Fixed vs SARON: the Swiss mortgage decision in 2026

This is the last of four pieces on Swiss mortgages, and the one about the choice you actually sign. The upstream questions are settled elsewhere: why the lending rules are so strict (why Swiss mortgage lenders are so conservative), how much a bank will lend (the FINMA math), and how to assemble the deposit (the 20% down problem). Once the bank has said yes and the equity is in place, one decision remains: fixed or SARON.

Fixed-rate (Festhypothek). The rate is locked for a defined term, typically 2 to 10 years, with some banks offering 15. Monthly payments are constant for the term, which is the entire appeal. The cost of that certainty is rigidity: exiting before the term ends triggers an early-repayment penalty (Vorfälligkeitsentschädigung), which in a falling-rate environment can be large.

SARON-linked (SARON-Hypothek). The interest tracks the money market. Daily SARON fixings are compounded over a roughly three-month interest period to produce the period rate, and a contractually agreed margin is added on top for the all-in rate. The rate is reset quarterly, at the start of each quarter, not daily, so the payment is stable within a quarter and steps between quarters. SARON contracts run 1 to 5 years, with 3 to 5 most common, under a framework agreement that sets the term.

SARON (the Swiss Average Rate Overnight) replaced CHF LIBOR as the Swiss money-market benchmark at the end of 2021. It is a secured overnight rate the SNB steers close to its policy rate, which is why a SARON mortgage moves almost in lockstep with SNB decisions, and why the rest of this piece spends so long on what the SNB is doing.

One product feature worth naming: UBS markets a SARON Flex option that lets a borrower convert from SARON into a fixed-rate mortgage within a few working days of a request (UBS SARON mortgage). It is a UBS-specific feature, not a universal one, and it matters for the risk discussion later: it bounds, without eliminating, the danger of being caught in a rising-rate cycle.

The numbers below are as of June 2026 and are indicative market ranges, not quotes; mortgage pricing varies by bank, loan-to-value, and creditworthiness, and moves week to week.

• SNB policy rate: 0.0%, held since the 19 March 2026 monetary policy assessment (also held in December 2025). Sight deposits above the threshold are remunerated 0.25 percentage points below the policy rate.
• SARON spot: near 0%, around −0.05%.
• SARON margins at the major retail banks run roughly 0.70% to 1.30% above compounded SARON 3M, with prime first-mortgage borrowers around 0.8% to 1.0%. That puts the all-in SARON rate at roughly 0.8% to 1.2% (Comparis; moneyland SARON FAQ).
• 5-year fixed: about 1.3% to 1.6%, an indicative ~1.5%.
• 10-year fixed: about 1.4% to 2.0%, with a wide lender spread (Comparis mortgage rates; key4.ch).
• 15-year fixed: offered by some banks at a modest premium over the 10-year; less commonly quoted.
• Direction: the SNB inflation forecast is 0.5% to 0.6% across 2026 to 2028, the low end of the band, signalling no tightening pressure; UBS does not forecast further cuts.

The headline picture: SARON is cheaper than the 5-year fixed by roughly 30 to 70 basis points at the short end, and the curve is flat to slightly down. A borrower choosing today is choosing between a known ~1.5% for five years and a floating ~1.0% that is cheaper now and moves with the SNB.

The floating product wins until rates rise, and then it does not.

Over long horizons, money-market mortgages have usually been cheaper than fixed, and it has not been close. The trade has always been lower expected cost for higher payment volatility.

The cleanest illustration comes from the decade to 2023. On a CHF 500,000 mortgage held from 2013 to 2023, a 10-year fixed cost about CHF 115,000 in total interest, while the money-market (SARON) route cost about CHF 44,000 (moneyland). The floating borrower saved roughly CHF 71,000, about 62% of the fixed interest bill, by riding the SNB's long descent into negative rates.

The same data carries the warning. In October 2023, after the SNB's 2022 to 2023 hiking cycle, SARON had drawn level with the six-year fixed at about 2.61%, and through 2024 SARON was, on average, briefly more expensive than fixed (moneyland, October 2023). The floating product wins until rates rise, and then it does not.

