About these calculators
Why I built these
I started the rent vs. buy calculator when I was considering buying a home myself. I wanted an equal-footing comparison between renting and buying, and I could not find one readily available online.
It began as a simple Python script and grew from there, as I kept thinking through different scenarios and how one decision or the other might shape my finances and my retirement.
Around the same time, a friend was selling their ten-year-old car, so I applied the same thought process to that. My own question was only whether the car was a good buy — but I was curious how the numbers would look if I bought new instead. That is how the car calculator came about.
The last one has a rather different story. I showed the rent vs. buy calculator to a friend for feedback, and he mentioned he had been wondering how buying a rental property compares with simply investing the same money. It is a common question, asked and argued over often enough both in conversation and online, and an interesting one — even though it concerns fewer people than the first two. That question became the rental property calculator.
Most of my family and friends are not especially technical, so I could not just hand them Python scripts and expect them to run the scenarios they cared about. A website seemed the easiest way to let them try their own numbers, and to be useful to other people while it was at it.
I hope it is useful to many. I welcome feedback, so please get in touch if you find any inconsistencies, bugs or logical errors.
What they have in common
Every calculator here answers the same shape of question: you have a sum of money and more than one thing to do with it, and the options are usually compared by their monthly payment, which is the wrong comparison.
They share a principle, not a formula. Every option is priced on an equal economic footing — money not spent on one choice is money still working somewhere else, and that has to be counted or the comparison is meaningless. Each option is then scored on what it leaves you with at a horizon you set. Income is identical under every option, so it cancels and is never modelled, and every assumption is an input you control.
How that principle is applied differs by question, because the questions are not the same shape:
- Rent vs. buy runs two households for thirty years and holds them to the same monthly spend. Whoever has the cheaper month invests the difference, so neither side is quietly handed extra money. The model refuses to report a result if the two lifetime totals ever diverge.
- Car: lease vs. buy compares three ways to pay for one car, where the spending is deliberately not equal — paying cash is a large outflow now, a lease is a small one every month. So instead of equalising the spending it charges every outflow the growth it never earned: a dollar paid in month t costs you that dollar plus what it would have become by the horizon.
- Rental property vs. investing equalises the capital. Both sides start with the same cash, and every month the property fails to pay for itself the landlord writes a cheque and the investor puts the identical amount into the market. Neither household needs anything here — both are investments, so the question is which does more with the same money.
Both answer the same underlying question — what is this genuinely costing me, once the money's alternative use is counted — and each calculator's methodology page sets out its own mechanism in full. Do not assume a figure from one was produced the same way as a figure from the other.
Rent vs. Buy
A month-by-month simulation of renting against owning. Not a payment comparison, and not a mortgage calculator with a rent field bolted on: it runs two households side by side for the whole horizon, and whoever has the cheaper month invests the difference.
Things most rent-versus-buy calculators skip:
- PMI with a real cancellation threshold, measured as loan-to-value against the original price.
- Property tax, insurance, and maintenance that can be linked to the home’s current value rather than held flat.
- A floor on those value-linked costs, so a price crash cannot drive a homeowner’s carrying costs to zero.
- Extra principal payments, including the early payoff, the reduced interest, the earlier end of PMI, and the drop in monthly outlay once the loan is gone.
- Selling costs at the horizon, so the buyer is scored on what they would actually walk away with.
- Holding periods measured in months, not just whole years, because a short hold is where buying looks worst and is worth being able to ask about.
It does not stop at one number. Three of the inputs — investment return, home appreciation, rent growth — are not measurements, they are guesses, and a single figure computed from a single guess at all three reads as far more certain than it is. So every result also carries:
- A scenario table: the same comparison under six rate regimes, with your own assumption slotted in among them, so you can see where yours sits rather than only what it produces.
- A robustness verdict in words — robust, assumption-sensitive, or finely balanced — because winning five of six scenarios and winning three of six are reported identically by a headline figure and mean entirely different things.
- A sensitivity matrix sweeping return against appreciation and colouring every combination by winner, so the question becomes how far the answer is from changing rather than what the answer is.
- An optional Monte Carlo: hundreds of correlated draws around your own numbers, reporting how often buying ends ahead at every holding period, with a confidence interval on the figure.
And, on every calculator, a solver that runs the whole thing backwards: instead of asking what happens, it asks how wrong you would have to be for the answer to change, and reports the tipping point for each input with the closest one first. Where an input has no tipping point, it says so — which is often the more useful answer, because it means that input is not worth arguing about.
None of that is a forecast. It is a map of how much the answer depends on assumptions you cannot verify.
Two smaller things that follow from the same idea. Every scenario is carried in the page’s address, so you can bookmark a result or send it to somebody and it opens as you left it. And because these tools ask for more inputs than anybody changes, each result says how many are still ours — an answer built from three of your figures and twenty-seven of our defaults is mostly a statement about our defaults, and it should say so.
