How SARB Rate Decisions Change Your Bond Repayment in 2026

From the repo rate to your prime rate: the transmission nobody explains

When people talk about "the Reserve Bank cutting rates", they almost always mean the repo rate — the rate at which the South African Reserve Bank (SARB) lends to commercial banks. In May 2026 the repo rate sits at 6.75%. You never borrow at the repo rate directly. What touches your bond is prime, the benchmark lending rate the banks set, which in 2026 is 10.25%. The gap between them — 3.5 percentage points — is the standard spread that has held for years, and it is the reason prime moves in lockstep with the repo rate.

Because that spread is fixed by convention, the transmission is almost mechanical: when the SARB's Monetary Policy Committee (MPC) moves the repo rate by 25 basis points (0.25%), prime moves by the same 25 basis points within a day or two. A basis point is one hundredth of a percent, so 100 basis points equals a full 1%. There is no negotiation and no lag worth worrying about — the banks announce the new prime rate the same week. This is different from fixed-rate mortgages overseas; in South Africa the overwhelming majority of home loans are variable and priced off prime, so an MPC decision lands in your bank account within one billing cycle.

Your personal rate is then quoted relative to prime: "prime less 0.5%" or "prime plus 1%", depending on your deposit, credit record, and the bank's appetite at the time. That margin is locked at registration, but the prime base underneath it floats for the whole 20 years. So if you signed at prime less 0.25% (effectively 10.00% in 2026) and prime rises to 11.25%, your rate becomes 11.00% — the discount stays, the base moves. Every example below uses prime itself; just shift by your personal margin to get your exact number.

The MPC cadence: when the decisions actually happen

The MPC meets roughly every two months — about six scheduled decisions a year, with the dates published well in advance. That cadence matters for planning because it caps how fast your repayment can change. A single bad inflation surprise cannot move your instalment overnight; it has to wait for the next meeting, and the committee typically moves in 25bp steps, occasionally 50bp when it wants to send a stronger signal. A 100bp move in one sitting is rare and reserved for genuine shocks.

What the MPC is steering toward is its inflation target, with the midpoint anchored near the bottom of the band. With CPI running around 3.1% in 2026 — comfortably contained — there is little pressure to hike, which is the backdrop to a stable 10.25% prime. But "stable now" is not "fixed forever". The committee reacts to the rand, to global rates, to fuel and food, and to its own forecast two years out. The honest planning assumption is not that rates stay flat, but that over a 20-year bond you will live through several full cycles of cuts and hikes.

This is why you should treat the current rate as a starting point, not a guarantee. A useful discipline: every time the MPC meets, check whether the decision changed your instalment and whether your budget still absorbs the next plausible move. Six checkpoints a year is not onerous, and it stops a slow drift of hikes from quietly breaking a household that only looked at the rate on the day it signed.

Worked example: what 25bp, 50bp and 100bp do to a R1.5m and R2.5m bond

Take a 20-year bond at the 2026 prime rate of 10.25%. On R1,500,000 the monthly instalment is R14,724.65. On R2,500,000 it is R24,541.08. Those are your baselines. Now move the rate and watch the instalment, holding the term at 20 years. A 25bp hike (to 10.50%) adds R251 a month on the R1.5m bond and R418 on the R2.5m. A 50bp hike (to 10.75%) adds R504 and R840 respectively. A full 100bp hike (to 11.25%) adds R1,014 a month on the R1.5m bond and R1,690 on the R2.5m.

Cuts work symmetrically, with a small twist: each step down saves slightly less than the same step up cost you, because you are applying the rate to a falling balance. A 100bp cut (to 9.25%) drops the R1.5m instalment by R987 to R13,738, and the R2.5m by R1,644 to R22,897. Annualised, that 100bp cut is worth about R11,840 a year on the R1.5m bond and R19,733 a year on the R2.5m — real money that either lands in your pocket or leaves it, decided by a committee you do not sit on.

The long-run number is the one that should change behaviour. On the R1.5m bond at 10.25%, you repay R3,533,916 over 20 years — R2,033,916 of that is pure interest. Push the rate to 11.25% for the life of the loan and total interest climbs to R2,277,322; pull it to 9.25% and it falls to R1,797,121. That is a R480,000 swing in lifetime interest from a single percentage point. It also explains why paying a little extra into the bond early, while rates are higher, compounds so powerfully: every rand of principal you knock out stops being charged at the prevailing prime rate for the remaining term.

