How to Value an Insurance Company

Introduction

Ask a candidate how to value an insurance company and you usually get one of two answers: a confident recitation of the DCF, or silence. Both fail for the same reason. An insurer collects cash today in exchange for promises it will settle over years or decades, invests that cash in the meantime, and books a liability whose true size nobody will know until the final claim closes. There is no factory, no inventory, no working capital cycle, and no meaningful EBITDA. What exists instead is a balance sheet whose liability side is the actual product.

The sector is unusually live right now, which is exactly why it shows up in interviews. US property and casualty insurers posted a 92.4% combined ratio in the first quarter of 2026, a sharp improvement from 99.2% a year earlier, with policyholders' surplus reaching $1.24 trillion, according to the Verisk and APCIA industry results release. Meanwhile Zurich agreed in March 2026 to buy Beazley for $10.9 billion, and Corebridge and Equitable announced a roughly $22 billion all-stock retirement and life merger, a figure that values the combined company rather than a price paid. If you are interviewing for a financial institutions group, insurance valuation is not an optional module.

This walkthrough covers why insurers break the standard toolkit, how property and casualty, life, and reinsurance businesses are each valued differently, and the two calculations you should be able to produce on paper: a decomposed combined ratio and a justified price to book from return on equity.

"Insurance" is a label covering economically distinct models. A short-tail auto insurer that pays most claims within a year has almost nothing in common with an annuity writer holding liabilities that run for forty years, and neither resembles a Bermudian reinsurer whose earnings depend on whether hurricanes make landfall. Before you pick a valuation method, diagnose which business you are looking at.

The common thread is that all three are balance sheet businesses, so equity value is the unit of account. The differences sit in how long the promises last and how confidently they can be measured, which is what pushes life insurers toward embedded value and leaves property and casualty insurers on book value.

Everything you learned first assumes a company that sells a product, incurs costs, and finances itself with some mix of debt and equity. Insurers scramble each of those assumptions at once, which is why the enterprise value bridge you use for an industrial simply cannot be built here. The same logic that governs how to value a bank applies, with a different set of moving parts.

Float Inverts the Business Model

An insurer is paid before it delivers. Premiums arrive up front, claims are paid later, and the money sitting in between belongs economically to policyholders but is invested for the shareholders' benefit. That pool is float, and it is the reason an insurer can earn an attractive return on equity while barely breaking even on the underwriting itself.

Float

Float is the pool of money an insurer holds between collecting premiums and paying out claims. It appears on the balance sheet as loss reserves and unearned premium, which are liabilities, but the insurer invests it and keeps the investment income. Float behaves like borrowed money whose cost is the insurer's underwriting result: an underwriting profit means the insurer is being paid to hold other people's money.

Float is also why the liability side of an insurer's balance sheet is not debt in any conventional sense. Subtracting reserves as "net debt" to reach equity value would be a category error, because those reserves are simultaneously the funding source and the obligation, and they are already reflected in book value.

Reserves Are Estimates, and Estimates Move Value

The largest liability on a property and casualty insurer's balance sheet is loss reserves, and it is an actuarial estimate rather than a contractual amount. Nobody knows the ultimate cost of a 2026 accident year until claims from that year finish settling, which for liability lines can take a decade. Change the reserve estimate and you change reported earnings, book value, and therefore the valuation, all without a single new policy being written.

That is a structural problem for valuation. The book value you anchor a multiple to is only as good as the reserve estimate embedded in it, which is why the quality of a company's reserving history carries real weight in how the market prices it.

No EBITDA, No Enterprise Value, No Unlevered DCF

Interest and investment income are core revenue for an insurer, not financing effects to be added back. Depreciation is trivial. Capital expenditure is essentially nonexistent, since growth consumes regulatory capital rather than plant and equipment. Strip all of that out and EBITDA has no economic content, so the enterprise value multiples you would normally reach for produce meaningless numbers.

Unlevered free cash flow fails for the same reason. There is no version of an insurer without its liabilities, because the liabilities are the funding, the product, and the risk in one. Valuation therefore runs entirely on equity: book value multiples, earnings multiples, embedded value for life businesses, and discounting at cost of equity rather than WACC.

Property and casualty is where most FIG interview questions land, because the metrics are concrete and the arithmetic is testable in a fifteen minute conversation. The entire model reduces to two questions: does the company make money underwriting risk, and does it make money investing the float?

From Premiums Written to Premiums Earned

Premiums written are what the insurer sold in the period. Premiums earned are the portion of those premiums whose coverage period has actually elapsed. Write a twelve month policy on October 1 and only three months of it is earned by December 31, with the rest sitting in unearned premium reserve. Every underwriting ratio uses earned premium in the denominator, so mixing the two is an easy way to produce a wrong answer under pressure.

