- What the number actually is
- The most numerous claim type owns the number
- Three things a settled claim does not tell you
- Which claims are in the denominator
- The ratio partly measures the age of the book
- What is published, and what is convention
- The questions that carry more information
- What to do with the number you have
- Common questions
What the number actually is
A claim settlement ratio is one division: the number of claims an insurer settled in a period, divided by the number of claims it had to deal with. Both terms are counts. Neither is a sum of money, and neither carries a clock.
Everything the metric is good for, and every way it misleads, follows from that construction. It is worth being unusually literal about it, because the ratio is quoted far more often than it is defined.
Three consequences fall straight out of counting. The claim type an insurer receives most of will dominate the number. A claim settled for a fraction of what was asked increments the numerator exactly as much as one paid in full. And a claim settled after fourteen months is indistinguishable, in the ratio, from one settled in a week.
None of that makes the figure dishonest. It makes it a measure of a different thing than most readers believe they are reading.
The most numerous claim type owns the number
Take a general insurer with two books. Round numbers, chosen to make the arithmetic visible — these are an illustration, not data about anyone.
It receives 99,000 motor claims in the year. Most are small, most are documented by a surveyor who saw the vehicle, and 98,000 are settled. It also receives 1,000 health claims, which are longer, more expensive and more contested. It settles 600 of them.
Total settled: 98,600. Total received: 100,000. The published ratio is 98.6%. The rejection rate on the health book is 40%. Both figures are true, and only one of them is on the marketing page.
The mechanism is not fraud or even spin. A claim settlement ratio is an average weighted by claim counts, and claim counts are decided by how often each kind of loss happens — which has nothing to do with which kind of loss you bought the policy for. A cheap, frequent, easily verified claim type carries the number. The rare, expensive, contested one you are actually insuring against barely registers in it.
Run the same arithmetic backwards and the reason established insurers cluster becomes obvious. Once the numerous claim type is settled routinely — and it usually is, because it is small and documented — the ratio is pinned near the top by construction. The variation between insurers gets compressed into the last percentage point or two, which is why comparisons between them read as noise.
Three things a settled claim does not tell you
“Settled” is a single flag on an outcome that has at least three dimensions. Here is what the flag drops.
How much of it was paid. A hospital bill of ₹4.2 lakh settled at ₹1.8 lakh is a settled claim, and the ratio records it identically to one paid in full. The gap between those two numbers is made up of deductions that operate entirely inside the wording: room-rent capping and the proportionate deduction it triggers, disease-wise sub-limits, and the list of consumables an insurer does not pay for. Call it the settled shortfall. It is the single largest thing this metric cannot see, and on a health policy it is usually discovered at the discharge desk rather than before.
How long it took. Turnaround is regulated separately, and the regulation is specific where the ratio is silent: a death claim carries a settlement window of 15 days, or 45 days where an investigation is warranted, and on a cashless hospitalisation the insurer has one hour to authorise, and three hours to grant final discharge authorisation. Those are obligations about time. Compliance with them is not what the ratio measures, and an insurer can be slow and still count every one of those claims as settled.
Whether the rejections were fair. This cuts both ways, and honesty requires saying so. A claim rejected for a fabricated hospitalisation and a claim rejected on a paperwork technicality both sit in the same bucket — not settled. So a lower ratio is not evidence of bad faith, and a higher one is not evidence of good faith. Compiling the ratio involves no reading of the rejected files; that happens only where a policyholder escalates one, and only for that one. The ratio counts outcomes. It does not judge them.
Which claims are in the denominator
The numerator is reasonably well behaved. The denominator is where two figures that look comparable stop being comparable.
| Question about the figure | Why it changes the answer |
|---|---|
| Count basis or amount basis? | Claims settled ÷ claims received is a count. Amount paid ÷ amount claimed is a different number entirely, and it is the one that would expose the settled shortfall |
| Which claim types? | A single insurer's motor, health and travel books behave nothing alike. An all-lines figure is dominated by whichever is most numerous |
| Claims received, or received plus pending from last year? | Carrying forward the previous period's unsettled claims into the denominator lowers the ratio; excluding them raises it |
| Where do withdrawn and abandoned claims sit? | A claim the insured never completed the documentation for is neither settled nor rejected on merits, and its treatment moves the number |
| What period, and how large is the book? | On a book of 400 death claims a year, one claim moves the ratio 0.25 percentage points. On a book of 40,000 the same claim moves it 0.0025 percentage points — a hundredth as much |
That last row is worth sitting with, because it is arithmetic rather than opinion: one claim divided by 400 is a quarter of a percentage point. A small insurer's ratio is a noisy statistic, and a year in which it moves a point may contain no information about the insurer at all.
