Why the usual rule fails
The conventional test asks what the money bought. A house or an education is good debt; a holiday or a phone is bad debt.
It fails in both directions, and the failures are not edge cases.
It approves loans that are dangerous. A home loan whose EMI consumes most of a single income, taken by someone with no emergency fund, is classified as good debt by this rule. It is the loan most likely to force a distressed sale.
It condemns loans that are sensible. A small, short, interest-free instalment on a washing machine you need, comfortably serviced, is classified as bad debt — while costing essentially nothing.
The rule sorts by what was purchased, when the thing that determines the outcome is the terms and your capacity to carry them. The same house at the same price is a sound decision for one household and a catastrophe for another, and the asset is identical in both.
Three questions instead
A specific loan, at a specific rate, for a specific borrower. Ask three things.
| The question | Why it decides | |
|---|---|---|
| 1 | What does it cost, as a rate? | The rate is the certain, unavoidable outflow. Everything else is a hope. |
| 2 | Does the borrowing produce value that exceeds it? | Income, a durable asset, or an earning capacity — not just satisfaction. |
| 3 | Can I service it if my income stops for six months? | This is the one that decides whether a bad year becomes a permanent loss. |
All three, not any one. A loan that fails the third question is dangerous regardless of how well it passes the first two, because the failure mode is not paying more interest — it is being forced to sell the asset at whatever price is available in the month you cannot pay.
Question three is also the one nobody asks, because at the point of borrowing the income is present and the shock is hypothetical.
What the rate is really telling you
The rate is the clearest signal available and it is largely a summary of the lender's view of the risk.
| Roughly | Typical of | What it implies |
|---|---|---|
| Single digits | Home loans, loans against securities | Well secured. Cheap enough that repaying early competes with investing rather than obviously beating it. |
| Low to mid teens | Personal loans, some education loans | Unsecured or weakly secured. Repayment usually beats any expected investment return. |
| High twenties and above | Revolving credit card balances, some short-term lending | No investment competes. Clearing it is the highest-return action available. |
Which produces a much simpler working rule than the good-debt taxonomy: the higher the rate, the less the purpose matters. Above the high teens, no justification about the asset rescues the arithmetic.
Debt is not good or bad — the ratio is
Here is the reframing worth taking away.
The single most predictive number is not what you borrowed for. It is what share of your take-home income leaves as loan repayments every month.
That figure determines resilience. At a low share, a job loss is survivable, a rate rise is absorbable, and you retain the ability to save. At a high share, the household is running with no slack — every obligation is fixed, income is not, and the first interruption forces borrowing at worse terms or selling at a bad time.
Lenders apply their own conventional thresholds and they are set to protect the lender's recovery, not your resilience. A loan you are approved for can be a loan you should not take, and the approval is not a second opinion.
The useful stress test is arithmetic: recompute the ratio against a meaningfully lower income — one salary instead of two, or a freelance year at 70% — and see whether the payments still work. If they do not, the debt is too large, whatever it bought.
“Good debt” is also a sales phrase
Worth naming, because the framing is not neutral.
“Good debt” is used to justify borrowing more than someone intended, and to reframe a large obligation as a sophisticated financial decision. Leverage does amplify returns, and it amplifies losses identically — a fact that survives every retelling of the first half.
Two specific cases worth thinking about carefully rather than by category. Borrowing to invest pays a certain interest cost for an uncertain return, which is the opposite of the trade in the order of operations. And a tax deduction on interest reduces the cost of a loan; it never makes the loan free, and it is not a reason to hold debt you could clear.
Common cases, judged on the test
- Home loan. Usually the cheapest borrowing available and it buys something you would otherwise rent. Passes questions 1 and 2 comfortably. Question 3 is where it is won or lost, and it is decided by the size relative to income rather than by the fact of it.
- Education loan. Buys earning capacity, which is the clearest version of question 2. Fails when the cost is unrelated to the realistic income the qualification produces — the asset is real and the arithmetic still has to work.
- Car loan. A depreciating asset, so question 2 is weak unless the vehicle produces income or replaces a larger cost. Frequently the loan that pushes an otherwise sound household past question 3.
- Personal loan for a genuine emergency. Expensive, but the alternative — selling long-term assets at a bad moment, or missing an essential payment — can be worse. This is exactly the situation an emergency fund exists to prevent.
- Revolving credit card balance. Fails all three. There is no version of this that is defensible, and clearing it comes before investing.
- Interest-free instalments on something you need and can afford. Passes on cost, so it comes down to whether it is genuinely interest-free — the cost is often in the price — and whether it is quietly accumulating alongside others.
Where repayment competes with investing
For anything above the mid teens the comparison is not close. For a cheap secured loan it genuinely is, and the honest way to settle it is to look at what the money would otherwise have done — including in the bad windows.
FNOTrader's Mutual Funds app runs contribution schedules against real NAV history — around 34 million NAV rows — reporting XIRR, and rolling returns across every start date rather than one flattering window. Set the worst window against the loan's certain rate: the loan's saving arrives with certainty and the investment's does not, and that asymmetry is the whole comparison.
Common questions
What is the difference between good debt and bad debt?
The conventional answer — debt for appreciating assets is good, debt for consumption is bad — sorts by what was purchased rather than by the terms and your capacity to carry them. A better test asks the rate, whether the borrowing produces value exceeding it, and whether you could service it if income stopped for six months.
Is a home loan always good debt?
No. It is usually the cheapest borrowing available and buys something you would otherwise rent, but a home loan whose EMI consumes most of a single income, held without an emergency fund, is the loan most likely to force a distressed sale. Size relative to income decides it, not the asset.
What matters more than what the debt was for?
The share of take-home income leaving as loan repayments each month. That figure determines whether a job loss is survivable and whether you retain the ability to save — it is far more predictive than the category of the asset.
How do I know if I have too much debt?
Recompute your repayment-to-income ratio against a meaningfully lower income — one salary instead of two, or a freelance year at 70%. If the payments no longer work, the debt is too large regardless of what it bought.
Does being approved for a loan mean I can afford it?
No. Lender thresholds are set to protect the lender's recovery rather than your resilience, so approval is not a second opinion on whether the borrowing is sensible for you.
Is it worth keeping a loan for the tax deduction?
A deduction reduces the cost of borrowing; it never makes the loan free. It is not a reason to hold debt you could otherwise clear, and rules change with the Finance Act.
Should I borrow to invest?
It pays a certain interest cost for an uncertain return, which inverts the usual trade. Leverage amplifies losses exactly as much as gains — a fact that tends to survive only the first half of most retellings.
Is a car loan bad debt?
It funds a depreciating asset, so it is weak on the value test unless the vehicle produces income or replaces a larger cost. It is frequently the loan that pushes an otherwise sound household past the point where a shock is survivable.
Continue reading
More in Personal Finance Basics · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article