How to Analyze an Insurance Company Before Buying the Stock

Insurance looks like a boring business until you try to analyze one — and then you discover the normal stock playbook barely works. Pull up an insurer's income statement expecting the usual revenue-and-margins story, and you'll find float, reserves, and a "combined ratio" that nobody taught you in the standard how-to-research-a-stock tutorials. Insurers are financial companies wearing an ordinary-company disguise.

That's why learning how to analyze insurance stocks is its own skill. The value isn't in the product; it's in pricing risk correctly today and getting paid to hold other people's money until the claims come due. Miss that, and you'll either overrate a company that's quietly under-reserving or dismiss a great compounder because its P/E looks weird.

This guide walks the pieces that actually matter for an insurer — the combined ratio, float and investment income, loss reserves, book value and ROE, and why the whole industry gets valued on price-to-book. It's an educational framework, not a verdict machine. The goal is to know which numbers to pull and what they're telling you.

Why insurance stocks break the normal playbook#

Start with the mental model, because it reframes everything else. An insurer does two jobs at once, and they're almost separate businesses:

• Underwriting — collect premiums, promise to pay future claims, and try to price that promise so premiums exceed claims plus expenses. This is the "sell insurance" business.
• Investing — take the pile of premium money it hasn't paid out yet (the float) and invest it. This is, effectively, an asset-management business bolted onto the first one.

A great insurer makes money on both. A merely okay one loses a little on underwriting and makes it back on investments. A dangerous one loses money underwriting, papers over it with optimistic reserve estimates, and hopes the investment portfolio bails it out.

Because so much of an insurer's balance sheet is financial assets and financial liabilities — bonds on one side, claim obligations on the other — the usual tools (revenue growth, gross margin, EV/EBITDA) either don't apply cleanly or actively mislead. You need the industry's own dashboard. If you want the general-purpose process first and the insurance layer second, keep our 12-step research checklist open alongside this one.

The combined ratio: are they any good at their actual job#

The single most important number in property-and-casualty insurance is the combined ratio. It answers one question: for every dollar of premium the company earned, how many cents went out the door on claims and expenses?

The formula is straightforward:

Combined ratio = (incurred losses + underwriting expenses) ÷ earned premiums

It's usually broken into two parts you'll see quoted separately:

• Loss ratio = claims paid and reserved ÷ earned premiums
• Expense ratio = underwriting costs (commissions, overhead) ÷ earned premiums

Add them together and you get the combined ratio. Here's the part that trips people up: lower is better, and the magic line is 100%.

A worked example makes it concrete. Say an insurer earns $10 billion in premiums, pays and reserves $6.5 billion in losses (a 65% loss ratio), and spends $3 billion on expenses (a 30% expense ratio). Its combined ratio is 95% — meaning it kept about 5 cents of underwriting profit per premium dollar, before a single dollar of investment income. Flip the losses to $7.5 billion and the combined ratio becomes 105%: a five-cent underwriting loss per dollar, now relying on the investment side to turn an overall profit.

Look for: a combined ratio consistently below 100%, and — even more telling — one that's stable or improving across a full market cycle, including bad-catastrophe years. Be wary of: a ratio that only dips below 100% in calm years, or improvements driven by one-off items rather than better pricing and claims discipline. One good year proves nothing; insurance is a game of cycles.

Float: the free(ish) money that funds the business#

Here's the quietly brilliant part of the model. An insurer collects premiums now and pays claims later — sometimes years later, especially for long-tail lines like liability. In between, it holds and invests that money. That reservoir of "collected but not yet paid out" cash is called float, and it's the reason insurance can be such a powerful compounding machine.

Think of float as a loan from policyholders. If the insurer runs an underwriting profit (combined ratio below 100%), it's essentially being paid to borrow — a negative cost of funds — and then it gets to invest the borrowed money on top. Even at break-even underwriting, the float is effectively an interest-free loan whose investment returns flow to shareholders.

