Skip to content
Risk & Psychology

The year in which nothing went wrong

Eleven renewals, no claim, and a household doing the sum out loud at the dining table. The arithmetic they are doing is correct and the question it answers is the wrong one — because a protection decision is designed around the outcome that has just happened for the eleventh time.

Risk & PsychologyIntermediate13 min read
Browse Risk & Psychology(130)

The renewal notice arrives in the same fortnight as the school fee, the car service and a wedding in Nagpur. Somebody at the table does the sum out loud: eleven years, ₹14,000 a year, one and a half lakh rupees gone, and not one rupee has ever come back. Nobody in the house has been ill. Nobody has died. The policy has, in the plainest sense available, done nothing at all for eleven years, and there is a fixed deposit in the same bank that at least shows a number going up. The arithmetic is right. The conclusion that usually follows it — that the money has been wasted and this is the year to stop — rests on a comparison the household has not noticed it is making: it is comparing what the policy produced against what a savings product would have produced, when the policy was never bought to produce anything. It was bought to make one particular future survivable, and the eleven quiet years are not evidence against it. They are the outcome it was designed around.

Think of it like this
The inverter in the corner of the room

The inverter cost money, takes up space by the shoe rack, needs distilled water topped up, and in a good year the power does not go for more than a few minutes. Judged as an appliance — what did it give me this year — it is the worst thing in the house. The fridge chilled things every day. The inverter chilled nothing. Nobody concludes from this that the fridge was the better purchase, because everybody understands intuitively that the inverter is not competing with the fridge.

In the market

A protection decision is not competing with the growing thing next to it. It is bought against a state of the world, and it is supposed to sit there producing nothing in every year that state does not arrive. The comparison that feels natural — this line went up, that line did nothing — is the comparison that has no meaning here.

Why the feeling gets stronger the longer it works

There is a cruel structure to this. Every quiet year adds a premium to the visible column and adds nothing to the invisible one, so the sense of waste does not stay flat — it accumulates. In year two the sum is ₹28,000 and easy to shrug off. In year eleven it is ₹1,54,000 and it feels like a serious amount of money, because it is. Meanwhile the actual need has usually been moving in the same direction: over those eleven years the household has taken a home loan, added a second child and reduced the number of earners from two to one. So the pressure to cancel is largest at roughly the moment the cover matters most, and it is largest for the arithmetical reason that the protection has been working, uninterrupted, for eleven years.

The two numbers that decide it, and the one that does not

A loss has two properties and people reliably worry about the wrong one. Frequency is how often it happens. Severity is what it costs when it does. Attention follows frequency, because frequency is what you experience — a phone screen cracks, a bike is scratched, a bill is higher than expected, and these are the losses that are vivid. Protection, however, is only ever worth its price against severity, and specifically against the severity you could not absorb. Which is why the extended warranty offered at the till and a large, dull hospitalisation policy feel like the same kind of purchase and sit in opposite corners of the same grid.

Happens oftenHappens rarely
Costs littlePay it and forget it. There is nothing here worth insuring — the premium plus the insurer's costs will exceed what you would claim, and the loss is one your monthly budget already absorbsIgnore it entirely. This is the quadrant that produces the most cluttered financial lives, because each individual cover sounds sensible and none of them changes anything
Costs a great dealThis is not an insurable risk, it is a lifestyle or a business problem. No insurer will price it for long, and if one does, the premium will approach the lossThe only quadrant where transferring the risk to somebody else is doing real work. Rare enough that the premium is small relative to the loss, severe enough that you cannot fund it yourself
The grid is a way of thinking, not a product list. What it settles is not whether to hold a particular cover but which question to ask about it: could this household absorb the loss out of what it already has, and how large is the loss relative to the price of not having to.

The break-even that almost nobody works out

You can put a number on the trade, and it is worth doing once because the number is usually much smaller than people expect. Ignoring the time value of money on both sides, a pure protection contract is a fair bet if the chance of the covered event over the whole term is greater than the total premiums divided by the amount paid out.

Break-even probability = Total premiums over the term ÷ Amount payable
Total premiums
The level annual premium multiplied by the number of years the cover runs
Amount payable
What the contract pays if the covered event occurs — the sum assured, in a life policy

Example: ₹14,000 a year for 30 years against ₹1 crore of cover: ₹4,20,000 ÷ ₹1,00,00,000 = 4.2%. Above a 4.2% chance across three decades the contract is favourable in plain rupees; below it, unfavourable.

