What Is a Rent-Controlled Apartment Worth?

We read a story in The New York Times — "As A.I. Money Floods the Market, San Francisco Renters Weigh Buyouts" — about landlords paying tenants to give up rent-controlled apartments. One tenant — 74 years old, retired, and 30 years in the same apartment — had been offered $40,000 to move out. He asked for $600,000 instead. And we thought: that is a present value problem. So we decided to dig in.

So which of those two numbers is more reasonable? Forty thousand dollars sounds like a lot of money to move. Six hundred thousand sounds like a joke. It turns out the joke number is much closer to defensible, and you can work out why with one idea from the rest of this site.

The setup

Paying below market is the same as being paid.

Rent control is a local law that limits how much a landlord can raise the rent on an existing tenant. San Francisco has one. Its Rent Board sets an allowable increase each year — for March 2026 through February 2027 it is 1.6%, calculated from 60% of the Bay Area Consumer Price Index. Rents out in the open market are not capped at all, and over the long run they have grown much faster than that.

The discount works like income. Not cash, though — it arrives as rent he never has to hand over. Every month he keeps the difference between what the apartment would rent for and what he actually pays.

That difference is where our starting number comes from. The Times reports that he pays around $1,100 a month for a two-bedroom in North Beach, against a citywide median asking rent in August of $4,395. Subtract one from the other and the gap he keeps is $3,295 a month — about $39,500 a year.

That gap did not appear overnight. He has lived in the apartment for 30 years. For three decades his rent went up by the small amount the city allows while the market went up by whatever it wanted, and the space between the two widened a little every year. He now pays about a quarter of what the apartment would fetch on the open market.

And the gap grows. If market rents rise 4% a year while his rent rises 1.6%, the two lines pull apart a little more every year. That is the whole engine of this article.

What is really going on here is compound interest wearing a different hat. Compounding is growth that builds on itself — each year's increase is a percentage of a number that already contains every increase before it. Both rents compound, his slowly and the market's quickly, so the space between them compounds too. The gap actually grows a shade faster than market rent does — about 4.8% a year at the start — because the slow-growing rent is being subtracted out of it.

That shape matters more than it looks. If both rents went up by a flat dollar amount each year instead of by a percentage, the gap in year 20 would be $6,459 a month. Because they compound, it is $8,119 — about $1,700 a month that exists for no reason other than the shape of the growth. You can watch the same effect at work in The Power of Compound Interest.

Which tells you what both sides are actually negotiating over. The tenant's asset is not really the apartment — it is the run: thirty years of increases that never kept pace with the market, and the prospect of more of them. He wants the run to continue. The landlord wants it to end, because a change of tenant lets the rent reset to market. Neither of them is arguing about this month's rent. They are arguing about who gets the run from here.

The core idea

Add up the future, then shrink it.

You cannot just add up twenty years of savings and call that the value. Money you receive in the future is worth less than money you have today, because money today can be invested and grow. To compare the two fairly, you shrink each future amount back to what it is worth right now. That shrunken figure is its present value, and the yearly rate you shrink by is the discount rate.

We will use 5.5% as the discount rate — roughly what a careful saver could earn on safe investments. Now the question has a real answer: what is the present value of every month of rent he will not have to pay?

That depends almost entirely on one thing nobody knows for certain — how long he would have stayed anyway. So we will price three versions.

That is the first surprise. A $600,000 counteroffer is not a stunt. It is what the lease is worth if this tenant stays about sixteen more years — which, at 74, would take him to 90. Meanwhile $40,000 is not a lowball so much as a rounding error.

On present value alone, the outrageous number is the reasonable one.

But present value on its own is only part of the story. Two things change the picture, and the first one is easy to miss.

Reality check one

The buyout is taxed. The rent he skips is not.

This is the part almost everyone forgets. A buyout is ordinary income — the same kind of income as a paycheck. It arrives in a single year, all at once, and it stacks on top of everything else he earned that year, which pushes it into his highest tax brackets. A marginal tax rate is the rate charged on the top dollar of income, and for a payment this size it is very high.

Rent he never pays, on the other hand, is not income at all. Nobody taxes it. So a dollar of buyout and a dollar of saved rent are not worth the same thing to him. He keeps every dollar of the rent he does not pay, but only about sixty cents of every dollar the landlord hands him. Comparing the two head-to-head makes the offer look about forty percent better than it is.