For life insurers there is a further question to settle before comparing anything: a life company also pays maturity and survival benefits, which are a contracted date arriving rather than a claim to be assessed. Whether those sit in the denominator alongside death claims changes the character of the figure completely. If they do, the arithmetic of the previous section applies with more force, not less.
The ratio partly measures the age of the book
Here is the mechanism that does most to explain why these numbers differ between insurers, and it is almost never mentioned alongside them.
Indian law puts a time limit on how long a life insurer may reopen the basis of a policy. Section 45 of the Insurance Act 1938 sets a window — three years — inside which a life policy can be called into question on stated grounds. Once it closes, that route closes with it. The exact grounds, and the date the clock starts running from, are matters for the section itself rather than for any summary including this one — it is short, and it is worth reading in the original. The shape of the rule is not in doubt: early claims are contestable and late ones are not.
Now think about what that does to a ratio. A death claim on a policy sold eleven months ago and a death claim on a policy sold in 2009 are not the same event. The first gets investigated; the second largely cannot be. So the share of an insurer's claims that are even capable of being repudiated depends on how much of its book is still young.
Which produces this: an insurer growing quickly, writing a large volume of new policies, has a higher proportion of claims inside the contestable window and will show a lower settlement ratio than a slower-growing competitor behaving identically. An insurer whose book has aged shows a higher one for the same reason. The number moves with the age distribution of the policies, not only with conduct.
Be careful about how far that goes. The direction is arithmetic and not arguable. The size of the effect is a different question, and answering it would need each insurer's claims split by policy age, which is not something a buyer has. So this is a reason two ratios may not be comparable — not a correction you can apply to one and then rank the results.
Two things follow that a buyer can act on. Comparing a young insurer's ratio against an old one's is comparing books, not behaviour. And the position that actually matters is not the insurer's average but where your policy will sit: a claim made in year two is assessed under a different rule from a claim made in year twelve. What decides the year-two claim is what you wrote on the proposal form — which is the one input in this entire system you control.
What is published, and what is convention
Separating these two is most of the work, because the second dresses as the first.
Published. Insurers file claim data in their regulatory returns and disclosures, and that filing is where an insurer-wise figure ultimately has to come from. What the process produces is a count, compiled by the insurer, on whatever definition the return specifies. It is a disclosure of an outcome tally, not an audit of individual claim files — no step in it re-examines a rejection to decide whether it was justified. So if a figure is quoted at you, the two things to ask for are which filing it came from and which period it covers. A ratio with neither attached is a marketing number.
Regulated. What regulation fixes is conduct and timing rather than any minimum ratio — the settlement windows above, the grievance machinery, and a route to the Insurance Ombudsman, which can hear complaints up to ₹50 lakh. Those are obligations with consequences. A ratio is a disclosure.
Convention. The thresholds that circulate — the idea that some round number separates a good insurer from a poor one — are comparison-site practice rather than a standard anyone set. You will find no such pass mark in this article, because a pass mark means nothing without an authority behind it. When a source quotes one, the question that settles it is whose rule it is: name the regulation, or it is a convention wearing a rule's clothes.
The practical rule that falls out: two ratios are comparable only if they cover the same period, the same claim types, the same basis and the same treatment of pending claims. A figure quoted without those four attached is not a number you can do anything with, however precise the decimal looks.
The questions that carry more information
None of this makes the ratio useless. It makes it a screen with one job — flagging an outlier far below the pack — and a poor tool for choosing between insurers that are close. These are the questions that do more work, roughly in order of how much.
- What is the ratio for the claim type I would actually make? The health line of a composite insurer, the individual death-claim line of a life insurer. The all-lines figure is the one distorted by count weighting.
- Is there an amount-basis figure? Amount paid against amount claimed is the number that would reveal the settled shortfall. Where it is not available, the wording is the substitute: read the sub-limits, the room-rent clause and the non-payables list, because those are what create the gap.
- How long, not just how many. The regulatory windows exist; ask what proportion of claims settle inside them.
