Should You Prepay the Home Loan or Invest the Money?

The decision

Somebody with a home loan and a surplus at the end of the month faces a genuine choice. The money can reduce the debt, or it can be invested, and both are defensible.

What makes the question hard is not that nobody knows which is better on average. It is that the two arms are different kinds of thing. Prepaying produces a certain outcome: reduce the balance and the interest that balance would have generated never arises, at exactly the borrowing rate, with no tax on the saving because avoiding a cost is not income. Investing produces an uncertain outcome with a wide range around it, and the gain is taxed when realised.

So the honest test is not which has the higher expected return. It is how often the uncertain arm beat the certain one, by how much, and — the part that decides it — what happened in the periods when it lost.

How this was tested

A lump sum, at every possible starting day in the record. One arm reduces the debt and compounds at the borrowing rate. The other buys the index and is sold at the end of the period, with capital gains tax charged from the rules file rather than from memory.

Two things had to be handled honestly.

The borrowing rate is not something this site holds. There is no Indian home-loan rate series in this repository, so rather than invent a representative figure the experiment is run at five different rates and the whole range is published. That is the same device used elsewhere here for costs that could not be sourced, and it has the advantage that the reader can find their own rate in the table.

What the record does carry is context for that ladder, though not the rate itself. Over the 11 financial years since 2016-17, the benchmark scheduled banks set their lending against — currently the MCLR (Overnight) — has ranged from 6.7% to 8.3%. That is not a home-loan rate and cannot be substituted for one: what a particular household pays depends on its lender and on its own circumstances, and no series for that exists here. What it does establish is that the span the ladder covers is a real one rather than an arbitrary spread around a guess, so a borrower reading their own rate off their own statement is placing it against something observed. The benchmark has also been redefined more than once, which is why the range above is taken within the current definition rather than across the whole record — a range spanning two definitions would measure the change in definition.

And both arms start with the same money on the same day and end on the same day. Any difference comes from the choice, not from the setup.

The result

If the debt costs	Investing came out ahead in	Typical difference	Worst period
7%	97.6%	+140%	-37%
8%	94.4%	+120%	-56%
9%	87.7%	+100%	-77%
10%	80.4%	+77%	-99%
11%	73.5%	+52%	-124%

Across 4,282 ten-year periods, investing won most of the time at every rate tested. At a borrowing rate in the middle of the ladder it came out ahead in 87.7% of periods, and the typical margin was large — roughly the original lump over again.

Now read the last column, which is the one that matters. Even at the middle rate, the worst ten-year period left the investor 76.75% of the lump behind where prepaying would have put them. That is not a bad quarter; it is the outcome of a full decade, and it is a real loss against a choice that had no bad outcome available to it at all.

That asymmetry is the whole decision. The prepayment arm's worst case and its best case are the same number. The investing arm has a distribution, and the question a household actually has to answer is not "which is higher on average" but "if I get the bad end of that distribution, what happens to me".

What the investing arm actually delivered

Separately from the comparison, it is worth seeing the range the market produced.

Over the same ten-year periods, the after-tax annualised return ranged from 4.8% to 20.9%, with the middle at 12.9%. One period in ten did worse than 8.7%.

And a limitation that must travel with those figures: in this record there is no ten-year period in which the investment lost money in nominal terms. That sounds like a strong argument for investing and it is mostly a statement about the sample. The Indian series covers a single stretch of a single market containing a long expansion; other markets have had ten-year periods, and twenty-year periods, that lost money. The absence of one here is not a law, and the article that sets out what this record can and cannot tell you should be read alongside this one.

The windows also overlap heavily. Thousands of ten-year periods drawn from twenty-seven years of history are nowhere near thousands of independent observations, and the confidence a large count suggests is not there.

What decides it for a particular household

The arithmetic says investing usually won. The decision is not made on the arithmetic alone, and four things move it.

Whether the bad case is survivable. If losing a substantial part of the lump against the alternative would force a change to something that matters — a goal missed, a house not bought, a loan that becomes hard to service — then the certain arm is the right one regardless of what usually happens. Most published advice on this question skips this entirely.

What else is unfunded. Someone without an emergency reserve, or without adequate insurance, is choosing between two long-term uses of money while a short-term hole sits open. Neither prepaying nor investing is the answer; the hole is.

Other debt. This comparison is about a home loan, which is usually the cheapest borrowing a household has. Applied to a credit card balance the answer reverses completely and stops being interesting — a card rate is far above anything the ladder here covers, and clearing it is not a choice that needs modelling.

How it feels to carry the debt. This sounds like a soft consideration and it is a real one. Someone who would sleep better without the loan, and who would abandon the investing plan in the first bad year, should prepay — because the plan that gets followed beats the plan that is optimal on paper and gets abandoned halfway.

Two things the test does not include

Tax relief on the loan. Interest and principal on a home loan may attract relief depending on the borrower's circumstances and the regime they are taxed under, and any such relief reduces the effective borrowing rate, which shifts the comparison towards investing. This page states no such provision, because those are statutory particulars that change and nothing in this repository sources them. Someone running this comparison for themselves should establish their own effective rate after any relief, and then read the ladder at that rate rather than the headline one.

The choice is not binary. Nothing requires the whole surplus to go one way. Splitting it captures part of each and is frequently the sensible answer for a household that cannot decide, particularly one for whom the bad case is uncomfortable but not ruinous.

There is also a decision inside the prepayment itself: whether to reduce the instalment or shorten the term. Shortening captures more of the benefit, because it removes more months of interest, and the amortisation schedule shows why prepaying early is worth so much more than prepaying late.

What to take away

Prepaying earns your borrowing rate, guaranteed and untaxed. Investing earned more than that in most ten-year periods on record, often by a wide margin — and in the worst periods it lost heavily against an alternative that could not lose at all.

The record contains no ten-year period where the investment lost money outright, and that says as much about a short sample from one market as it does about equities. So do not treat the historical margin as a forecast.

Decide on the bad case rather than the average. Fill the emergency reserve and the insurance first, clear expensive debt before considering either, establish your own borrowing rate after whatever tax relief actually applies to you, and split the surplus if you genuinely cannot choose. Then prepay early rather than late, and shorten the term rather than the instalment.