How to Value an Insurance Company: The Guide for IB Interviews

You've got your DCF down cold. You can walk through EV/EBITDA in your sleep. Then an interviewer asks, "How would you value an insurance company?" and suddenly none of that works.

That's because insurers break almost every rule of standard valuation. The good news: once you understand why, the right approach is simple and very interview-friendly. Let's walk through it, with worked numbers you can actually use.

Why Insurers Are Different

With a normal company, the balance sheet supports the business. With an insurer, the balance sheet is the business.

Here's how an insurer makes money:

• It collects premiums upfront from policyholders.
• It pays claims later, sometimes years later.
• In between, it invests that money (called the float) mostly in bonds.

So an insurer earns from two engines: underwriting (premiums minus claims and expenses) and investing (returns on the float and its own capital).

Why EV/EBITDA and a Standard DCF Fall Apart

Here's the short version you can give in an interview:

• Interest is an operating item, not a financing one. Investment income is a core revenue source, so stripping it out (as EBITDA does) removes half the business.
• "Debt" is hard to define. The biggest liabilities are reserves owed to policyholders. Those are operating liabilities, not financing, so enterprise value gets murky.
• Free cash flow isn't meaningful. Capex is small, working capital isn't a clean concept, and cash is tied up by regulators.

The fix: value the equity directly. Every method below focuses on equity value, not enterprise value.

The Key Metrics You Need to Know

1. Combined Ratio (P&C Insurers)

This is the single most important measure of underwriting quality for property and casualty insurers.

Combined ratio = Loss ratio + Expense ratio

• Loss ratio = Claims and claim-handling costs (loss adjustment expenses) ÷ Premiums earned
• Expense ratio = Underwriting expenses (commissions, admin) ÷ Premiums

Below 100% means the insurer makes money on underwriting itself. Above 100% means it loses money on underwriting and relies on investment income to stay profitable.

2. Return on Equity (ROE)

ROE = Net income ÷ Shareholders' equity

For insurers, ROE is the engine of value. Hold that thought, because it's the link to everything in the valuation section.

3. Book Value per Share

Since assets are mostly financial (bonds, stocks, cash) and marked at or near market value, book value is a decent proxy for what the business is worth on a liquidation basis. That makes it far more meaningful than it is for, say, a software company.

Worked Example: Building an Insurer's Earnings

Let's take a simple P&C insurer, "SafeHarbor Insurance" (in $ millions):

Now the ratios:

• Loss ratio: 650 ÷ 1,000 = 65%
• Expense ratio: 300 ÷ 1,000 = 30%
• Combined ratio: 95%, so a 5% underwriting margin
• ROE: 120 ÷ 1,000 = 12%

Notice that investment income ($100m) is twice the underwriting profit ($50m). That's typical, and it's exactly why you can't ignore interest when valuing an insurer.

How to Value an Insurer: The Three Core Methods

Method 1: Price to Book (P/BV) and the ROE Link

This is the go-to multiple. Insurers trade on P/BV (and often P/TBV, which strips out goodwill and intangibles), alongside P/E.

But here's the part that impresses interviewers: P/BV isn't arbitrary. It's driven by ROE versus cost of equity.

Justified P/BV = (ROE − g) ÷ (Cost of equity − g)

For SafeHarbor, assume:

• ROE = 12%
• Cost of equity = 10%
• Long-term growth (g) = 4%

Justified P/BV = (12% − 4%) ÷ (10% − 4%) = 8% ÷ 6% = 1.33x

With $1,000m of book value, that gives an equity value of about $1,333m.

The intuition is the key takeaway:

• ROE above cost of equity → trades above book (it creates value)
• ROE equal to cost of equity → trades at book (1.0x)
• ROE below cost of equity → trades below book (it destroys value)

If an interviewer asks why one insurer trades at 2.0x book and another at 0.8x, this is your answer.

Method 2: Dividend Discount Model (DDM)

Since FCF doesn't work, you discount what shareholders actually receive: dividends.

The catch is that dividends are constrained by regulatory capital. An insurer must hold enough capital to stay solvent (under risk-based capital rules in the US, or Solvency II in Europe). It can only pay out what's left after keeping its capital at the required level.

