The money mistakes that actually cost Indian households

The six, ranked by what they cost

Ranked by expected cost to an Indian household: property as the default investment, no health cover for ageing parents, insurance bundled with investment, education funded out of retirement, gold held without a stated job, and a fixed deposit treated as the safe home for long-horizon money.

The ordering is a judgement, so here is the method. Each mistake is scored on three things multiplied together: annual drag, the share of net worth it touches, and the years it survives unnoticed. The third term decides the ranking, and it is the one every other list leaves out. Every figure below is illustrative arithmetic with stated inputs.

Here is why the third term dominates, with the other two held still. Take a drag of 1 percentage point a year on 5% of net worth, spotted and fixed in year three: about 0.15% of net worth, ignoring compounding. Now take the same 1 percentage point on 60% of net worth, running the full 20 years because nothing ever printed a figure that would have prompted the question: about 12%. Identical annual mistake, eighty times the bill — and every item below is the second kind.

1. The flat as the default, not the decision

Property is not a bad asset. It is a badly measured one, which is a different problem carrying a much larger price tag.

The return almost everyone computes is sale price divided by sticker price. A flat bought at ₹80 lakh and sold ten years later at ₹1.4 crore is 1.75 times the money, about 5.8% a year. That is the figure the family repeats.

Now use money that actually left and actually arrived. Suppose stamp duty, registration, brokerage and the interior spend took the real outlay to ₹88 lakh on day one, and exit brokerage left ₹1.37 crore in hand. That is 1.56 times over ten years — about 4.5% a year, before a rupee of maintenance, property tax or vacancy, and before any interest on the loan. The gap is made entirely of costs that were paid and forgotten.

The trade-off, because 4.5% is not the whole picture either. An owner-occupied flat also delivered ten years of housing that would otherwise have been rented, and a let-out one delivered rent; neither shows up in a price-to-price calculation, and the honest number adds it back. So the point is not that the flat did badly. It is that the 5.8% the family repeats is neither of the two defensible figures — and the decision to buy the next one is being made on it.

The loan compounds the illusion, because the monthly instalment — the EMI — is experienced as saving. In the early years most of it is interest rather than principal: rent paid to a bank. The schedule that shows that split instalment by instalment comes with every home loan, and almost nobody opens it.

Then concentration: one asset, one building, one micro-market, bought with borrowed money. Nobody would accept that shape in a share portfolio; it passes unremarked here only because there is no daily price — until you write it into your net worth.

2. The cover for your parents that quietly stopped being available

Health cover is priced by age, and it excludes what you already have. Those two facts together create a deadline that nobody announces.

Premiums are age-rated, so the cover gets dearer exactly as it becomes likelier to be used; and a condition that exists before the policy starts is not covered until a stated waiting period passes. So the option to insure a parent does not expire with a refusal — an insurer will still sell you something. It expires by exclusion: you buy the year after the diagnosis and find the one thing you wanted it for carved out.

What puts this second is not the hospital bill but how it gets paid. The adult child is India’s insurer of last resort, so the liability sits on your balance sheet whether or not you wrote it down — and an uninsured admission is funded by liquidating whatever sells fastest. A probabilistic health event becomes a realised portfolio loss, and emergencies do not avoid the months when markets are down, which is why the funding matters as much as the cover.

The honest trade-off: a senior-citizen policy is expensive, and it arrives with limits. You typically pay a fixed share of every claim yourself — a co-payment — the daily room charge it will reimburse is capped, and some things are not covered at all. But a budgeted premium with partial cover is a structurally different exposure from an unbounded one — so what a policy does and does not pay is worth reading first.

3. The policy that is also an investment

A traditional life policy or a unit-linked plan makes one premium do three jobs: buy cover, pay whoever sold it, invest the remainder. You are never shown the split. The two fail differently — a unit-linked plan at least publishes a unit value, so the investment leg is visible even where the charges deducted in units are not, while a traditional participating policy publishes nothing anyone could put beside a rival. The arithmetic below is the traditional kind.

Compare the unbundled version. Term insurance buys cover and nothing else, priced as a premium against a sum assured — the amount paid on death. A mutual fund buys investment and nothing else, priced as an annual charge stated as a percentage of what you hold — the expense ratio — against a published daily price per unit, its NAV. Either sits beside a rival’s number in a minute; the bundle produces neither, by design.

Do the arithmetic on the shape of it. Suppose a policy takes ₹50,000 a year for 20 years, each premium paid at the start of the year, and pays a stated maturity value of ₹22 lakh. You put in ₹10 lakh and got back ₹22 lakh, which sounds like more than doubling. Run the internal rate of return on cashflows landing on different dates — XIRR — and it compounds at a shade over 7% a year. The 20 years did the work, not the product, which is why a total-return figure flatters a staggered contribution.

Two mechanisms make that worse than it looks. The money inside a participating policy is invested under limits that keep it heavily debt-weighted, so the ceiling on what it can return is set by prevailing bond yields rather than by anyone’s skill; and early surrender values are punitive, which turns leaving into a sunk cost that keeps charging. You cannot correct a mistake you cannot measure, and you will not correct one that charges you to leave — hence the case for reading how these plans are built first.

4. Funding education out of retirement

There is a lending market for a child’s education. There is none for your retirement. That asymmetry settles this one.

