Is a Rental Property a Retirement Plan?
"We were thinking of buying a rental instead of putting more in the brokerage." I have heard a version of that sentence from four different people this year, and in every case the reasoning was the same: the rent covers the mortgage, so after thirty years you own a house for free.
That sentence contains one assumption doing all the work, and it is almost always wrong by about half.
Where the rent actually goes
Here is a reasonable rental run against the mid-career sample household. The Hendricks family, 44 and 42, $163,800 of household income, $343 a month of slack. They buy a $340,000 property at 50 with 25% down, financing the rest over thirty years at today's 7.24%, and let it out at $2,200 a month — the number on the listing.
The plan inflates both the price and the rent to the year of purchase, so at 50 it is a $394,296 property with a $309,296 loan and a $2,108 monthly payment, asking $30,616 of rent across that first year. Here is what becomes of that rent:
The property loses money in its first year, and it is not close. The vacancy allowance is 7.6%, which is the national rental vacancy rate from the Census Bureau's Housing Vacancy Survey for the third quarter of 2026 — not a pessimistic assumption, the actual measured figure. The $9,264 is upkeep at 1% of the building's value a year, property tax at the US average effective rate, and an insurance premium. Every one of those is a field you can overwrite with your own county's tax bill and your own insurance quote, and you should.
None of that money is lost, exactly. It went into a building. But it did not go into a retirement account, and the question on the table is which of those two things funds a retirement.
What it does to the plan
Run it to the end and the rental's effect on the investment portfolio is almost nothing:
Fifty-one thousand three hundred and ninety-seven dollars, on a portfolio of eight and a half million, across forty-two years. The household goes from $343 a month of slack to $180 a month under water to get it. If the rent were the point, this would be the end of the post.
A note on the figure in the box, because I got this wrong in the first version of this article and someone may have followed the old advice. Type the gross rent. planbend takes the vacancy, the management fee, the upkeep, the property tax and the insurance off for you. Netting them off by hand first means they come off twice, and the plan will tell you the property is a disaster: entering $1,100 instead of $2,200 on this same rental reports $1,358 a month under water and $481,981 taken off the portfolio. That number is not a more conservative answer. It is a wrong one.
Twelve years before it pays for itself
The rent rises with inflation and the mortgage payment does not, so the shortfall closes. At age 62 — twelve years after the purchase — the property clears its own costs for the first time, by $293. Follow it to 70 and the picture is genuinely better:
Twenty years of tenants to reach $490 a month of profit. "The rent covers the mortgage" turns out to be a statement about the 2040s, and in the intervening twelve years the difference comes out of a retirement plan that was already tight.
Then look at the last row. Four hundred and sixty-six thousand dollars of equity, built out of $85,000 down and twenty years of somebody else paying most of the interest. That is where the money is. It was never in the cash flow, and every version of this calculation I have run says the same thing.
One detail in the tax treatment is worth pulling out, because it is the reason a rental's cash flow and its tax bill disagree so violently. In year one the plan deducts $22,295 of mortgage interest and $11,470 of depreciation against $28,289 of banked rent, so on paper the letting loses money — while also losing money in the bank. The two numbers are never the same, and the depreciation that shelters the early years is the same depreciation that gets recaptured at the end.
The losses you cannot use yet
That paper loss does not simply vanish, and it does not simply offset your salary either. A rental is a passive activity. An owner who actively participates — choosing tenants, approving repairs, setting the rent — may deduct up to $25,000 of a passive loss against ordinary income, but that allowance falls by fifty cents for every dollar of income above $100,000 and is gone entirely at $150,000. The Hendricks household earns $163,800.
So for the whole of their working lives they deduct none of it. The losses are suspended, not lost, and they accumulate:
The year they stop working, their income drops under the phase-out and the allowance opens. From 60 onward the plan starts drawing the old losses down against their retirement income. By 70 the letting is turning a paper profit, and that profit is absorbed by what is still suspended rather than taxed — which is why the taxable rental income at 70 is zero against $5,881 of real cash.
Then they sell, and the rest of it arrives at once. A fully taxable disposition frees every suspended loss that is left: $73,507 of deductions, accumulated a few thousand at a time over twenty years, all landing in the single year with the largest gain in the whole plan. That is not an accident of the rules; it is the design of them. Fifteen years of being told no end in one year of yes.
Not everything you spend is a deduction
The upkeep allowance covers two different things that the tax code treats nothing alike. A repair — a service call, a broken window, a coat of paint — is deducted in the year you pay it. An improvement — a roof, a boiler, a rewire — is not deducted at all. It is added to what you paid for the property and written off over 27.5 years, the same as the building.
The plan now splits the allowance in half between the two, which is a working convention rather than a measurement, and the field is editable. Over twenty years the Hendricks rental accrues $50,361 of capital improvements. That has three consequences, and only the first is obvious:
- A smaller deduction now, because half the upkeep is no longer expensed when paid.
- A larger one later, because each year's improvement starts its own 27.5-year clock — a roof replaced in year twelve is still being written off in year thirty-nine, long after the building itself is fully depreciated.
- A higher basis at sale, which cuts the gain — against more accumulated depreciation to recapture, which does not.
A tool that deducts the whole allowance when it is spent, which this one did until November 2026, flatters the early years and overstates the gain at the end.
Until November 2026 this plan modelled none of that. It floored a rental's taxable income at zero, which was conservative in any single year and wrong across a property's life, because it quietly threw away a six-figure deduction that the sale would have released.