When you analyze an insurer, look at two things here:

• The size and growth of float relative to the company. Growing float from profitable underwriting is a genuine engine. Growing it by writing bad policies just to collect premium is borrowing trouble.
• How the float is invested. A conservative bond-heavy portfolio behaves very differently from one reaching for yield in riskier assets. Investment income is real, but it carries the risks of whatever it's invested in — and in a bad year, investment losses and insurance claims can hit at the same time.

Look for: float that grows on the back of disciplined, profitable underwriting, invested sensibly. Be wary of: an insurer whose earnings lean heavily on aggressive investing to compensate for chronically weak underwriting. That's not an insurance business; it's a leveraged bond fund with a claims department.

Loss reserves: the number they get to estimate#

This is where insurance analysis gets genuinely tricky — and where the most important judgment calls hide. When an insurer writes a policy, it sets aside loss reserves: its own estimate of what it will eventually have to pay on claims that have happened (or will happen) but aren't yet fully settled.

That word estimate is the whole game. Reserves are management's best guess about the future. Guess too low, and today's earnings look better than they really are — the bill just hasn't arrived. Guess too high, and profits are understated now but can be released later. Because reserves flow directly through the income statement, they're one of the most powerful levers a management team has over reported results, which is exactly why you should watch them closely.

Reserve releases and strengthening#

Over time, reality catches up with those estimates, and the company adjusts. You'll see this described as development:

• Favorable development (reserve release): claims came in lower than reserved, so the company releases the excess back into earnings. It boosts current profit.
• Adverse development (reserve strengthening): claims came in higher than reserved, so the company adds to reserves. It reduces current profit.

Neither is automatically good or bad, but the pattern matters enormously. A company that consistently releases modest reserves is showing a conservative reserving culture — it tends to over-provide and give money back. A company that keeps strengthening reserves is telling you its past estimates were too optimistic, which raises the question of whether its current numbers are too.

The trap: reserve releases can flatter a soft underwriting year, making the combined ratio look better than the underlying business earned. When you see a suspiciously good result, check whether it leaned on prior-year reserve releases. A durable insurer earns its combined ratio on current business; a fragile one borrows from the reserve cookie jar.

Look for: a long history of neutral-to-favorable development and transparent reserve disclosure. Be wary of: repeated adverse development, or a company whose profits depend on releasing reserves year after year. These are third-party facts you can pull from the filings — the annual 10-K and quarterly reports break reserve development out in detail.

Book value and ROE: how insurers actually compound#

For most companies, book value (assets minus liabilities, i.e. shareholders' equity) is an afterthought. For an insurer, it's close to the main event — because an insurer's assets and liabilities are mostly financial, and financial items sit on the balance sheet at or near fair value. Book value is a reasonably honest snapshot of what the company is worth, in a way it simply isn't for, say, a brand-driven consumer company whose real value is intangible.

That makes two metrics central:

• Book value per share, and its growth over time. A well-run insurer compounds book value per share year after year — underwriting profit plus investment income, minus dividends, all stacking into equity. Steady book-value growth is the clearest sign the compounding machine is working.
• Return on equity (ROE). Net income divided by shareholders' equity tells you how hard the company's capital is working. Because an insurer is essentially deploying its equity plus float into risk and investments, ROE is the headline efficiency number. If you're fuzzy on the mechanics, our return on equity explainer walks through what a "good" ROE means and how leverage can flatter it.

Look for: consistent growth in book value per share and a durable, mid-to-high ROE earned without extreme leverage or reckless investing. Be wary of: an ROE propped up by an aggressive investment portfolio or thin capital, and book value that stalls or slips despite reported profits (often a sign that unrealized investment losses or reserve issues are eating equity from the side).

Why insurers are valued on price-to-book#

Now the piece that confuses newcomers most: walk into insurance-land expecting P/E ratios and you'll find everyone quoting price-to-book (P/B) instead. There's a logical reason.

