How SONIA Swap Rates Control Your Fixed Mortgage Price

Here's a question that trips up a lot of people renewing a mortgage: the Bank of England's Monetary Policy Committee meets only eight times a year, and between meetings the base rate simply doesn't move. So why did the fixed-rate quote from your bank change three times in the last fortnight? The short answer is that fixed mortgages aren't priced off the base rate at all — they're priced off a different, faster-moving number called the SONIA swap rate, and understanding the difference is the key to understanding why fixed and variable mortgages can behave in completely opposite ways in the same week.

SONIA — the Sterling Overnight Index Average — is the Bank of England's own benchmark interest rate, and the Bank is literally its administrator: every London business day, it collects data on real overnight loans between financial institutions and calculates the average rate at which they actually lent to each other. It replaced LIBOR, the older and less reliable benchmark that was based on estimates rather than real transactions, with sterling LIBOR settings permanently stopping publication at the end of 2021. SONIA is now the reference point for an enormous share of UK financial markets — the Bank of England's own figures put it at over £90 trillion in referenced transactions annually. SONIA itself tracks just under whatever the Bank of England's base rate happens to be, since it's measuring overnight borrowing costs directly shaped by that policy rate — with Bank Rate held at 3.75% since the Committee's sixth consecutive hold on 17 September, SONIA itself sits close to that level too.

That close tracking at the overnight level is precisely why SONIA on its own doesn't explain fixed mortgage pricing. What does is the SONIA swap market — specifically, an instrument called an Overnight Indexed Swap. Think of it as insurance against interest rates moving over a period of years. A bank funding a five-year fixed mortgage doesn't want to be exposed to whatever happens to interest rates over the next five years, so it enters into a swap with another institutional counterparty: the bank agrees to pay a fixed rate for five years in exchange for receiving the floating SONIA rate over that same period. Whatever fixed rate the bank locks in through that swap becomes the foundation of the mortgage rate it can profitably offer you — because from that point forward, the bank's own cost of funding is fixed, regardless of what actually happens to interest rates afterward.

That swap rate isn't set by today's base rate — it's set by what the market collectively expects the average interest rate to be over the life of the swap. If traders expect the Bank of England to cut rates repeatedly over the next two years, the two-year swap rate will sit below today's base rate, pricing in those expected cuts in advance. If they expect rates to stay elevated or rise, the swap rate sits above it. This is the entire reason fixed mortgage rates can fall even while the Bank of England is holding or raising its base rate, and why they can rise even during a rate hold — the swap market has already moved on updated expectations before the Bank's next meeting even happens.

From that swap rate, a lender builds the actual number you're quoted through four layers stacked on top, and it's worth writing the formula out explicitly rather than leaving it implicit: quoted retail rate equals the swap rate, plus a credit and property risk margin, plus a wholesale funding and liquidity spread, plus the lender's own profit margin. Using current figures for a hypothetical five-year fix rather than a promotional best-buy rate: start with the five-year SONIA swap rate itself, around 4.65% as of mid-September — that's the lender's raw wholesale cost of locking in fixed funding. On top of that sits a credit and property risk margin, roughly 0.35%, sized according to the loan-to-value tier and the borrower's own risk profile. Then a wholesale funding and liquidity spread, covering the regulatory capital the lender must hold against the loan and the cost of the funding itself, perhaps another 0.30%. Finally, the lender's own commercial profit margin, commonly around 0.40%. Add those together — 4.65% plus 0.35% plus 0.30% plus 0.40% — and you land around 5.70%, a fairly typical quoted rate rather than the very best deal on the market. The lowest advertised five-year fixes sit meaningfully below that, in the low-to-mid 4.70s as of mid-September, because they're priced with a much thinner margin — usually reserved for the strongest loan-to-value bands and reflecting a lender's appetite to win market share on a particular product that month. The structure is consistent either way: swap rate as the foundation, with the margin above it varying by lender, loan-to-value and how aggressively that lender wants to compete that month.

The gap between the best-buy rate and a typical quoted rate is itself worth watching, because it tends to widen when swap rates move quickly. Lenders that offer the lowest advertised rates are pricing on thin margins and can't absorb much swap-rate volatility without repricing, so a fast-moving market often means best-buy deals get pulled and replaced faster than the more typical, thicker-margin products around them — which is part of why headlines about lenders "pulling deals" tend to cluster around exactly the best-buy end of the table.

It's worth being precise about what this does and doesn't apply to. Tracker mortgages and most variable and standard variable rate deals move directly with the Bank of England's base rate — when the Bank cuts or hikes, those rates typically follow within weeks, sometimes days. Fixed rates are the ones governed by the swap mechanism described here. This is also why the base rate decision itself, on the day it's announced, often has almost no effect on fixed mortgage pricing — the market usually saw it coming, and priced that expectation into the relevant swap rate weeks beforehand.