For SafeHarbor, to grow book value by 4% a year, it must retain earnings:

• Retention ratio = g ÷ ROE = 4% ÷ 12% = 33%
• Payout ratio = 67%
• Dividend = $120m × 67% = $80m

Using a simple Gordon Growth DDM:

Equity value = $80m ÷ (10% − 4%) = $1,333m

Same answer as Method 1. That's not a coincidence.

Method 3: Excess Returns Model

This one frames value as book value plus the value of earning more than your cost of equity.

Equity value = Book value + PV of excess returns

For SafeHarbor:

• Excess return = (ROE − Cost of equity) × Book = (12% − 10%) × $1,000m = $20m per year
• PV of excess returns, growing at 4% = $20m ÷ (10% − 4%) = $333m
• Equity value = $1,000m + $333m = $1,333m

Once again, $1,333m.

This is a great point to make in an interview: under consistent assumptions, justified P/BV, the DDM and the excess returns model are just three ways of expressing the same economics. They all come down to how much ROE beats cost of equity, and for how long.

Adjustments That Separate Good Answers From Great Ones

Strip Out Unrealized Bond Gains and Losses

Insurers hold huge bond portfolios. When rates move, the market value of those bonds swings, and under US GAAP much of that flows into AOCI (accumulated other comprehensive income) on the balance sheet.

That can make book value jump around even when nothing changed in the business. So analysts often look at book value excluding AOCI to get a cleaner, more stable figure.

Question Reserve Adequacy

Reserves are management's estimate of future claims. If they're too low, book value and earnings are overstated today, and a painful reserve charge comes later.

A sharp answer: "I'd look at reserve development history. If prior-year reserves keep needing to be increased, I'd haircut book value."

Normalize for Catastrophes

One hurricane year can blow up a combined ratio. When comparing insurers or projecting earnings, analysts typically normalize catastrophe losses to a long-run average rather than using one unusual year.

What About Life Insurers?

Life insurers write long-dated policies (life cover, annuities) where profits emerge over decades. Accounting profit in any single year tells you little.

So they're often valued using embedded value (EV):

Embedded value = Adjusted net asset value + Value of in-force business

• Adjusted net asset value is roughly the shareholders' capital.
• Value of in-force business is the present value of future profits from policies already on the books.

Analysts then use Price to Embedded Value (P/EV) much like P/BV. Under IFRS 17 (the accounting standard used outside the US), there's also the contractual service margin (CSM), which represents unearned profit on existing contracts. Analysts frequently look at book value plus after-tax CSM as a fuller picture of equity.

You won't usually need deep detail here for a generalist interview. Knowing that life insurers use embedded value and P&C insurers lean on P/BV, ROE and the combined ratio is plenty.

Common Interview Questions (and How to Answer Them)

"Why don't you use EV/EBITDA for an insurer?" Because investment income is core to the business, so interest isn't a financing item, and policyholder reserves make "debt" hard to separate from operations. We value equity directly instead.

"What multiples would you use?" P/BV (or P/TBV) and P/E for P&C insurers, P/EV for life insurers, always cross-checked against ROE.

"Why does one insurer trade at a higher P/BV than another?" Higher sustainable ROE relative to cost of equity, and possibly higher growth. Justified P/BV = (ROE − g) ÷ (COE − g).

"What does a combined ratio of 105% mean?" The insurer pays out $1.05 in claims and expenses for every $1.00 of premium, so it loses money on underwriting and depends on investment income to be profitable.

"Why use a DDM instead of a DCF?" Free cash flow isn't meaningful for an insurer, and regulatory capital rules determine how much can actually be paid out, so dividends are the best measure of cash returned to shareholders.

The Bottom Line

If you remember one thing, make it this: an insurer's value comes down to how much ROE beats its cost of equity. P/BV, the DDM and the excess returns model are just different lenses on that one idea.

Nail the combined ratio, explain why EV/EBITDA doesn't work, and walk through a quick justified P/BV calculation. Do that, and a question that trips up most candidates becomes one of your strongest answers.