After tax, the $40,000 offer covers ten months of the rent gap. Not ten years. Ten months.

Turn it around and it is starker still. To hand this tenant, after tax, the full value of what he is giving up, the landlord would have to write these checks:

Look at the middle number. Measured this way, the $600,000 counter is not aggressive at all — it is almost exactly what the landlord would have to pay, before tax, to leave him whole for ten more years in the apartment.

Reality check two

Nobody knows how long he would have stayed.

Every number above depends on one figure that neither side actually has: how many more years the tenant would have stayed. In present value terms that is the number of payments — and it is the single biggest input in the whole calculation.

The tenant wants to be paid for a long stretch of time. He is giving up an apartment he may well keep for the rest of his life, and he prices it that way. The landlord has no way of knowing how long that is. At 74 and retired, this tenant is not going to be moved by a new job or a growing family — the things that would end his tenancy are health, family, or a decision to leave the city. Nobody at the negotiating table can put a date on any of them.

So the two sides are not really arguing about present value. They are arguing about duration. And neither of them can settle it, because it depends on a future that has not happened yet.

There is one piece of evidence, and it points the tenant's way: he has already been there thirty years. Someone who has not moved in three decades is not an obvious candidate to move now. That is not proof of anything — but it is the only real information either side has.

Here is what the lease is worth at three different answers.

That is a spread of half a million dollars, and it comes entirely from a number nobody can know. It is why these negotiations are so hard.

But notice what the range does not include. Even the shortest case — five more years, a genuinely conservative assumption — is worth $189,458. That is still more than four times the $40,000 on the table. You have to assume the tenant was about to leave within roughly a year before $40,000 starts to look like fair value.

The market

So what do these deals actually settle for?

We have been working out what the lease is worth. That is not the same as what buyouts actually cost, and San Francisco is one of the few places where you can check. Every buyout in the city must be filed with the Rent Board, and the filings are public.

This is the honest answer to why the market clears where it does. Buyout prices are not set by present value. They are set by bargaining power and by who needs money sooner. A tenant sells a $400,000 asset for $40,000 because he cannot sell it to anyone else, cannot borrow against it, cannot take it with him, and is negotiating against someone who does this for a living.

Every signed buyout is a tenant whose real reservation price was lower than the economic one — because he was moving soon anyway, or needed cash now more than housing in 2046, or was frightened of being evicted, or simply never did this arithmetic.

One number in that chart is worth pausing on. The 2026 median has jumped to $50,000 from $36,959 the year before. When money floods into a city, the price of getting a tenant to leave goes up along with everything else.

The verdict

So which number is right?

The tenant's, and it is not close. Taken at face value, $600,000 prices about sixteen more years, which would take him to 90. But that is the figure before tax. After tax it leaves him roughly $367,000, which buys back just under ten years of the rent gap — so his counter is close to the honest price of staying to 84. Meanwhile $40,000 is off by a factor of ten. Even the shortest case we ran — five more years, to age 79 — is worth $189,458. There is no reasonable set of inputs that produces $40,000. That figure prices a tenancy of about one more year.

Is $40,000 a good offer? No — but it is a normal one, which is a more interesting fact. It is the market median, and roughly half the tenants who are offered something like it say yes. As an opening bid it is perfectly rational for the landlord. As a final settlement it moves something like $300,000 of value from the tenant to the owner.

What should he actually do? Understand what his own number is for. $600,000 will probably not be paid — no San Francisco buyout on record has ever reached it. But it is an excellent anchor, and it is honestly derived, which matters if the conversation gets serious. The useful move is to name the assumption out loud: how many more years are we pricing? Once both sides are arguing about that instead of about a lump sum, the arithmetic does the rest. And whatever the number, he should price it after tax — a $250,000 buyout is really a $150,000 buyout.

Two things would change the answer. If he was going to leave soon regardless — moving closer to family, or into care — the present value collapses, and $40,000 to $80,000 becomes genuinely fair. And if market rents stop rising, the widening-gap engine switches off and the whole valuation deflates with it.

None of which is really about apartments. It is about what happens when you take something that pays you a little every month, for a long time, and try to price it in one lump. The arithmetic is the same whether it is a lease, a pension, a lottery annuity, or a royalty check. Add up what is coming, shrink it for time, subtract what the tax collector takes, and then ask how likely it all is to actually arrive.