- What do the complaints say? Grievances per thousand policies rather than raw counts — a large insurer will always have more complaints in absolute terms, so an unnormalised count reads as an argument for buying from the smallest insurer available, which is not what it shows. How they were disposed, and how many went as far as the Ombudsman, are a signal about handling that a settlement count cannot carry: unlike a claim, each grievance represents someone unhappy enough to escalate.
- Who handles the claim? An in-house team or a third-party administrator, and for health cover, whether the hospitals you would realistically use are in the network. A cashless authorisation that has to travel through an extra party is a different experience from one that does not.
- What does the wording say? For anything that pays on a defined event rather than a bill — critical illness and disability cover above all — the definition in the policy decides the claim, and it is readable before you buy rather than after diagnosis.
Notice that four of those six are about your own policy and only two are about the insurer. That ordering is the article's argument in miniature.
What to do with the number you have
Use it as a floor, not as a ranking. An insurer sitting well below the pack is telling you something, and it costs nothing to notice. Two insurers a fraction of a point apart are telling you nothing you can act on, and choosing between them on that decimal is one of the more expensive ways to feel like you did your research, because the hours go into a number that will not affect your claim while the proposal form and the policy wording, which will, go unread.
Behind that is the trade-off every easy metric makes. A claim settlement ratio is public, single-figure and comparable across insurers, which is exactly why it is quoted. What you pay for that convenience is everything the count discards: the amount, the delay, the claim type and the merits.
The reason any of this matters is the reason insurance exists at all: you are buying a payment at the worst moment of your financial life. The metric that dominates the buying conversation is one that cannot see how much arrived, or when. That is worth knowing before it is your claim.
FNOTrader does not sell insurance, is not affiliated with any insurer, and is not a SEBI-registered investment adviser. This explains how a published metric is constructed; it is not advice on which policy or insurer to buy. Policy wordings differ and are revised, and regulatory timelines change — read your own schedule and the current regulations before deciding anything.
Common questions
What is the claim settlement ratio?
The number of claims an insurer settled in a period divided by the number of claims it had to deal with. Both terms are counts of claims, so the figure carries no information about how much of each claim was paid or how long settlement took.
Does a high claim settlement ratio mean my claim will be paid?
Not on its own. The ratio is an average weighted by claim counts, so it is dominated by whichever claim type the insurer receives most of — typically small, easily verified ones. An insurer can settle almost every claim in its largest book and still contest a high proportion of the smaller, more expensive book you would be claiming under.
Does the ratio count a claim that was only partly paid?
Yes, and identically to one paid in full. A hospital bill of ₹4.2 lakh settled at ₹1.8 lakh increments the numerator by exactly one. The shortfall comes from room-rent capping and the proportionate deduction it triggers, disease-wise sub-limits and non-payable items — none of which the ratio can see.
Why do most insurers' claim settlement ratios look so similar?
Because the numerous, routine claim types are settled routinely, which pins the aggregate near the top by construction. Real differences in how an insurer handles contested claims get compressed into the last percentage point or two, where they are hard to distinguish from noise — particularly on a small book, where a single claim can move the figure a quarter of a percentage point.
Is there a minimum claim settlement ratio set by the regulator?
None that anyone quoting one can point to a regulation for. The round numbers that circulate as pass marks are comparison-site convention rather than a rule, which is why no threshold appears in this article — a pass mark means nothing without an authority behind it. What regulation does fix is conduct and timing: settlement windows, cashless authorisation windows, the grievance machinery and the Ombudsman route. If a source quotes a floor, ask it to name the regulation.
Why would a fast-growing insurer show a lower ratio?
Because a life policy can be called into question only inside a limited window from the start, and a young book contains a higher share of claims still inside it. Two insurers behaving identically will report different ratios if one has been writing new policies quickly and the other has an older book — the number moves with the age distribution of the policies, not only with conduct.
What should I look at instead of the claim settlement ratio?
The ratio for the specific claim type you would make rather than the all-lines figure; an amount-basis figure if one exists; what proportion of claims settle inside the regulatory windows; grievance and Ombudsman data, which records policyholders unhappy enough to escalate; who administers claims; and the policy wording, which is what actually decides the claim.
Is the claim settlement ratio audited?
It comes out of insurers' own regulatory returns and disclosures, as a count compiled on whatever definition the return specifies. Compiling it involves no re-reading of rejected claims to judge whether the rejection was justified — that happens only where a policyholder escalates one. Which is why a low ratio is not proof of bad faith and a high one is not proof of good faith.
Continue reading
More in Insurance · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article