How much can you borrow for a Swiss home? The FINMA math

Every prospective Swiss buyer asks the same question first: how much will a bank actually lend me? In most countries the answer tracks your real monthly payment. In Switzerland it does not. The answer is set by a self-regulation formula that banks must apply, and it turns on four fixed numbers, none of which is the rate you will actually pay.

The four numbers are these.

• 33.3%: your total housing cost cannot exceed one-third of gross household income.
• 5%: the imputed mortgage rate banks test you against, regardless of the real rate.
• 1%: the annual maintenance allowance, as a share of property value.
• 15: the number of years over which the second mortgage must be amortised (or fewer, if retirement comes first).

What unites the four is that not one of them is your real monthly payment. Three of them (the stress rate, the maintenance allowance, the forced amortisation) are costs the bank constructs for the test, and the fourth is the ceiling it measures them against. A buyer can know their true cash outflow to the franc and still have no idea what a bank will lend, because the bank is solving a different equation.

This piece walks through each of the four, then puts them together with a concrete worked example and solves for the maximum house price a real household can buy. If you want the history behind why the regime is built this way, the regulatory and cultural context lives in our companion piece, why Swiss mortgage lenders are so conservative. This one assumes you either know the why or do not need it, and gets straight to the arithmetic.

The anchor of the whole calculation is the affordability ceiling: a household's total annual housing cost must not exceed one-third (33.3%) of gross household income. Gross, not net; the test is run before tax, on the headline salary.

Housing cost here is not the rent-equivalent or the real mortgage payment. It is a constructed figure with three components: the stressed mortgage interest, the maintenance allowance, and the mandatory amortisation contribution on the second mortgage. Add those three, divide by gross income, and the result has to come in at or under a third.

The 33.3% is not a bank's preference; it is the FINMA-recognised minimum standard that the Swiss Bankers Association self-regulation imposes across the industry (SwissBanking mortgage-market regulation; FINMA mortgage-market fact sheet). A bank can apply a stricter internal limit, and some treat 38% as an absolute outer bound for exceptional cases, but 33.3% is the number that governs the overwhelming majority of lending decisions. A household that clears it gets financed; one that does not, broadly, does not, however comfortable its real cash flow looks.

One point worth fixing early: the test is run on gross household income, so a dual-earner couple is assessed on the combined figure, and a bonus or variable component is usually discounted or excluded by the bank rather than taken at face value. The 33.3% is applied to income the bank considers durable, not to the best year on a payslip. That tends to make the effective ceiling stricter still for anyone whose pay is heavily variable.

The crucial move is that two of the three cost components are deliberately set above reality. That is where the conservatism actually bites, and it is the subject of the next two sections.

The largest of the three cost components is interest, and it is computed at an imputed rate of 5%, not the rate you will sign for. In 2026 that gap is enormous. The Swiss National Bank has held its policy rate near zero, and Swiss mortgages price off that: SARON-based mortgages sit around 1% to 1.5%, and ten-year fixed rates around 1.5% to 2.5% (SNB monetary-policy decisions). Banks underwrite at 5% anyway.

The arithmetic of the wedge is stark. Take a CHF 1,000,000 mortgage. At a real SARON rate of 1.5%, the annual interest is CHF 15,000. At the 5% stress rate, the same loan is assessed at CHF 50,000 of annual interest. The bank runs the affordability test against the CHF 50,000 figure. Your actual cash outflow of CHF 15,000 is irrelevant to whether you qualify.

The rationale is the whole point of the regime. A mortgage is a multi-decade commitment, and the bank is not asking whether you can pay today's rate; it is asking whether you could still pay if rates returned to a normal, or a 1990s, level over the life of the loan. The 5% is a long-run through-the-cycle assumption, chosen so that affordability does not evaporate the moment rates rise. It is the single mechanism that most constrains who can buy, and it is fixed: it does not move when SARON moves, so a falling-rate environment gives a buyer no extra borrowing room at all.

The 5% is not an arbitrary scare number. Swiss mortgage rates reached 7.9% as recently as October 1992, during the property downturn that shaped the entire regime (SNB rate history). Measured against that, a 5% underwriting assumption is not pessimism; it is the midpoint of lived Swiss experience. A borrower who finds the stress rate punitive is really objecting to the memory encoded in it.

The 5% has held as the standard underwriting assumption across the Swiss market for years, through rate environments far lower than it (FINMA mortgage-market fact sheet). A borrower should treat it as a fixed feature of the system, not a number that will soften because market rates are low. It is the same figure whether you take a SARON-based tracker or a ten-year fixed, and the same figure in a 0.5% world or a 2.5% one.

The second cost component is upkeep, assumed at 1% of the property value every year. On a CHF 1,500,000 home that is CHF 15,000 added to the annual housing cost in the affordability test, whether or not you actually spend it.