An education is financeable: a term, a stated interest rate, a moratorium while the course runs, a borrower with forty working years ahead. Retirement is the one large liability nobody will lend against. Draining the unborrowable to fund the borrowable is the wrong way round, and it happens constantly because the alternative feels like failing your child.

Timing makes it costlier than it looks. The withdrawal usually lands in your early fifties, when the accumulated retirement pot — the corpus — is at its largest and a rupee therefore carries the biggest absolute future value. Take ₹20 lakh pulled out at 50 with ten years left, growing at an assumed 11%: the multiplier is about 2.84, so that withdrawal is roughly ₹57 lakh not there at 60. Not a forecast — just what the assumption implies.

The trade-off, because the loan is not free either. An education loan carries interest and puts a repayment on a starting salary. What differs is the shape: the loan is bounded, dated, refinanceable and shareable, while a retirement shortfall is unbounded, undated, and funded by the same child, ten years later, without a stated interest rate. The transfer was deferred, not avoided.

5. Gold with no job

Gold is not the mistake. Gold held without an answer to “what is this here for?” is, and the two are easy to confuse because the metal behaves identically either way.

Gold produces no cashflow: no coupon, no dividend, no earnings. The entire return is what the next buyer pays, so there is nothing to discount and therefore no sense in which gold is cheap or dear on fundamentals.

Its two defensible roles are insurance-shaped, and they are not equally solid. Gold is priced internationally in dollars, so a weakening rupee lifts the rupee price even when the dollar price has not moved — that one is mechanical. That gold has often risen when equities fell is the second, and it is a tendency in the record rather than anything the structure guarantees. A household holding gold for growth has bought a hedge and labelled it an engine.

Then the failure mode that is specifically Indian. Nobody in an Indian household ever decided an allocation to gold. It accretes — weddings, inheritance, a little each festival — and because it was never chosen as a percentage it is never reviewed as one. Anything you cannot state as a share of the total is not part of your asset allocation.

Jewellery adds a cost financial gold does not: you buy at a price including making charges and sell at one that excludes them, a spread paid on day one and never returned. Telling, then, that the commonest financial use of household gold here is not selling it but borrowing against it.

6. The deposit that cannot show a loss

A fixed deposit does not remove risk. It swaps one risk for another and conceals the one it kept.

What it removes is price volatility — the number cannot fall. What it keeps is purchasing-power risk, and what it adds is a tax-timing drag: FD interest is taxed as it accrues, every year, whether or not you withdraw, so tax leaves the compounding base annually rather than at the end. Over one year, trivial. Over fifteen it is why a debt fund and an FD with identical gross yields do not finish in the same place.

Now the inflation half, inputs stated so you can redo it with your own. Suppose the deposit pays 7%, your income-tax slab leaves you 5% after tax, and prices rose 5.5% that year. The deposit grew in rupees and shrank in what those rupees buy — and the statement recorded a credit both times, because a deposit statement has no field capable of showing a loss. Inflation over long horizons is slow enough to lose to for a decade unnoticed.

For money needed inside a year the trade-off reverses: nothing improves on a deposit. The mistake is not the deposit. It is the horizon — money with a fifteen-year job parked in an instrument built for a one-year one, and the result called caution.

What all six have in common

Read the six together and one property does all the work: not one of them ever produces a negative number that anybody sees.

The endowment policy reports a fund value that only rises. The flat is marked to a neighbour’s asking price, the gold in rupees per gram, the deposit by a quarterly interest credit. The uninsured parent and the raided retirement account report nothing at all, because an absence has no statement. Meanwhile the mutual fund portfolio — usually the smallest holding in the house — prints red several times a year, and is therefore the one thing that gets scrutinised and occasionally fixed.

That is the ranking, restated. An asset that cannot display a loss is never sold, so the error runs for the full holding period instead of being caught in year two — which is the third term in the scoring above doing its work. The most expensive holdings in an Indian household are not the riskiest ones; they are the unmarked ones. Measurement is what triggers correction.

Which makes the fix mechanical rather than motivational. Force each holding to produce a number, once.

• The policy: every premium and the maturity value, with dates, and the XIRR of that series.
• The flat: all money out, acquisition costs and upkeep included, against what would arrive net of brokerage today — plus the rent it saved or earned, which belongs in the total.
• The gold: as a percentage of net worth.
• The deposit: the return after tax and after inflation, negative sign included.
• The parents: the premium you are not paying, beside the largest bill you would fund without it.
• The retirement account: any withdrawal multiplied by growth over the years remaining.

Six numbers, none needing a view on markets. This describes how the structures work; it is not advice about your money.

Computing the two that are genuinely tedious

Four of those six are a spreadsheet afternoon. Two are not: they need a long price history rather than your own records.

What a horizon actually held. Whether a fifteen-year job was better served by a deposit or a fund is answerable from history — not from one trailing return, but from every window of that length, including the worst. FNOTrader’s Mutual Funds app computes rolling-return distributions and worst peak-to-trough falls — drawdowns — across the published NAV history from AMFI, the mutual fund industry body, around 34 million NAV rows.

What a contribution schedule produced. The same XIRR arithmetic runs against a real NAV series for the same amounts on the same dates, with a stated period and scheme. Past performance is a record of what happened, not an indication of what will.

FNOTrader is not a SEBI-registered investment adviser and does not give investment advice.

Common questions

Continue reading

More in Personal Finance Basics · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article