Growth in net premiums written is the leading indicator, since it tells you what is coming. Net premiums earned is the recognition measure that drives the current period income statement. Industry premium growth slowed sharply to 2.9% in the first quarter of 2026, well below the mid to high single digit growth of the prior two years, which is the classic signature of a softening market.

Decomposing the Combined Ratio

The combined ratio is the single number that tells you whether the underwriting business works. It sums the cost of claims and the cost of running the business, both expressed as a share of earned premium.

The first term is the loss ratio, which captures claims paid plus reserves established plus loss adjustment expenses, the cost of investigating and settling claims. The second term is the expense ratio, covering commissions to brokers, premium taxes, and general operating costs. Loss ratio movements tell you about pricing and claims severity; expense ratio movements tell you about distribution economics and operating scale.

Combined Ratio

The combined ratio is an insurer's incurred losses plus underwriting expenses divided by net premiums earned, expressed as a percentage. A combined ratio below 100% means the insurer earned an underwriting profit, taking in more premium than it paid out in claims and costs. Above 100% means it lost money underwriting and depends on investment income to be profitable overall.

A sub-100 combined ratio is the headline every insurance management team wants, but it is worth being precise about what it means. It says the underwriting operation is self-funding: the company is being paid to hold float rather than paying for it. Chubb reported a P&C combined ratio of 81.2% in the fourth quarter of 2025, and Travelers posted a consolidated combined ratio of 83.6% in the second quarter of 2026, per its second quarter results release. Those are exceptional numbers, not typical ones.

A Worked Combined Ratio Calculation

Take an insurer with net premiums earned of $4.0 billion, incurred losses and loss adjustment expenses of $2.6 billion, and underwriting expenses of $1.1 billion.

The loss ratio is 65.0%, the expense ratio is 27.5%, and the combined ratio is 92.5%. The underwriting profit is the residual: $4.0 billion minus $2.6 billion minus $1.1 billion equals $300 million, which is exactly 7.5% of earned premium, the complement of the combined ratio.

Now layer in the float. If the company holds an investment portfolio generating $700 million of net investment income, pre-tax operating income is roughly $1.0 billion rather than $300 million. Investment income is more than double the underwriting profit, which is the normal shape of the business and the reason interest rate levels matter so much to insurance valuations.

Reserve Development: The Biggest Judgment Area

Reserve development is where insurance valuation stops being arithmetic and starts being judgment. Favorable development means prior estimates were too conservative and reserves are released into earnings. Adverse development means claims are settling above the reserve, forcing a strengthening charge that hits income and book value at once.

The reason analysts obsess over it is that development is both material and persistent in character. Companies with disciplined reserving tend to release reserves year after year; companies with weak reserving tend to strengthen repeatedly, often in the same problem lines. Travelers recorded $578 million of net favorable prior year reserve development in the second quarter of 2026, a meaningful share of the quarter's result.

Float and reserving together explain the framing Warren Buffett made famous at Berkshire Hathaway: if an insurer can underwrite at or below a 100 combined ratio, the float it holds is effectively free or better than free leverage, and the investment income on that float accrues entirely to shareholders. The corollary matters just as much. An insurer running consistently above 100 is paying for its float, and the higher the combined ratio, the more expensive that funding becomes.

Insurance metrics are where FIG candidates most often freeze: Work through combined ratio, book value, and return on equity questions with full worked answers on the practice platform, start practicing interview questions for free and find the gaps before an interviewer does.

With no EBITDA and volatile annual earnings, the balance sheet becomes the valuation anchor. Price to book is to insurers what EV/EBITDA is to industrials, and understanding why it works, and what makes it move, is the core of the interview answer.

Why Book Value Is the Right Anchor

An insurer's assets are mostly marketable securities carried at or near fair value, and its liabilities are estimates of future cash payments. That makes reported book value a genuine approximation of economic net worth in a way that an industrial company's historical cost balance sheet never is. It is not perfect, since reserves can be wrong and some fixed income is held at amortized cost, but it is close enough to serve as a valuation base.

Analysts often refine this into price to tangible book value, stripping goodwill and other intangibles from equity on the grounds that goodwill from past acquisitions cannot pay a claim. The wedge can be large: Chubb reported book value per share of $188.59 and tangible book value per share of $126.22 at the end of 2025, so the same share price implies two very different multiples depending on which denominator you use. Always confirm which one a comps page is quoting before comparing companies.