The 1% is a long-run convention, meant to approximate the average annual cost of maintaining and renovating Swiss owner-occupied property across its life, including the lumpy big-ticket items (roof, heating, kitchen) amortised over many years (SwissBanking mortgage-market regulation). Some banks refine it: a newer property may be assessed at around 0.7%, an older one at 1.5% or more. But 1% is the standard figure the affordability test uses, and unlike the stress rate it tracks a real, if averaged, cost. It is the least aggressive of the three components, and the one a buyer is least likely to argue with.

Two details matter. The 1% is charged on the property value, not the loan, so it does not shrink as you amortise; a fully owned home still carries the assumed upkeep in any later affordability reassessment. And it is not waived because you intend to do the work yourself or because the building is new. The test treats maintenance as an unavoidable cost of holding the asset, which, averaged over the decades a Swiss household typically holds a home, it is.

The third cost component is forced repayment. A Swiss mortgage splits in two. The first mortgage, up to two-thirds (66.7%) of lending value, carries no obligation to repay principal. The second mortgage, the slice from two-thirds up to 80%, must be amortised back down to two-thirds within 15 years, or by retirement age, whichever comes first (SwissBanking, Basel III Final self-regulation). That annual amortisation payment counts inside the 33.3% housing cost.

For most buyers the 15-year horizon governs. A CHF 200,000 second mortgage amortised over 15 years adds about CHF 13,333 a year to the affordability test.

The retirement-age trigger is the catch, and it is where late-career buyers get caught. The amortisation must be complete by the time the borrower reaches the standard retirement age, even if that is sooner than 15 years out. A 55-year-old buying at 80% loan-to-value has roughly ten years to retirement, not fifteen, so the same CHF 200,000 second mortgage must be cleared over 10 years, at CHF 20,000 a year rather than CHF 13,333. The larger annual contribution pushes up the housing cost in the test and tightens the affordability ratio for an identical loan.

There is a second reason the amortisation sits inside the affordability test rather than beside it. The rule is not only asking whether you can service debt; it is engineering your loan-to-value ratio downward over time, so that within 15 years the bank is exposed to at most two-thirds of the property value, the level it considers safe through a price fall. Forced deleveraging is the mechanism, and making the borrower demonstrate they can fund it out of a third of income is how the rule guarantees the deleveraging will actually happen.

The practical consequence: a mortgage taken at 55 is not the same product as one taken at 35, even on an identical house and identical income. The bank is lending into a shorter amortisation window, and the FINMA math makes the older borrower look less affordable for reasons that have nothing to do with their actual ability to pay today.

Take a married couple with a gross household income of CHF 200,000, looking at a CHF 1,500,000 house in the Zurich area. Work it through step by step.

This is the wedge that defines Swiss mortgage lending. The household's real cash flow says yes at 23%; the FINMA stress test says no at 44%. The stress test is the binding constraint, and it is binding by design. The couple can afford the house in any ordinary sense and still cannot borrow for it.

So what can they buy? Solve for the maximum property price P where the stressed housing cost equals one-third (33.3%) of income. With an 80% loan, the cost is 5% on the loan, plus 1% of value, plus the second-mortgage amortisation over 15 years:

(0.05 × 0.8P) + (0.01 × P) + ((0.8P − 0.667P) ÷ 15) ≤ 0.333 × 200,000

With 0.333 × 200,000 = CHF 66,600, that reduces to roughly 0.0589 × P ≤ CHF 66,600, so P ≈ CHF 1,130,943. The couple who wanted a CHF 1,500,000 house is held, by the FINMA math, to about CHF 1.13 million, roughly a quarter less, despite a real payment that would consume less than a quarter of their income.

Run your own figures on the Swiss mortgage affordability calculator, which applies exactly this stress formula on live cantonal data and returns the maximum price and the equity it requires.

The FINMA math is not trying to estimate what you can afford. It is trying to estimate what you could still afford in a far worse world.

Three things follow for anyone planning a Swiss purchase.

Plan against the stress, not the actual. The income a bank requires is meaningfully higher than the income your real monthly payment would suggest. In the worked example the gap was a 44% stressed ratio against a 23% real one. Budget your purchase price from the stressed number, because that is the one the bank uses.

Wage growth helps; cheap rates do not. Because the stress rate is fixed at 5%, a low-rate environment gives you no extra borrowing capacity. The only levers that raise your FINMA maximum are a higher gross income, a larger down payment, or a cheaper house. Waiting for rates to fall does nothing; they are already far below the rate you are tested against.

Age changes the product. The retirement-age amortisation trigger means a late-career buyer faces a tighter test for the same loan, because the second mortgage must be cleared over fewer years. Buying at 55 with a 25-year horizon is not the same transaction as buying at 35, and the affordability gap is structural, not a matter of negotiation.

The honest summary: the FINMA math is not trying to estimate what you can afford. It is trying to estimate what you could still afford in a far worse world. Once you read the 5% and the 33.3% that way, the maximum it hands you stops looking arbitrary and starts looking like what it is, a deliberately conservative answer to a deliberately pessimistic question.