solvency II rules as relevant to the 2026 period and flags the incoming changes for later release-date checking. No future rule is silently imposed retrospectively. A material regulatory update before eventual publication would require targeted revalidation of the affected paragraphs and evidence, rather than a wholesale rewriting of otherwise stable scholarly analysis.

U.S. statutory measurement and supervision

The U.S. regime combines state-based solvency supervision, statutory accounting, investment eligibility requirements and risk-based capital frameworks that differ in construction from the EU's Solvency II formula. Investment portfolios reported to the NAIC must therefore be analyzed through their actual statutory categories, affiliated holdings, valuation rules and individual insurer context. Applying a European 39% or 49% standard-formula factor as though it were a universal U.S. equity capital requirement would be a legal and analytical error. This monograph deliberately avoids a single numerical NAIC RBC factor because the correct treatment depends on current RBC instructions, security classification, insurer type and potential changes during the reporting period.

The NAIC year-end 2025 asset-mix report and its PE-owned insurer report jointly highlight why exposure and ownership must be separated. The first measures positions across insurers; the second identifies an ownership-defined subset and describes its asset mix. The latter's high bond and mortgage shares contradict simplistic descriptions of private-equity-owned insurers as holding predominantly private-equity stakes. Relevant supervisory issues also include affiliate transactions, asset manager affiliations, reserve assumptions and reinsurance arrangements, as the BIS review of PE-related links explains.

Cross-country comparisons should therefore focus on comparable economic questions—portfolio risk, legal-entity liquidity, leverage, transparency, capital resilience and policyholder claims-paying security—while reporting distinct statutory measures explicitly. A rank ordering of national portfolio percentages without denominator and valuation harmonization is misleading.

Risk appetite and governance obligations

An effective insurer investment policy should be a written set of investable-universe definitions, prohibited structures where appropriate, risk limits, delegation and escalation rights, capital and liquidity thresholds, valuation independence, affiliated-party transaction controls and reporting expectations. The board retains accountability for the risk appetite; investment executives implement it through documented mandates; independent risk and actuarial functions evaluate liability compatibility; valuation experts verify pricing; finance reconciles statutory and general-purpose accounts; compliance monitors applicable rules; internal audit tests operation. External asset managers can perform delegated tasks, but the insurer's understanding and governance responsibilities remain.

A formal due-diligence file for a private or real-asset investment should document ultimate beneficial economic exposures, investment mandate, partnership legal terms, leverage, bridge and subscription financing, manager compensation, related-party business, capital-call and default provisions, redemption and transfer rights, side letters, collateral, valuation policy and independent evidence. It should include adverse-case valuations, cross-fund concentration, investment and claim cash-flow calendars, relevant currency and geography exposure, disaster and transition risk, legal enforceability and operational arrangements. A stress number divorced from those terms is not adequate assurance.

A board should also see the differences among current observable value, best-estimate model value, conservative stress value and rapid-disposition proceeds, with explicit methods and measurement dates. The reporting process must distinguish a manager's reported fund NAV from an insurer-verified economic valuation. Independent challenge can use benchmark transactions, subsequent exits, independent valuations, comparable listed businesses and realized fund cash flows. A valuation exception or missing underlying holding should trigger a documented response commensurate with its potential effect, not an assumption that the diversification benefit is intact.

Liquidity governance and contingency plans

The cash-flow ladder should project investment income, realizations, fund drawdowns, pension or annuity payments, general insurance claims, expense requirements, tax, collateral and reinsurance settlements by relevant deadline. Management actions, such as asset sales, repo financing, derivative unwinds or subsidiary dividend extraction, should be modeled only when legally and operationally executable. In stress, repo availability, counterparties and market depth may decline together. A capital injection that has not been contractually committed cannot be modeled as certain. Likewise, a planned private-fund secondary sale should use a stressed time-to-completion and price range, not an assumption of immediate liquidation at last-reported NAV.

Concentration control should operate through multiple maps: issuer and counterparty, investment manager and originator, funding bank, underlying collateral, region, economic sector, vintage, currency, property hazard, infrastructure concession, legal entity and reinsurance affiliate. Correlations across these maps can be substantial even when fund-name diversification appears high. The investment policy should impose escalation when exposures cannot be traced, expected liquidity is based on discretionary gates, or a manager's investment discretion materially changes underlying risk.

The EIOPA supervisory analysis of alternatives links private-market complexity and valuation vulnerabilities to financial stability monitoring. Its significance for governance is procedural: a capital ratio can remain apparently comfortable in normal conditions while operationally available cash and promptly recognized loss values deteriorate. Liquidity, model risk and capital need coordinated monitoring, including credible contingency actions.