US income tax is modelled and switched off by default, so the headline figures are pre-tax until you ask otherwise. Switched on, it prices the mortgage interest deduction against the standard deduction — which is usually why that deduction turns out to be worth nothing — and the IRC 121 exclusion on the gain, which the renter’s portfolio has no equivalent of and which is normally the larger effect by some distance.
It still leaves out moving friction, rent control, lumpy maintenance, and the value of flexibility. The calculator and its methodology page explain each omission and why it matters.
Car: Lease vs. Loan vs. Cash
Three ways to get the same car — lease it, finance it, or pay cash — priced on one basis. A dollar paid in month t costs you not just the dollar but the growth it never earned, so every option is one stream of outflows plus whatever the car is worth at the end. Lower net cost wins.
Things a payment calculator cannot tell you:
- The opportunity cost of the down payment. Cash on the hood is the most expensive money in the deal, because it stops compounding first.
- A lease cannot be cancelled. Signing one commits you to the whole term on day one, so the model carries the payments still owed as a liability at every month in between.
- Maintenance tracks the car’s age, not the calendar. Re-leasing keeps resetting the car to nearly new, which is where a lease earns back part of its cost.
- The residual is the whole lease. It sets the payment, and it sets whether buying out at the end is a bargain or a mistake.
- Negative equity. A financed car can be worth less than its loan for years. The model shows the gap and how long it lasts.
Sales tax has always been modelled here, and it is the tax that decides this comparison: a purchase is taxed on the whole price at delivery, a lease on each monthly payment in most states. Income tax is off by default; switched on it adds the annual property tax some states charge on a vehicle’s value — which costs most on whichever option holds the newest car, and over a long horizon that is the lease chain — and a business-use deduction, taken by a different method on an owned car than on a leased one.
It leaves out mid-lease exits and transfers, gap insurance premiums, and EV credits. It also leaves out fuel and tyres — on the assumption that you are comparing one car driven the same distance every way, in which case they cancel. That assumption breaks down precisely when you compare a new lease against a used purchase, where economy can differ and a car handed back every three years may never need tyres at all. The methodology page says so plainly.
Two things the car model is explicit about. Depreciation is the largest lever in it, and the shipped curve is a market average rather than a forecast for any particular car — so every result says whether it is using that estimate or resale figures you supplied for your own model, year and mileage. And when the APR is zero it names the cash rebate at which the financing stops being worth taking, because subsidised rates are offered instead of a rebate rather than alongside one.
Rental Property vs. Investing
Buy a property and rent it out, or put the same money in the market. Most rental calculators stop at cash flow, which tells you whether the property feeds itself but never asks what the money would have done somewhere else. This runs that alternative alongside it on equal capital.
Things it does not skip:
- Vacancy and bad debt. Nobody collects twelve months of rent every year forever. It is the input most often set to zero.
- A capital reserve separate from maintenance. The roof, the HVAC and the flooring arrive in lumps years apart, and they arrive whether or not anyone budgeted for them.
- Management at 8% of rent collected — charged on what you actually collect, not what was scheduled. Setting it to zero does not make the work disappear; it means you are doing it.
- The ratios landlords actually screen on — cap rate, cash-on-cash and DSCR — with net operating income calculated before debt service, which is why a property can show a healthy cap rate and still bleed cash every month.
- Leverage in both directions. A 25% down payment buys four dollars of property per dollar of your capital, which magnifies a gain and magnifies a loss just as hard. Unlike a portfolio, the mortgage payment is due whether or not the tenant pays.
Tax is off by default here too, and that default matters more on this calculator than on the other two, because the tax treatment is a large part of the case for owning a rental at all. Switched on, it models depreciation on the building, recapture at sale at a rate above the capital-gains rate, and the passive loss rules that for most earners defer the shelter rather than granting it. Which way it nets out depends on income, and not in a straight line. On the shipped scenario tax favours investing at every income — but by only about $1,800 at $125,000, where half the $25,000 allowance survives and the bracket is high enough for it to be worth something, against roughly $24,000 once the allowance is gone above $150,000. That cliff is the allowance disappearing, and it is the single largest tax effect on this calculator.
How they are checked
Each model asserts its own invariants and refuses to report a result if any of them break — every dollar of a loan either repaid or still owed, a fixed-rate loan gone at scheduled payoff, and the two sides of a comparison never diverging in what they have spent. A comparison that quietly hands one side more money than the other is worse than no comparison, so those checks are enforced rather than assumed.
Where the numbers come from
Nowhere. Every input is yours to set, and the defaults are only a plausible starting scenario — not a market survey, and certainly not a quote. Put your own numbers in.
The inputs that dominate the answer — investment return, home appreciation, rent growth, depreciation — are all unknowable. Use these models to find out how much your assumptions matter, not to predict what will happen.
It is not advice
These are arithmetic models of the assumptions you give them. They have no view on whether you can afford the thing, whether you will enjoy it, or what your money is for.