A Worked Price to Book Versus ROE Example

The link between the multiple and the return is a formula, not a vibe. The justified price to book multiple is:

Take an insurer earning a sustainable 14% return on equity, with a cost of equity of 9% and long-run book value growth of 3%:

On book value per share of $60, that justifies a share price of about $110. Now hold everything else constant and drop ROE to 8%, below the cost of equity. The multiple falls to 0.05 divided by 0.06, or 0.83 times book, implying roughly $50 per share. At an ROE of exactly 9%, matching the cost of equity, the multiple is precisely 1.0 times book.

That is the whole relationship in three data points. Book value barely moved; value moved a lot, and it moved because the spread between return on equity and cost of equity changed. Decomposing where that ROE comes from, which our guide to ROIC, ROE, and DuPont analysis breaks down step by step, is how you turn a multiple into an argument.

Life insurance breaks the property and casualty framework because the timing is entirely different. A policy sold today may generate profits for the next thirty years, and GAAP earnings in any single year reveal almost nothing about whether the business written that year was profitable. The industry's answer is embedded value.

Inside an Embedded Value Calculation

Embedded value estimates what the existing book of policies is worth today, ignoring any business the company has yet to write.

Adjusted net worth is the market value of assets backing capital in excess of what is required to support the liabilities. Value of in-force business is the present value of future after-tax profits expected from policies already sold, discounted at a risk-adjusted rate and reduced by the cost of holding required capital against those policies.

Embedded Value

Embedded value is a life insurer's adjusted net worth plus the present value of future profits from policies already in force. It is used instead of earnings-based valuation because a life policy generates profit over decades, so a single year of accounting income says little about economic value. Analysts often value life insurers as a multiple of embedded value, and separately value the new business the company expects to write.

The companion metric is value of new business, the present value of profits from policies sold during the year, measured at the point of sale. Value of new business is the closest thing life insurance has to organic growth: it shows whether the company is writing profitable business now, at current pricing and current interest rates. A life insurer with a large embedded value but declining value of new business is a run-off story dressed as a growth story.

The Long-Duration Liability Problem

Life liabilities can run for forty years, and the assets backing them rarely match that duration exactly. That mismatch is the central risk of the business. When interest rates fall, the present value of long-dated liabilities rises faster than the value of shorter assets, compressing economic capital even when nothing has happened to policyholder behavior. When rates rise, the reverse happens, but unrealized losses appear on the bond portfolio.

Embedded value is highly sensitive to the discount rate and the assumed investment return, which is why disclosures include sensitivity tables showing the effect of a 100 basis point rate move, a change in lapse rates, or a change in mortality assumptions. Reading those sensitivities is the fastest way to understand what a life insurer actually is: a leveraged bet on rates, spreads, and policyholder behavior wrapped in an insurance policy.

Why Life Insurers Trade at Persistent Discounts

Life insurers have long traded at lower price to book multiples than property and casualty peers, and often below book value outright. Several forces compound.

None of this makes life insurers uninvestable, and it is exactly why the sector generates so much M&A and reinsurance activity. It does mean that applying a property and casualty style price to book screen to a life insurer will hand you a list of apparent bargains that are mostly not bargains.

Insurance sits inside a broader FIG technical set: Download the comprehensive 160-page PDF, access the IB Interview Guide covering insurance, banks, and every core valuation framework interviewers test.

Insurers keep two sets of books, and confusing them is a fast way to get an answer wrong. Regulatory capital, not GAAP equity, determines what an insurer can actually do.

Two Sets of Books

GAAP financials are prepared for investors and emphasize matching revenue with expense over time. Statutory financials, filed with state regulators under NAIC rules, are prepared for solvency and are deliberately conservative: acquisition costs are expensed immediately rather than deferred, many assets are non-admitted and excluded entirely, and reserves are set on prescribed conservative bases.

The practical consequence is that statutory surplus is usually lower than GAAP equity, and dividends from an operating subsidiary to the holding company are constrained by statutory surplus rather than GAAP book value. When you model an insurance holding company's ability to fund buybacks or service debt, statutory dividend capacity is the binding constraint.

Risk-Based Capital and Rating Agency Models

The NAIC's risk-based capital framework sets a required capital amount reflecting asset risk, underwriting risk, credit risk, and interest rate risk, then compares total adjusted capital against it. The NAIC's risk-based capital framework triggers escalating regulatory intervention as the ratio deteriorates: a trend test between 200% and 300%, a company action level at 200%, a regulatory action level below 150%, authorized control below 100%, and mandatory control below 70%, at which point the regulator is obliged to take over. Well-capitalized insurers typically run several multiples of the trigger, so the binding constraint in practice is almost never the regulator.