What the history does not say is that SARON wins going forward. The period since 1991 has been largely disinflationary, which is exactly the environment in which a rolling short rate beats a locked one. Long-run studies favour the borrower who keeps rolling the short term, but conditional on the SNB not running a hiking cycle through the middle of the hold. The evidence is a description of a regime, not a forecast of the next one.

Five factors decide it, and none of them is the current rate alone.

The rate-environment view.The first question is where the SNB is heading. In June 2026 the answer is nowhere in the near term: inflation sits at the low end of the target band, no tightening pressure is signalled, and UBS does not forecast further cuts. A flat, low policy path is the environment in which SARON's lower starting rate compounds in the borrower's favour.

Risk tolerance. The second question is whether the household can absorb a shock. At current levels, a 100-basis-point rise in SARON roughly doubles the monthly interest. A household with income headroom and savings can take that; one already near the affordability ceiling from the FINMA math cannot, and should buy the certainty of fixed.

Optionality. The SARON Flex switch-to-fixed feature (UBS specifically) means a SARON borrower who senses a hiking cycle can convert into a fixed rate within days. That bounds the tail risk, but it does not remove it: you may read the turn late, and the fixed rate available when you switch will already have moved up in anticipation. Optionality narrows the downside; it does not floor it.

Holding horizon. A three-year SARON contract inside a ten-year hold gives several decision points to switch or refinance, and lets the borrower capture more of the disinflation evidence the history rewards. A buyer who wants to think about the mortgage once and never again is, by temperament, a fixed-rate buyer.

The post-2028 reform overlay. Once mortgage-interest deductibility is curtailed under the imputed-rental reform (effective 2028 at the earliest, with the Federal Council signalling 2029, EFD), the tax shield on mortgage interest shrinks. That favours carrying less interest in absolute terms, which today points to SARON because SARON currently carries less interest; the direction would flip if SARON became the more expensive product. This is a directional read of the deductibility mechanics, not a sourced product-specific claim.

Take the couple from the down-payment piece: a CHF 1,200,000 property in Zug, CHF 240,000 down, leaving a CHF 960,000 mortgage. At June 2026 rates:

• SARON at 1.0% all-in: CHF 9,600 of interest in year one.
• 5-year fixed at 1.5%: CHF 14,400 in year one.
• SARON saves CHF 4,800 in year one.

If rates hold for five years, the gap simply compounds: SARON costs CHF 48,000 in interest over the term, fixed costs CHF 72,000, and SARON saves CHF 24,000.

The timing of the hike is the whole story. A hike at the start of year two hands the five-year comparison to fixed by CHF 14,400. A hike delayed to the start of year four still leaves SARON ahead by CHF 4,800. The break-even sits between a year-three and a year-four hike: roughly, if SARON survives the first half of the term without a 100bp rise, it wins the five years; if the rise comes early, fixed wins.

For the long-horizon reference, the 2013 to 2023 figure (a CHF 500,000 mortgage, about CHF 71,000 of cumulative SARON saving over ten years, moneyland) is the historical norm for a disinflationary regime, not a projection of the next decade.

A closing note on the optionality: the UBS SARON Flex switch bounds this stress case, since the couple could convert to fixed when they saw the cycle turning. The residual risk is timing. You switch too late and eat the start of the move. Model your own version of this on the mortgage affordability calculator, which shows what a given rate does to the maximum price you clear.

• In June 2026, with SARON at roughly 1.0% all-in and the five-year fixed near 1.5%, SARON is materially cheaper at the short end. The SNB is on hold and the inflation picture does not signal an imminent hike.
• The risk to SARON is an SNB hiking cycle; the mitigation is the switch-to-fixed option (UBS SARON Flex specifically). It is bounded, not eliminated, and timing is the residual exposure.
• For risk-averse buyers, including anyone who cannot absorb a 100bp jump in the monthly payment, fixed buys payment certainty at a known premium of roughly 30 to 70 basis points today.
• Over long horizons, the historical evidence favours rolling short-term borrowing, conditional on monitoring the rate environment rather than setting and forgetting.
• The 2028 reform marginally favours carrying less interest in absolute terms; it does not structurally favour either product.