Making the Call
Rent by default. Buy when you clear all four gates below, and understand that they are gates, not a scorecard — three out of four is a no, because each one describes a way the position kills you independently of the others.
Gate 1: Price. Does the rent justify the purchase? This is the arithmetic gate, and it has a closed-form answer. You are indifferent between renting and buying at the price where the annual rent equals the annual unrecoverable cost of ownership. Setting and solving for gives the most you should pay for a specific house given what its equivalent rents for:
Invert it and you get the price-to-rent ratio directly — the same ratio quoted as folk wisdom in section “Calculating Financial Impact Of Buying A House”, now derived rather than asserted:
Run the earlier example. With , , , , and , the denominator is 5.85%, so . The $2,000,000 house renting for $7,500 a month carries a price-to-rent ratio of . That is well above the threshold, which is the same verdict the dollar comparison reached — and the formula also tells you the price at which the answer flips: . Not $2 million. If you want that house, that is the number to negotiate toward, and if the seller will not go there, the market is telling you to rent it instead.
Notice what governs the threshold. The familiar advice — “above 20, rent; below 15, buy” — is simply this formula evaluated at carrying costs of 5% and 6.7%. The dominant, most volatile term is , and it moves with rates. Drop to 2% in a cheap-money regime and rises to 22.5: the identical house, at the identical rent, becomes a defensible buy. Nothing about the property changed. This is why people who bought in 2021 and people who declined in 2024 can both have been right, and why anyone quoting a fixed price-to-rent threshold without naming their cost of capital is repeating a number rather than making an argument. Recompute it at today’s rate, with your own tax position, every time.
Gate 2: Horizon. Will you hold it long enough? Round-trip friction alone requires roughly 8.5% of appreciation to break even (section “True Affordability in the Housing Market”), and full break-even including carry lands in the five-to-ten-year range. The gate is not “do I plan to stay?” — everyone plans to stay. It is whether you would still be here if your employer relocated, the relationship ended, or the school turned out to be wrong. If any of those is live, you are buying an asset with a two-month exit and a two-year break-even. Rent.
Gate 3: Balance sheet. Does it fit without deforming everything else? Two tests, both after closing, not before. First, liquidity: reserves outside the house sufficient for the runway of section “Underwriting Your Own Income”, which for most readers means twelve months of PITI still sitting in T-bills the day after you get the keys. Second, concentration: the equity you put in, measured against investable net worth, at a level you would tolerate in any other single illiquid position (section “Housing’s Role in Optimal Portfolios”). A purchase that passes the price test and fails this one is not a good deal you can afford — it is a good deal for someone else.
Gate 4: Correlation. Is this the same bet you have already made? If your income, your employer equity, and this house all depend on one industry in one metro, the diversified position is the rental (section “Underwriting Your Own Income”).
Then price what the arithmetic missed — do not wave at it. Suppose you clear gates 2 through 4 and fail gate 1 by $460,000 of purchase price. You are not obliged to walk. You are obliged to convert that into an annual number and look at it. Amortized over a ten-year hold, paying $2,000,000 for a house worth $1,538,000 on rent economics costs roughly $27,000 a year in unrecoverable carry above what renting the same shelter would cost. That is the price of the thing arithmetic does not capture: permanence, control, the freedom to renovate, not negotiating with a landlord, and the school district you actually want.
State it as a purchase and the decision becomes easy to make honestly. Twenty-seven thousand dollars a year is roughly a very good car, every year, forever. Some households will look at that and buy without hesitation, and they are not being irrational — a primary residence is consumption, and consumption is allowed (section “Consumption or Investment? Get the Definition Right”). What is not allowed is pretending the number is zero, or that appreciation will refund it. Name the premium, decide you want it, and buy. The failure mode this entire section exists to prevent is paying the premium while believing you made an investment.
| Signal | Points to renting | Points to buying |
| Price-to-rent vs. | Market ratio above your computed threshold | Market ratio below it |
| Confident holding period | Under five years, or genuinely uncertain | Ten years or more |
| Cost of capital | High-rate regime; large loan above the $750,000 deductibility cap | Low-rate regime; small or no loan |
| Post-closing liquidity | Down payment consumes the reserves | Twelve months of PITI remains outside the house |
| Income and location | Job, equity, and metro all one bet; volatile or specialized income | Portable skills, diversified local economy |
| Marginal dollar | Tax-advantaged space still unfilled (section “The Down Payment Versus the Retirement Account”) | Space filled; surplus cash beyond it |