Two-year and five-year swap rates don't always move together, and the gap between them tells you something useful. Two-year swaps are highly sensitive to near-term expectations — the next one or two Bank of England decisions, the latest inflation print, anything that shifts the near-term policy path. Five-year swaps reflect a longer-run average expectation and tend to move more slowly, since a single data release rarely changes what the market thinks average rates will be five years out. When the two-year rate sits meaningfully above the five-year rate, the market is effectively pricing in higher rates in the near term followed by cuts later — sometimes described as an inverted curve. When they sit close together, or five-year is higher, the market expects rates to hold roughly steady or drift up over the longer run.

For buy-to-let landlords, the two-year-versus-five-year swap gap has a second, more mechanical effect on top of pricing: it changes how much a lender will actually let you borrow. Buy-to-let affordability is set by an Interest Coverage Ratio test — rental income must cover mortgage interest by 125% for basic-rate taxpayers and limited companies, or 145% for higher and additional-rate taxpayers — but lenders don't apply that test to the rate you're actually being offered. For a two-year fix, most lenders add a stress margin of roughly two percentage points on top of the pay rate before running the ICR calculation. For a five-year fix, many lenders instead stress-test at the fix's own pay rate, with no additional buffer, on the reasoning that five years of rate certainty removes most of the near-term repricing risk the buffer exists to cover. The practical effect is that when swap rates are elevated, a landlord can often borrow meaningfully more against the same rental income on a five-year fix than a two-year fix, even when the five-year rate itself isn't much cheaper — a detail that matters more to how much you can borrow than to what the deal actually costs.

Swap rates themselves don't move in a vacuum — UK government bond yields, or gilts, are the deeper driver underneath them. Gilts are the benchmark for risk-free UK borrowing costs across every maturity, and swap rates track them closely because both are pricing the same underlying thing: where the market expects interest rates and inflation to sit over that time horizon. When gilt yields spike — as they did through the second half of 2026 amid renewed inflation concerns and heavy government borrowing — swap rates spike with them, and fixed mortgage pricing follows within days. This is precisely why mortgage rates can rise sharply even in a month when the Bank of England hasn't moved its base rate at all: the pressure is coming from the bond market, not from Threadneedle Street.

September 2026 is a clean illustration of the whole mechanism working at once. The Bank of England's Monetary Policy Committee held Bank Rate at 3.75% on 17 September, its sixth consecutive hold, on a 6-3 vote — Huw Pill, Megan Greene and Catherine Mann again preferring an immediate quarter-point cut to 3.50%. That decision, on its own, changed nothing about swap rates, because the market had already priced a hold as the near-certain outcome beforehand. What moved swap rates in the surrounding weeks was the inflation and gilt-market backdrop the vote sat inside: August's CPI print at 3.1%, up from July, driven by energy and fuel costs; gilt yields sitting at multi-decade highs on renewed fiscal-supply concerns; and the Committee's own language judging the inflation outlook as more evenly balanced than in July, opening the door to a possible November cut. None of that is captured in the base rate figure itself, which is exactly why watching Bank Rate alone would have told you almost nothing useful about where fixed mortgage pricing was actually heading through September.

One thing worth being realistic about: there's no fixed, universal lag between a swap rate move and a lender actually repricing its mortgage book. Lenders generally respond to sustained trends rather than reacting to every intraday tick, which means a brief, reversed-within-days swap movement often passes through to high-street pricing partially or not at all, while a move that holds for one to two weeks is far more likely to show up in a lender's next repriced product range. A useful real example: after the Bank of England's rate cut in August 2025, five-year swap rates actually climbed by nearly a quarter of a percentage point over the following two weeks as markets reassessed the inflation outlook — and lenders responded by raising fixed rates shortly after, even though the base rate itself had just been cut. The lesson isn't a specific number of days to watch for; it's that a swap move needs to hold for it to matter, and once it does, lenders tend to follow within a couple of weeks rather than months.

For a homeowner or remortgager, the practical upshot is this: base rate headlines tell you what's already happened to short-term policy, not what your next fixed quote will look like — swap rate movement, which happens continuously and often ahead of the news, is the thing actually driving that number. Most lenders let you lock in a rate three to six months before your current deal ends, which means the swap rate on the day you apply, not the day your deal completes, is often what matters most. If you're watching for a good moment to lock in, tracking swap rate trends over the preceding weeks tells you far more than waiting for the next Bank of England announcement — by the time that announcement happens, the swap market has usually already told you what to expect.