Worked example
The household's sum, and the sum it did not do
Illustrative — a level ₹14,000 premium against ₹1 crore of cover, eleven years in
What has gone outEleven renewals at ₹14,000. This is the number said out loud at the table, and it is entirely accurate₹1,54,000
What has come backNo claim. Measured as a savings product the return is minus one hundred per cent, and it will be minus one hundred per cent in most years the household ever hasNil
What is standing behind it today₹1,00,00,000 ÷ ₹1,54,000 ≈ 65. That ratio is the thing actually being bought, and it does not appear anywhere on a statement₹1 crore, about 65 times the premiums paid so far
The break-even over the full term₹4,20,000 of premiums against ₹1,00,00,000 of cover. The contract is favourable in rupee terms if the chance of the event across the thirty years exceeds roughly one in twenty-fourAbout 4.2%
What that calculation quietly leaves outIt ignores discounting, and the two sides are not affected alike: premiums go out from year one onward, while the ₹1 crore is a single sum that mostly falls decades away and is worth a fraction of that in today's money — so on balance the plain-rupee sum flatters the contract. It ignores the insurer's costs and margin, which have to come from somewhere and come from the buyer. And it assumes all thirty premiums are paid even where an early claim would have stopped them, which runs the other wayThree simplifications, and they do not all point the same way
So what is the honest expected valueYou do not need to settle the simplifications to know the sign. An insurer prices a book so that what comes in covers what goes out plus its own costs, and one that priced at break-even would not survive. Every buyer of protection is knowingly accepting a slightly bad bet in rupeesNegative, by construction
That last line is the one worth carrying out of the lesson, because it is usually taught as a scandal and it is not one. Protection is negative expected value in rupees and can still be the right decision, for a reason the rupee figure cannot express: losing ₹1 crore does not hurt a hundred times as much as losing ₹1 lakh, it hurts in a different category altogether, and it can end the plan rather than dent it. You are not buying a favourable average. You are paying a known small amount to remove an outcome you could not come back from. Which also means the reverse is true, and worth saying plainly so the lesson is not read as a recommendation: where the loss is one the household could genuinely absorb, the same negative expected value is simply a bad deal with nothing bought in exchange — and a great many things sold as protection in India sit in exactly that position.

None of this settles the product question — how much life cover a household needs, why bundling protection with investment does both badly, and what a health policy's exclusions actually say are all worked through in the Market Basics track and are not repeated here. What this lesson is trying to change is narrower and comes earlier: the instinct that a run of quiet years is information about whether the decision was sound. It is not. It is information about which years you happened to live through.

Check yourself

A household holds a pure protection policy costing ₹14,000 a year against ₹1 crore of cover, and has made no claim in eleven years. Which statement about those eleven years is correct?

Simple bhasha mein
Inverter ne is saal kuch nahi diya

Gyarah saal se ₹14,000 har saal — kul ₹1,54,000 nikal chuke aur wapas ek rupaya nahi aaya. Hisaab bilkul sahi hai. Par peeche jo khada hai woh ₹1 crore hai, yaani jitna abhi tak diya uska lagbhag 65 guna — aur woh number kisi statement pe nahi chhapta. Inverter se aap yeh nahi poochte ki is saal kitna kamaya; poochte ho ki bijli gayi thi tab chala ya nahi. Gyarah shaant saal is faisle ki galti nahi, uska design hain. Sawaal sirf ek hai: woh cheez ho jaaye toh ghar us kharche ko apne bal pe utha lega ya poora plan hi khatam ho jaayega.

What to remember
  • The ordinary experience of a correct protection decision and a wasteful one is identical: you pay, and nothing happens.
  • The felt waste accumulates every quiet year, so the pressure to cancel is usually greatest after the cover has worked longest.
  • Attention follows frequency; protection is only worth its price against severity you could not absorb.
  • Break-even probability is total premiums divided by the amount payable, and it is usually a small number.
  • Protection is negative expected value in rupees by construction, and can still be right — because a loss that ends the plan is not just a bigger version of one that dents it.
You reached the endMark it done and keep your streak going.
Up nextThe cover that lapses in the year it was neededPrevious: The plan somebody else has to run
Finished this lesson?

Mark it done to track your progress through the curriculum.

Common questions

Short, direct answers to what people ask about this topic.

why does term insurance have no maturity value
Because a term plan is a pure protection contract — the premium buys a payout only if the insured person dies during the term, and nothing is set aside to be handed back if they do not. That is exactly what makes it cheap: a policy that also returns money has to fund that return out of your own premium, so the same cover costs considerably more. A run of years with no claim is the contract behaving as designed, not the contract failing.
the two properties of a loss that decide whether it is worth insuring are
Frequency — how often it happens — and severity, what it costs when it does. Attention naturally follows frequency, because small repeated losses are the ones you actually experience, but transferring a risk to an insurer only earns its price against severity you could not absorb yourself. A loss your monthly budget already swallows is cheaper to pay than to insure, since the premium has to cover the insurer’s costs on top of the claim.
how to calculate the break even probability of a term insurance policy
Divide the total premiums payable over the whole term by the sum assured. A level ₹14,000 a year for 30 years is ₹4,20,000 of premiums against ₹1 crore of cover, so the break-even is ₹4,20,000 ÷ ₹1,00,00,000 = 4.2% — above roughly that chance of the covered event across the term, the contract is favourable in plain rupees. The sum ignores the time value of money on both sides and the insurer’s own costs, so read it as a sense of scale rather than a valuation.
is buying insurance a bad deal mathematically
In plain rupees, yes — protection is negative expected value by construction, because an insurer must collect more in premiums than it pays out in claims plus its own costs, or it would not survive. That does not settle whether a particular cover makes sense, because a loss large enough to end a household’s plan is not simply a bigger version of one that dents it. Where the loss is one you could comfortably absorb, though, the same negative expected value is just a bad deal with nothing bought in exchange.
is it worth continuing a term plan if I have never claimed
A run of years without a claim carries almost no information about that question, because a contract written against a rare event produces nothing in the overwhelming majority of years by design. What the answer turns on is the exposure that still exists today: whether anyone would be financially damaged if the insured person died, and whether that loss is one the household could fund out of what it already holds. Those are questions about the current balance sheet, not about the quiet years behind it.