How this model works
Most rent-vs-buy calculators compare a mortgage payment to a rent cheque and stop there. That comparison is close to meaningless, because the two households are not doing the same thing with their money — and because a mortgage payment is not purely a cost. Part of it is principal, which is you paying yourself.
This model runs both households month by month for the whole horizon and enforces one rule: they spend the same total, every month. Whoever has the cheaper month invests the difference at the return you set. Both start with the same cash — the buyer sinks it into a down payment and closing costs, the renter puts it in the market. At the horizon the home is sold, selling costs come out, and the two net worths are compared.
That equal-outlay rule is enforced, not assumed. If the two households’ lifetime spending ever diverges by more than a rounding error, the model refuses to report a result at all. It is the single assumption that makes the comparison mean anything, so it is checked rather than trusted.
Does property tax go up when your home appreciates?
In most places, yes. Property tax is typically ad valorem — assessed against market value, land included. So this model scales your year-one tax figure by however far the home has moved from its purchase price. Your inputs imply an assessed rate, and that rate is then charged against the current value.
The same question has a different answer for each cost, and it is worth being honest about how confident each one is:
- Property tax — strongest case for tracking value. But regimes vary enormously: caps, acquisition-value systems, and exemptions all break the link. If you live under an assessment cap such as California’s Proposition 13, turn tracking off and set a fixed growth rate instead.
- Insurance — weaker. Premiums track rebuild cost, which covers the structure but not the land. In land-scarce markets, appreciation raises the home’s value without raising what it would cost to rebuild, so tracking overstates your premium.
- Maintenance — weakest. A roof costs what a roof costs. Upkeep follows the structure and the price of labour and materials, not the value of the dirt underneath.
This matters more than it sounds. If you leave these costs fixed while testing different appreciation rates, you have quietly rigged the test in favour of high appreciation — the home gets more valuable but never more expensive to own. That is why tracking is the default here.
Why a housing crash does not make buying a good deal
Link costs to value without a floor and the model produces something absurd. A deep price decline drives property tax, insurance, and maintenance toward zero. The buyer’s carrying cost collapses, they invest the savings, and the model cheerfully reports that a 50%-a-year housing crash leaves the buyer better off than steady 3% appreciation.
That is an artifact, not a finding. A home that has lost half its value still needs a roof, still gets a tax bill, and still has to be insured.
The value-link floor prevents it. At 100%, a value-linked cost can rise with the home’s value but never falls below what it would have cost on its own inflation path. On the upside the floor does nothing at all — appreciation already dominates — so it does not touch the headline result. It binds only on the downside, which is exactly where the unfloored model went wrong. Set it to zero and you can watch the absurdity reappear.
PMI: 80% LTV is not 20% equity
These sound like the same thing and are constantly confused. Private mortgage insurance is cancelled at a loan-to-value threshold, measured against your original purchase price. Eighty percent means stop once the balance falls to 80% of the price — which is the same moment you reach 20% equity. If you enter 20% thinking in equity terms, you will model PMI running almost the entire life of the loan.
The thresholds themselves are worth knowing, because there are two of them:
- 80% — the point at which you may request cancellation.
- 78% — the point at which, under the Homeowners Protection Act, the lender must terminate it automatically whether you ask or not.
- FHA loans follow different rules entirely, and depending on when the loan was written the premium may never cancel.
Setting a PMI rate with no cancellation threshold at all is rejected as an error rather than silently modelled, because no conventional loan behaves that way. On a typical scenario the difference is not small: it bills roughly twice the PMI a real loan would ever charge.
Should you pay extra principal, or invest the money?
The answer depends almost entirely on one comparison: your investment return against your mortgage rate. Paying down a 4.5% loan is a guaranteed 4.5% return. Whether that beats the market is the whole question, and nobody knows the answer in advance.
What the model does show is the mechanics, all of which fall out of the amortization rather than being bolted on: the loan retires early, total interest drops, PMI ends sooner because the balance falls faster, and once the loan is gone the payment disappears entirely — so the buyer’s monthly cost falls off a cliff and they start investing the difference instead.
One modelling decision is worth stating plainly, because it is what keeps this honest: the renter invests the same extra amount. Extra principal is money the buyer spends, so under the equal-outlay rule the renter must be handed the same sum to invest. Without that, the buyer would be quietly given free money the renter never receives, and every comparison involving extra principal would be meaningless.
The crossover is not a breakeven
The crossover is the first month buying pulls ahead of renting. It is deliberately not called a breakeven, because it says nothing about whether buying stays ahead. The lead can change hands more than once, and when it does, the model says so — a single crossover date is not a durable verdict.
Something else shows up once you test this: the crossover tends to be all-or-nothing rather than gradual. On a typical scenario, nudging rent down a few hundred dollars moves the crossover from a few years to never, with very little middle ground. The question people arrive with is “how many years until buying wins”, and the honest answer is usually that it is either fast or it never happens at all.
What this model leaves out
Every model is a simplification, and the omissions here are as important as the mechanics:
- Income tax. No mortgage interest deduction, no SALT deduction, no cap. For most households since 2018 the standard deduction makes this smaller than folk wisdom suggests, but it is not zero.
- Investment tax. The invested difference grows untaxed. A real taxable brokerage account would not.
- Moving and transaction friction beyond the closing and selling percentages you enter.
- Rent control, vacancy, and landlord behaviour. Rent grows at a smooth rate here; real rent moves in steps and sometimes not at all.
- Maintenance timing. Upkeep is charged smoothly each month. Real roofs fail all at once.
- Flexibility and risk. Renting is liquid and mobile; a house is neither. Owning concentrates a large, leveraged, undiversified bet in a single asset in a single town — often the same town your job is in. None of that carries a dollar figure here, and for many people it is the deciding factor.
The numbers this page produces are only as good as the assumptions you feed it, and the three that matter most — investment return, home appreciation, and rent growth — are genuinely unknowable. Treat the output as a way to test how much those assumptions matter, not as a prediction.