Confronting the principal decision tradeoffs

A board deciding whether to increase alternative allocations has to consider expected net return and possible diversification alongside capital absorption, cash-flow uncertainty, lockups, catastrophe overlap, pricing opacity, fees, governance and related-party conflicts. A long investment horizon can improve the capacity to tolerate temporary market fluctuations only if liabilities and capital support that holding period. A fund's lower reported volatility may be informative about appraisal timing rather than economic risk. A high fund IRR can depend on timing and leverage. An infrastructure asset may support long-term cash flow while exposing the insurer to construction and regulatory events. These observations arise jointly from standard asset pricing, cash-flow analysis and supervisory evidence; they do not amount to a new proprietary investment doctrine.

One important limit of current evidence is causal attribution. Supervisory allocation data and academic fund samples document holdings and historically measured returns; they do not alone show whether a chosen insurer's higher private-asset allocation caused better claims-paying resilience. Higher allocations may reflect changes in ownership, liabilities, capital position, reporting categories and investment opportunity. Without matched, consistently measured counterfactuals, the defensible conclusion concerns identified risk channels and necessary tests, rather than a precise aggregate welfare effect of alternative investments.

Scholarly and editorial verification of findings

This Dossier's empirical claims are confined to specified dates, populations and accessible primary reports. The NAIC year-end 2025 industry balance sheet and the EIOPA 2026 dashboard have deliberately different denominators; they are not recombined. Ownership-defined data are separated from investment-asset categories. Formulaic EU capital charges are distinguished from realized or forecasted equity losses. The 2027 EU revisions are identified as future-applicable at the research cutoff. All portfolio, leverage, volatility, correlation, surplus and liquidity examples are explicitly hypothetical, and their arithmetic can be reproduced from stated inputs. Research papers are interpreted within their samples and without claiming full replication of proprietary underlying datasets.

The study is limited by available published material: it is a documentary and analytical synthesis rather than a global census of insurer holdings, a licensed extraction of every proprietary fund dataset, an audit of a particular insurer's financial records, or external peer review. Peer-reviewed findings and public supervisory reports support the risk mechanisms and reporting context; they do not validate any bespoke numerical risk-calibration recommendation. Private fund contract terms, national supervisory rules and market liquidity may evolve, so a future release should check only genuinely time-sensitive statements against their then-current authoritative texts.

Conclusion

Equity and alternative investments place insurers in economic claims whose timing, pricing and legal availability differ from the cash obligations they have undertaken to fulfill. Listed equity exposes the balance sheet to frequent observable repricing; real estate and infrastructure equity embed operating, financing, regulatory and physical risks; private partnerships add commitment, valuation, fee, leverage and transfer restrictions; and some other alternatives introduce derivatives, funding and hidden look-through exposures. These instruments may contribute return and diversification under well-specified circumstances. Their assessment is complete only when the underlying financial exposure is linked to insurer-specific liabilities, cash availability, risk capital and accountable governance.

The conceptual distinction between an investment in private equity, a private-equity-owned insurer and the use of a PE-affiliated manager is essential to empirical interpretation. The NAIC's asset holdings and PE-ownership reports measure different populations, while EIOPA statistics use separate regional reporting definitions. Differences in insurance segment and liability design matter at least as much as headline investment categories. An asset-allocation comparison that ignores these distinctions cannot support a reliable solvency inference.

Valuation must also retain its correct meaning. Accounting fair value, model-estimated economic value and forced-sale proceeds are distinct measurements, each relevant to particular decisions. Delayed NAV updates can suppress measured covariance and volatility while economic exposure remains. Fund return measures such as TVPI, DPI, IRR and PME reveal different aspects of investment performance and carry assumptions that limit their interpretation. The analytical evidence therefore supports direct, dated cash-flow analysis and stress sensitivity alongside traditional portfolio statistics.

The hypothetical balance-sheet experiment demonstrates a general arithmetic truth: simultaneous asset repricing and liability revaluation determine economic surplus, while restricted cash and stressed asset sales can add incremental loss. The covariance experiment illustrates how dependence and volatility assumptions alter measured diversification. Neither experiment estimates an actual insurer's loss distribution. Their purpose is to make assumptions visible, replicable and subject to challenge.

International supervision requires attention to both quantitative and qualitative investment risks. IAIS ICP 15 and ICP 16 establish a common supervisory logic; the EU prudent person principle, equity risk modules and look-through provisions operationalize parts of that logic within Solvency II; the U.S. statutory and NAIC environment uses its own reporting and capital construction. Regulation informs rather than substitutes for enterprise analysis. The incoming EU changes applicable from January 30, 2027 require proper temporal identification.