Because an insurer's book value is a meaningful, mostly-marked-to-market measure of its net worth, the market naturally asks: how much am I paying for a dollar of that net worth? That's exactly what P/B measures — price divided by book value per share. It's a cleaner yardstick than earnings for insurers, whose reported earnings swing with catastrophes, investment gains, and reserve adjustments from year to year. Book value is steadier, so the market anchors to it. Our price-to-book ratio guide covers the mechanics and the pitfalls in general terms.

The crucial link — and the thing that separates a naive P/B read from a useful one — is the relationship between P/B and ROE. An insurer that reliably earns a high return on its equity deserves to trade at a higher multiple of that equity than one that barely earns its cost of capital. Intuitively: if a company turns one dollar of book value into fifteen cents of profit every year, that dollar is worth more than a dollar that produces only six cents.

So the two numbers must be read together:

• A low P/B on a business earning weak, declining ROE may be cheap for a reason, not a bargain — a possible value trap.
• A higher P/B on a business compounding book value at a strong, durable ROE can be perfectly rational.

The real question is never "is the P/B low?" It's "does this P/B make sense given the ROE, the reserving quality, and the durability of the underwriting?" — which is only answerable after you've done the combined-ratio, float, and reserve work above. Valuation is the last step, not the first.

Look for: a P/B that's justified by the ROE and reserving track record you've already assessed. Be wary of: anchoring on a "cheap" P/B without asking why it's cheap, or paying a premium multiple for an ROE the company can't sustain through a hard market.

Putting it together#

Notice how interlocking this is. The combined ratio tells you if underwriting works. Float and investment income tell you how the held money compounds. Reserves tell you whether to believe the reported combined ratio in the first place. Book value and ROE measure the compounding. And P/B, read against ROE, tells you what you're paying for all of it. Skip any one and the others can mislead you — a beautiful combined ratio built on reserve releases, a great ROE built on a fragile bond portfolio, a low P/B hiding adverse development.

That web of cross-checks is exactly the kind of work an educational research desk is built to run. Valarn approaches a company like an insurer with up to about 25 specialist AI analysts across five categories — Core Research, Market Structure, Debate & Risk, Financial Quality, and Events/Sector & Macro — so the financial-quality analysts can scrutinize reserving and book value while the valuation analysts weigh P/B against ROE, and a dedicated risk committee stages a structured bull-versus-bear debate before anything is synthesized into a single neutral research view (Bullish, Cautious Bullish, Neutral, Cautious, or Bearish — never a buy/sell instruction). Every factual claim traces back to a filing or licensed source with an as-of date, and each report carries two 0–100 scores: a confidence score reflecting data quality (not a price prediction) and an agreement score measuring how much the analysts converge. Instead of a single price target, you get a Scenario Range — bear, base, and bull — with a Reference Price and a Risk Level.

You can see how that framework handles a real company in a full sample research report, or orient on a specific ticker from a company research page before you dig into filings yourself. For the underlying concepts, the Valarn glossary defines combined ratio, float, reserves, and the rest one term at a time. Insurance and banking share a lot of this financial-company DNA, so our companion guide on how to analyze bank stocks is a useful next read.

The bottom line#

Analyzing an insurer isn't harder than analyzing a normal company — it's just different, and the difference is the whole point. Get comfortable with the combined ratio (are they good at underwriting?), float (how does the held money compound?), reserves (can you trust the reported numbers?), book value and ROE (is capital compounding?), and P/B read against ROE (what am I paying for it?), and you'll understand an insurance business the way the industry actually understands itself.

None of this hands you a "buy." It hands you the right questions — and a way to tell a disciplined compounding machine from a leveraged bond fund hoping the claims stay small. When you're ready to see the process run end to end, explore a sample report or generate your own free research run and grade it against everything above.

Valarn is an educational research tool, not investment advice. It does not tell you to buy, sell, or hold anything, and nothing here is a recommendation or a promise of results. Always do your own research and consider consulting a licensed financial professional.