Reinsurance is insurance for insurers, and it functions as both a risk transfer mechanism and a capital management tool. A primary insurer that cedes a share of its portfolio reduces the capital it must hold against that business, effectively renting someone else's balance sheet. Quota share treaties cede a fixed proportion of premium and losses, while excess of loss treaties cover the layer above a retention, which is how catastrophe protection is typically bought.

For a valuation, reinsurance usage changes the picture in two directions. It lowers earnings volatility and frees capital, which supports a higher multiple. It also transfers away profitable premium and introduces counterparty credit risk, which caps how much of the upside shareholders keep. Reinsurers themselves are valued much like property and casualty insurers, on price to tangible book against return on equity, with the added feature that their results swing on catastrophe outcomes. The sector earned a headline return on equity of roughly 17% to 18% in 2025, and January 2026 renewals saw the Guy Carpenter US property catastrophe rate-on-line index fall 12% as abundant capital pushed pricing down.

Catastrophe Modeling and Tail Risk

Because a single hurricane can consume a year of earnings, catastrophe exposure is modeled rather than estimated from history. Vendor models simulate large numbers of synthetic event years across a portfolio to produce loss distributions, and the outputs analysts care about are the probable maximum loss at specified return periods and the aggregate expected annual loss.

Recent experience has been unusually benign, which flatters current combined ratios. Global insured natural catastrophe losses reached $46 billion in the first half of 2026, down from $84 billion a year earlier and roughly 28% below the ten-year average. A candidate who notices that current underwriting margins are partly a function of light catastrophe activity, rather than assuming they are structural, is showing real sector judgment.

Deal activity in insurance runs on a handful of repeating logics, and knowing them lets you talk credibly about live transactions. Scale in distribution and data drives consolidation in personal lines. Specialty capability drives premium-priced acquisitions, which is the clearest read on Zurich paying $10.9 billion for Beazley to gain a Lloyd's platform and a leading cyber franchise, announced in Zurich's offer statement. Capital relief and legacy management drive the run-off market, where specialists acquire closed books of business to manage them to expiry.

Life and retirement consolidation is its own category, driven by the search for scale in asset management economics, as in the roughly $22 billion Corebridge and Equitable merger announced in March 2026. Elsewhere the sector saw Sompo acquire Aspen for $3.5 billion and Mapfre agree to buy Safety Insurance for $1.54 billion in July 2026.

The deal math looks different from an industrial transaction. Consideration is usually framed as a multiple of book or tangible book value, the key diligence question is reserve adequacy rather than revenue synergies, and acquirers care intensely about tangible book value dilution at close and the earn-back period required to recover it. A buyer paying two times tangible book for a target with questionable reserves is buying a liability it has not sized, which is why actuarial diligence dominates the process.

Interview questions in this space are predictable, because the sector's logic compresses into a few relationships. Expect to be asked why you cannot use EV/EBITDA, how the combined ratio decomposes, what a 95 combined ratio implies, what float is and why it matters, and why one insurer trades at 0.8 times book while another trades at 2.5 times.

The strongest answers connect the operating metric to the multiple rather than reciting either in isolation. Say that the combined ratio drives underwriting profit, that underwriting profit plus investment income on float drives return on equity, and that return on equity relative to cost of equity sets the price to book multiple. That chain is the entire sector in one sentence, and it works equally well as an answer for banks, insurers, and other balance sheet businesses. It is the same instinct that separates good from average answers in real estate valuation on FFO and NAV, where the accounting also has to be translated before the multiple makes sense.

If you can then add one live data point, that the industry ran a 92.4% combined ratio in the first quarter of 2026 helped by light catastrophes, or that reinsurance pricing fell 12% at January renewals, you have moved from textbook to informed. For context on the coverage teams that do this work, our overview of the financial institutions group covers what FIG bankers actually advise on.

Insurance valuation looks intimidating from the outside because the vocabulary is unfamiliar, but the underlying structure is simpler than a manufacturing DCF. Diagnose the business model first, since property and casualty, life, and reinsurance genuinely require different methods. Anchor on the balance sheet, because for a portfolio of financial assets and estimated liabilities, book value means something. Then explain the multiple with the return, because price to book is nothing more than a statement about whether the company earns more than its cost of equity.

The 2026 backdrop makes those relationships easy to illustrate. Combined ratios are unusually strong, catastrophe losses are running well below average, reinsurance pricing is falling as capital floods in, and consolidation is accelerating across both specialty and retirement. Each of those facts maps directly onto one of the drivers above, which means a candidate who understands the framework can read a headline and immediately say what it does to value. That is what a FIG interviewer is actually testing.