So count the equity properly
The portfolio figures above exclude the property, because the investment portfolio is what funds retirement spending and a duplex does not. Which means that for a rental, the headline number understates the case badly. Here is the honest comparison, with the property equity added back:
Held to the end of the plan the rental is worth $1,163,703 — a portfolio $51,397 heavier and $1,112,306 of bricks. Sold at 70 and reinvested, it is worth $1,871,904, and that is after the tax on the sale.
Selling beats holding by $708,201, and the reason is not complicated: the plan grows property with inflation and grows a portfolio at the portfolio's rate. Once the leverage has done its work — once the loan is small relative to the value — the asset is an inflation tracker, and leaving $466,478 in it is a choice to earn inflation on half a million dollars.
The bill at the end
Selling an investment property is taxed in a way that surprises people, and until November 2026 this plan did not charge any of it. Here is the sale at 70:
Three things in that table are worth sitting with. The first is how close the gain and the cash are — $409,817 against $427,712 — and that they are two unrelated numbers that happen to have landed near each other. Paying off a loan does not reduce a gain. Here the capital improvements raised the basis by enough to keep the gain just under the cash; on a property improved less, or refinanced along the way, it is entirely possible to owe tax on more money than the sale put in your hand.
The second is that three fifths of the gain is depreciation recapture — $247,140 of it. Twenty years of deductions cut a $444,657 basis to $197,517, and most of what looks like profit is that gap. It is not mostly appreciation; it is mostly the tax code handing back what it lent you. You are required to claim that depreciation, and section 1250 recaptures it whether it ever helped you or not, which is why a landlord who never bothered to claim it still owes this.
The third is the rate. $45,773 on a $409,817 gain is 11.2%, which is lower than most people expect, for two reasons that compound. The recapture is taxed at your ordinary rate capped at 25% rather than at a flat 25%, and this household is well past its earning years by 70 — and the $73,507 of released passive losses has just wiped out what ordinary income it did have. A household selling the same property at the top of its career would pay considerably more on both counts.
And the gain pushes this household's MAGI to $409,817 in that one year, which is the thing nobody plans for: a one-off sale lifts a Medicare premium two years later. The plan shows that now. It could not before.
Two variants worth running
The second home nobody rents out. Same $340,000, same date, rental income zero.
Two and a half thousand dollars a month under water, and $1,225,998 off the portfolio. The roof, the tax bill and the insurance do not care whether there is a tenant under them, which is the whole difference between this and the version above. The equity is still built, so this is not money set on fire — but it is a lifestyle purchase with a thirty-year loan attached, and the only honest way to judge it is as a lifestyle purchase. The plan is useful here precisely because it refuses to call it an investment.
Paying cash. Here the plan stops and says no. $340,000 in today's money is $394,296 by the time they buy at 50. The Hendricks household has $1,007,995 that year — and only $172,524 of it outside retirement accounts. Draining every taxable account they own still leaves them $167,476 short.
So the cash-or-mortgage question everybody asks is not available to this household. They could reach it by taking a six-figure early withdrawal from a 401(k), paying income tax and a 10% penalty on the way out, and that is a different question with a worse answer. The useful output is not a number here. It is that the option is not there, which no amount of comparing end balances would have told them.
An earlier version of this section said paying cash left them around half a million dollars ahead of the brokerage. That figure was wrong, and a reader checking the arithmetic is what surfaced it. The projection took a down payment out of one named account, emptied it, and forgave whatever was left of the price — so the household was buying a $394,296 property for about $172,000, and the comparison flattered cash by exactly the amount it had not paid. The engine now spends down every taxable account in order and reports what it still cannot cover, rather than writing it off.
What I would actually do with this
Replace the defaults with the real numbers. The vacancy rate, the tax rate and the insurance premium that ship in those boxes are national averages, and your property is not national. The county assessor publishes the tax rate, an insurer will quote the premium in ten minutes, and the upkeep default of 1% of value a year is a rule of thumb rather than a measurement — the one figure in there with no statistic behind it. A 1960s duplex deserves worse than 1%.
Put in what you actually paid. If you already own the place, the cost basis box matters more than anything else on the form. Left blank it assumes you paid today's value, which means the plan taxes only future appreciation and understates the bill at the end — badly, if you bought in 2012.
Model the year it goes wrong. A single $14,000 roof, or four months empty, is not an edge case over thirty years — it is Tuesday. The upkeep percentage is an average across decades, not a promise about any one year. Add the bad year as a one-off Withdrawal from the account you would actually raid, and see whether the plan notices. A rental that only works in the years nothing breaks is not working.
Compare it against the thing you would otherwise do. Not against zero. The brokerage line in every table above is the alternative, and it is doing better than people expect, with no tenants and no roof.
Decide the exit before the entry. The largest controllable number in this whole post is the difference between selling at 70 and holding, and nobody ever decides that at purchase. Put an age in the box. You can change it later; you cannot un-not-think-about-it.
Run it on the sample plan
That is the real application against a finished household, fully unlocked, saving nothing anywhere. Scenarios tab, Scenario Toggles, + Add a change, Buy Property. Put in the price in today's dollars, the age you would buy, the down payment, today's rate, and the gross asking rent. The panel will show you what the rent nets, line by line, before you commit to anything.
For your own plan, planbend is free to start and never asks for a bank login. If the purchase is the house you live in rather than one you let out, what giving up a low mortgage rate costs is the same arithmetic pointed at your own front door, and the big-purchase version of it covers everything else with an upkeep bill. Before you commit capital to a property, it is worth a conversation with someone licensed — there are directories on the Connect with a professional page.