The final substantive conclusion is that alternative-asset suitability is contingent on the insurer's capacity to understand, value, finance, govern and, when required, realize its investments while meeting policyholder obligations. A well-supported investment decision documents the economic exposure, identifies the states in which investment and insurance cash flows deteriorate together, and demonstrates resilience under executable funding arrangements. The appropriate scientific contribution of this Dossier is an evidence-grounded integration of existing insurance, financial-economic and supervisory knowledge, with explicit limitations and tested illustrative reasoning. No new authorial theory, empirical priority or regulatory prescription is asserted.

References

  • National Association of Insurance Commissioners (NAIC). (2025, August). U.S. insurance industry private equity-owned insurers: Year-end 2024 (Capital Markets Special Report). NAIC PE ownership report.

  • National Association of Insurance Commissioners (NAIC). (2026, May). U.S. insurance industry asset mix: Year-end 2025 (Capital Markets Special Report). NAIC 2025 asset-mix report.

  • International Association of Insurance Supervisors (IAIS). (2024, as accessed 2026). Insurance Core Principles and ComFrame (especially ICP 15 Investment and ICP 16 Enterprise Risk Management for Solvency Purposes). IAIS Core Principles.

  • European Parliament and Council. (2009, consolidated online text accessed November 2026). Directive 2009/138/EC (Solvency II), Article 132: Prudent person principle (EIOPA Single Rulebook). Solvency II Article 132.

  • European Insurance and Occupational Pensions Authority (EIOPA). (2026, October). Insurance Risk Dashboard: October 2026. EIOPA October 2026 dashboard.

  • Brunnermeier, Markus K., & Pedersen, Lasse Heje. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22(6), 2201–2238. Journal article.

  • European Commission. (2015, as amended and accessed 2026). Commission Delegated Regulation (EU) 2015/35, Article 169: Equity risk submodule (EIOPA Single Rulebook). Solvency II Article 169.

  • European Commission. (2015, as amended and accessed 2026). Commission Delegated Regulation (EU) 2015/35, Article 172: Symmetric adjustment of the equity capital charge (EIOPA Single Rulebook). Solvency II Article 172.

  • Organisation for Economic Co-operation and Development (OECD). (2020). Assessing institutional investment in infrastructure. In Green Infrastructure in the Decade for Delivery. OECD Publishing. OECD infrastructure study.

  • IFRS Foundation. (2026 access). IFRS 13 Fair Value Measurement: issued-standard overview. IFRS 13 overview.

  • Kaplan, Steven N., & Schoar, Antoinette. (2005). Private equity performance: Returns, persistence, and capital flows. Journal of Finance, 60(4), 1791–1823. Kaplan and Schoar article.

  • Harris, Robert S., Jenkinson, Tim, & Kaplan, Steven N. (2014). Private equity performance: What do we know? Journal of Finance, 69(5), 1851–1882. Harris and colleagues article.

  • Getmansky, Mila, Lo, Andrew W., & Makarov, Igor. (2003). An econometric model of serial correlation and illiquidity in hedge fund returns (NBER Working Paper 9571); later published in Journal of Financial Economics, 74(3), 529–609 (2004). NBER working paper.

  • Bank for International Settlements (BIS). (2024, September). Shifting landscapes: Life insurance and financial stability (BIS Quarterly Review). BIS insurance stability review.

  • IFRS Foundation. (2026 access). IFRS 9 Financial Instruments: issued-standard overview. IFRS 9 overview.

  • European Insurance and Occupational Pensions Authority (EIOPA). (2026, June). Financial Stability Report: June 2026. EIOPA June 2026 report.

  • National Association of Insurance Commissioners (NAIC). (2026 access). Private Equity (insurance topic and supervisory information). NAIC Private Equity overview.

  • European Commission. (2015, as amended and accessed 2026). Commission Delegated Regulation (EU) 2015/35, Article 84: Look-through approach (EIOPA Single Rulebook). Delegated Regulation Article 84.

  • European Commission. (2015, as amended and accessed 2026). Commission Delegated Regulation (EU) 2015/35, Article 168: General provisions (EIOPA Single Rulebook). Delegated Regulation Article 168.

  • European Commission. (2015, as amended and accessed 2026). Commission Delegated Regulation (EU) 2015/35, Article 171a: Long-term equity investments (EIOPA Single Rulebook). Delegated Regulation Article 171a.

  • European Insurance and Occupational Pensions Authority (EIOPA). (2026 access). Solvency II: Regulation and policy (including Solvency II review application timetable). EIOPA Solvency II overview.

  • European Insurance and Occupational Pensions Authority (EIOPA). (2026, June 24). EIOPA’s Financial Stability Report shows continued resilience of insurance amid geopolitical uncertainty (news release). EIOPA June 2026 release.