In the first week of September, Halifax, TSB and HSBC all trimmed fixed mortgage rates while the wholesale costs those fixes are built on went sharply the other way. Moneyfacts’ Rachel Springall said on 2 September that swap rates “have risen dramatically over the past week, reaching 30-day highs”, and that the UK 10-year gilt yield had climbed above 5.2%, its highest since 2008.

Only one of those facts can win. Moneyfacts figures for 2 September put the cheapest remortgage two-year fix at 4.55% and the cheapest homebuyer two-year fix at 4.47%, against an average two-year fix of 5.59% on 3 September. Those best buys are, for now, priced below where the market says funding costs. Here is why that happens, why it never lasts, and why panic is still the wrong response.

What are swap rates and why do they set fixed mortgage rates?

A fixed-rate mortgage is a promise to take a set rate for two or five years whatever happens in between. That is a risk for a lender funding itself at rates that move, so lenders hedge it with interest rate swaps: contracts that exchange a floating stream of interest for a fixed one over a matching term.

The price of that contract is the swap rate, in effect the market’s forecast of average short-term rates over the term. A lender prices a fix by starting from the matching swap and adding its costs, risk margin and profit, so when the swap moves the poster rate moves with it. Gilt yields, the benchmark for long-term sterling borrowing, pull in the same direction: a 10-year yield at its highest since 2008 drags swaps up because both price the same expectation of rates staying higher for longer.

Why does Bank Rate matter less to a fixed rate than most people think?

Bank Rate has sat at 3.75% since 18 December 2025 and was held again on 30 July, yet fixed-rate pricing has moved throughout. Bank Rate is a today number; a fixed rate is a next-five-years number.

Fixes follow the market’s forecast of Bank Rate, which is far more volatile than Bank Rate itself. That is why UK mortgage rates went up in August while Bank Rate did not.

The forecast has shifted because the inflation picture has. Governor Andrew Bailey said on 30 July that inflation had fallen faster than expected but that the Middle East conflict meant high and volatile energy prices. Ofgem gave that worry a number on 26 August: the price cap rises 4% to £1,723 from 1 October, on wholesale gas up 11% in three months. That is what makes investors demand a higher yield.

Why would a lender cut rates into a rising market?

Three reasons, usually together.

The first is volume. The Bank of England’s Money and Credit release on 1 September showed 56,053 mortgage approvals for house purchase in July, the lowest since January 2024 and down from a revised 58,215 in June. A weak summer leaves lending targets short, and competition for a thin pool of borrowers drove the week’s cuts: Halifax by up to 0.11 percentage points for purchase and 0.13 for remortgage, TSB by up to 0.20 points across two, three and five-year products, HSBC on two- and five-year remortgage rates.

The second is timing. A cut announced on 1 September was almost certainly agreed on August swap rates; the lender has not yet caught up.

The third is hedging. Lenders buy swaps in tranches, in advance, to cover planned lending. A tranche bought before the move is locked at the old price, so fixes can be sold against it until it runs out. The next tranche costs what swaps cost now.

What were the cheapest deals on offer this week?

The best buy sits more than a full point below the average in every column, a reminder of why the advertised mortgage rate is not the rate most borrowers are offered. The best buys also carry the thinnest margin over the swap, so they go first.

Why does the window close so quickly?

When swaps jump, the deals closest to the old funding cost turn loss-making first. A lender at 4.47% has less room than one at 5.59%, so the market leader acts first, usually by withdrawing the product at short notice before it attracts applications the lender no longer wants at that price. That is why Moneyfacts urged borrowers on 2 September to secure a new deal quickly.

None of this is a prediction: swaps rose over one week and can fall back the next. The narrower point is that a lender cutting into a rising market is spending down a hedge or chasing a target, and both run out.

Does any of this mean a borrower should rush?

Here is the case against the headline. Nobody can time swap rates, including the lenders. A borrower who grabs a two-year fix because five-year swaps look expensive, while planning to stay put for a decade, has solved a small problem by creating a large one. Early repayment charges, or refixing in two years into a market nobody can foresee, will generally cost more than a rate that moved a fifth of a point during the decision. The term is the decision; the rate is only its price. A deal ending within six months can usually be reserved now and switched if pricing improves.

The 17 September decision deserves the same treatment. Whether the MPC holds at 3.75% or not, fixes may do nothing, because the swap market has already priced in what it expects. The decision that moves trackers is not the one that moves the two-year fix.

Frequently asked questions

What is the difference between a swap rate and Bank Rate?
Bank Rate is the rate the Bank of England sets today. A swap rate is the market’s price for exchanging floating interest for fixed over a set term, so it reflects where investors expect Bank Rate to average over that term.

Why did a rise in the 10-year gilt yield affect two-year mortgage rates?
Gilt yields and swap rates price the same expectations for inflation and interest rates, so a higher-for-longer view in gilts feeds straight into the swaps lenders use to price fixes.

If lenders are still cutting, is the market really getting worse?
The cuts in the week to 1 September were agreed on earlier swap levels to win volume after weak July approvals. They describe lenders’ targets, not where funding costs are heading.

Will fixed rates fall if the Bank cuts on 17 September?
Not necessarily. If a cut is expected, the swap market has already priced it in and fixes may not move. Only trackers and variable rates follow the decision directly.

The week’s rate cuts and the week’s rise in swap rates are not a contradiction; they are a lag. Lenders are selling last month’s funding costs into a market that has moved, and the deals nearest the edge go first. Worth knowing if a fix is ending soon, but not a reason to choose the wrong term in a hurry.