Why rent-vs-buy and rentvesting maths are so rate-sensitive

Buying is a leveraged, rate-exposed bet; renting is not. When you buy, your largest monthly cost — the bond instalment — is set by prime, and a 100bp move changes it by roughly R1,000–R1,700 in these examples. Your rent, by contrast, escalates on a contractual schedule (the engine models 6% a year) and is completely deaf to the MPC. So the entire rent-vs-buy comparison hinges on a rate the buyer cannot control and the renter is insulated from. Raise rates and buying looks worse on cash flow; cut them and the buyer's instalment shrinks while the renter keeps paying the same escalating rent.

Rentvesting — renting where you want to live while owning a rental property elsewhere — is doubly geared to the rate, and that is the whole point of stress-testing it. You earn a yield on the investment property (the framework models gross yields near 9.0% in Cape Town and 13.7% in Johannesburg) and you pay a bond rate on it. When prime is below your gross yield, the property's own rent helps carry the bond; when the gap narrows, the shortfall comes out of your salary. On a R1,500,000 flat, a Johannesburg-level 13.7% gross yield throws off about R17,125 a month in rent against a R14,725 instalment at 10.25% — a positive carry. A Cape Town-level 9.0% yield throws off about R11,250, which does not cover the same instalment, so the position only works if appreciation and tax efficiency make up the difference.

Now push prime up 100bp. The Joburg instalment rises to R15,739 and the buffer over rent thins from roughly R2,400 to under R1,400 a month; the Cape Town position bleeds an extra R1,014 every month on top of an already-negative carry. The maths does not just get worse linearly — it flips. Below some rate the rentvesting case clears comfortably on cash flow; above it you are subsidising a tenant out of after-tax salary and betting entirely on capital growth. Knowing where that flip sits for your specific yield, deposit and margin is the difference between a strategy and a hope.

How to stress-test before you sign

Do not buy on the rate you can afford today; buy on the rate you can survive. A workable rule is to add 200bp to the current prime rate and confirm the instalment still fits inside your budget with room to spare. On the R1.5m bond that means checking you can carry roughly R16,800 a month (prime at 12.25%), not just the R14,725 at 10.25%. On the R2.5m bond it means stress-testing around R28,000 rather than R24,541. If the higher number forces you to cut essentials, you are buying at the top of your range and a normal hiking cycle will hurt.

Build a literal buffer for it. If you can comfortably afford the stressed instalment, pay the difference into the bond now while rates are flat — that R1,000–R2,000 a month both reduces lifetime interest and creates an access-bond cushion you can pull back if the MPC hikes and cash gets tight. This converts rate risk into prepaid principal, which is the cheapest insurance available to a homeowner. The same discipline applies to a rentvesting bond: size the shortfall you would face if prime rose 100–200bp and your yield held, and make sure your salary can absorb it without touching emergency savings.

Finally, separate the cash-flow question from the wealth question. A higher rate can make a property painful to hold month-to-month while still leaving you better off over a decade once appreciation, rent escalation and tax treatment are counted — and the reverse is also true. The only way to see which way your specific numbers tip is to model both the buy and the rentvesting paths across a range of rates. Treat this as general information rather than advice for your exact situation, and run the scenarios with your real deposit, margin and target suburb before committing.

Pressure-test your own decision in the rentvesting engine

The figures above use prime at 10.25% and round numbers for the bond; your reality has a personal margin, a specific deposit, a real yield for your suburb, and transfer and bond costs the site models at roughly 3.5–4.5% of price. Small changes to any of those shift where the rent-vs-buy answer flips. The point of stress-testing is not to find a single "right" rate — it is to see how fragile or robust your plan is across the range the MPC could realistically deliver over six meetings a year.

Open the rent-vs-buy net-worth engine, drop in your numbers, and slide the rate up by 100bp and 200bp to watch your instalment, your carry and your projected net worth move together. Do the same for a rentvesting structure and compare the two paths side by side. If your decision survives a hiking cycle on paper, it will survive one in real life — and if it only works at today's rate, the tool will show you that before the bank does.