Two structural facts break normal financial statements.
Fact 1: the cost of goods sold is an estimate. When a general insurer closes its year, most of the claims it'll eventually pay on this year's policies haven't yet been reported, negotiated, or settled. So it books a reserve, its best estimate of ultimate claims, including IBNR ("incurred but not reported"). Reported profit = premiums earned − an opinion about the future − expenses. Book that opinion 5% too low and profit is overstated. You only find out the truth years later, in the reserve-development table you'll learn to read. It's the largest accrual you'll ever analyse, and management's own actuaries set it. Everything Phase 2 taught you about accrual discretion applies here with the dial turned to maximum.
Why does profitable life insurance look like a loss?
Fact 2: profitable growth looks like a loss (life insurance). A life insurer selling a 20-year policy incurs almost all its costs up front: first-year commission (routinely 15 to 35% of the first premium), medical underwriting, onboarding. It also has to set up a policy reserve straight away. Year-one outgo typically exceeds year-one premium margin. So under India's statutory accounts (Indian GAAP, IRDAI-format), writing profitable new business produces an accounting loss in year one. That loss is the new-business strain, and the profits emerge over the following 20 years.
Think about what that does to the numbers. A fast-growing, highly profitable life insurer reports depressed statutory profit. A shrinking one gushes reported profit, because old policies release margins and no new strain is taken. Statutory P/E is therefore close to useless for life insurers. To fix that, the industry built a parallel actuarial reporting stack. Embedded Value (EV) covers the existing book and Value of New Business (VNB) covers each year's sales, and both discount the future profits the accounts refuse to show.
Advertisement
How do the main reporting regimes compare?
The reporting regimes, side by side. Teach yourself the economics once, then note where each regime puts the truth:
| Regime | Who uses it | Core logic | Where the economics show up |
|---|---|---|---|
| Indian GAAP (IRDAI formats) | Indian insurers' statutory accounts | Premiums when due; acquisition costs expensed immediately; conservative reserves; separate Policyholders' and Shareholders' accounts (Revenue Account "Form A-RA" vs P&L) | Barely. Hence the EV/VNB actuarial supplements every listed life insurer publishes (and analysts live in the investor presentation, not the annual accounts) |
| IFRS 17 (effective 2023; Ind AS 117 notified in India but insurers stay on IRDAI formats until the regulator's roadmap lands) | Europe, most of Asia-Pacific incl. AIA; India eventually | Measure the contract liability at current fulfilment value; park the unearned profit in a Contractual Service Margin (CSM) released over the coverage period; onerous contracts hit P&L immediately | The CSM balance is, conceptually, a cousin of the value-of-in-force in EV: the profit store the old accounting hid. New-business CSM ≈ VNB in spirit |
| US GAAP (incl. LDTI, effective 2023) | US insurers | Long-duration liabilities re-measured with updated assumptions (discount-rate changes through OCI); market-risk benefits at fair value; DAC amortized | Partly. US analysts still lean on book value per share ex-AOCI vs ROE, adjusted operating EPS, and statutory data. EV disclosure is rare in the US (see price to embedded value) |
You don't need to account under any of these. You need to know which supplement to reach for. In India, that's the EV/VNB disclosures and the IRDAI public-disclosure forms. In the US, it's the 10-K loss-development table and statutory Schedule P for P&C, and book value ex-AOCI for life. In Europe and Asia, it's the IFRS 17 CSM, often alongside surviving EV reporting.