The Bank of England published its Money and Credit figures for July on 1 September, and two lines in it point in opposite directions.

Lenders approved 56,053 mortgages for house purchase, the lowest monthly total since January 2024 and well short of the roughly 59,500 that economists had expected. In the same month, net consumer credit, meaning unsecured borrowing on credit cards, overdrafts and personal loans, rose by £2.006bn, ahead of the £1.8bn forecast and the biggest monthly increase since November 2025.

Fewer people committing to a mortgage. More people borrowing without security. Those are not contradictory findings; they are two symptoms of the same thing, and it is worth understanding what that is before drawing the wrong conclusion from either.

What the figures do and do not say

Approvals are a forward-looking measure. They count mortgages agreed, not completions, so they show what buyers were deciding to do in July rather than what happened in the market months earlier. A fall in approvals means fewer people chose to commit that month.

Consumer credit is a net figure, meaning new borrowing minus repayments. A £2bn net rise does not mean households borrowed exactly £2bn; it means borrowing outpaced repayment by that much. On its own, rising unsecured credit is not automatically distress. It can reflect confidence, or spending brought forward, or simply more people putting large purchases on a card for the points. But a jump in unsecured borrowing in the same month that mortgage commitments hit a two-year low is a combination that deserves a closer look.

Why the two moved in opposite directions

Three explanations are doing most of the work, and they are not mutually exclusive.

Mortgage decisions are postponable, and everyday costs are not. Buying a house is one of the few large financial commitments a household can simply defer. Groceries, energy, insurance renewals and car repairs cannot be. When budgets tighten, the mortgage decision slides to next year while the shortfall in the current month goes on a credit card. That is the mechanically simplest reading of these two numbers, and probably the largest part of it.

Inflation ticked back up. CPI rose to 2.9% in July from 2.6% in June. That is a long way from the peaks of a few years ago, but it is moving the wrong way, and it lands on households who have already spent several years absorbing higher prices. Rising prices reduce the monthly surplus that makes a mortgage feel affordable, while increasing the gap that unsecured credit fills.

Affordability, not the headline rate, is the binding constraint. Bank Rate has sat at 3.75% since December, and lenders have been trimming fixed rates. Yet approvals fell anyway. That is a reminder of something we set out in detail when looking at why the advertised rate is not the rate you get offered: the rate on the poster is not what decides whether a purchase happens. Deposit size, the lender’s stress test and existing credit commitments decide it. And that last one matters here, because unsecured borrowing feeds directly into a mortgage affordability assessment.

The trap hiding in the combination

This is the part worth acting on, and it is easy to miss.

If you are planning to buy in the next year or two, the unsecured borrowing you take on now directly reduces what a lender will let you borrow later. Monthly commitments on credit cards, car finance and personal loans are deducted from the income a lender is willing to lend against. A £250 a month car finance payment can reduce your maximum mortgage by roughly £15,000 to £20,000, depending on the lender’s method.

So the household that defers the mortgage and fills the gap with credit can end up in a worse position next year than it was this year, even if its income has risen. The deferral was rational month to month, and the cumulative effect works against the original goal.

Two practical implications follow. First, if a purchase is genuinely still the plan, treat unsecured borrowing as borrowing against your future deposit and your future loan size, not just as this month’s convenience. Second, if you have existing balances, clearing or consolidating them before applying often does more for your borrowing power than adding the same amount to the deposit would.

There is a savings-side version of the same logic. Money earmarked for a deposit in the next two or three years usually belongs somewhere accessible and capital-secure, which is why the cash side of the ISA rules matters to buyers specifically, and why the proposed £12,000 cash ISA cap is worth understanding before April 2027 if you are saving hard toward a purchase.

What one month of data is worth

Not very much on its own, and it is worth saying so plainly. Monthly lending figures are volatile, get revised, and are distorted by seasonality and one-off effects. July also sits in the middle of the summer, which is never a representative month for property.

What makes this release worth reading is not the single month but the direction of two series at once. If August and September repeat the pattern, then a genuine shift in household behaviour is underway. If they do not, July was noise. The next Money and Credit release is the thing to watch rather than this one, and the Bank’s next rate decision after that.

Frequently asked questions

Does a fall in approvals mean house prices will drop? Not necessarily, and not quickly. Approvals measure demand for purchase mortgages, and prices respond to the balance of demand and supply, which includes cash buyers and how many sellers actually list. Weaker approvals reduce upward pressure rather than guaranteeing falls.

Is rising consumer credit a sign of trouble? It can be, but the figure alone does not tell you. The same £2bn could be people financing holidays confidently or people covering bills reluctantly. Arrears and default data, published separately and with a lag, are the better distress signal.

Should I delay buying because approvals are falling? The wrong question. Other people’s decisions do not change your affordability. Whether the numbers work for you, at the deposit and rate you can actually access, and whether you intend to stay long enough to absorb transaction costs, is what decides it.

Does clearing a credit card really improve my mortgage chances? Generally yes, and by more than most people expect, because lenders deduct the monthly commitment from the income they lend against. Reducing balances also helps your credit profile, though avoid closing long-held accounts immediately before applying.

The useful takeaway from July is not that the market is weak or that households are struggling, because one month cannot establish either. It is the reminder that these two decisions are connected. The credit you take on while waiting for the right moment to buy is quietly changing what will be possible when that moment arrives.

Late August produced two mortgage numbers that both describe the same market and look nothing alike. The lowest two-year fixed rate being tracked was 4.39%. The average two-year fixed rate was 5.07%. On a £200,000 mortgage that difference is roughly £75 a month, and the borrower paying the higher figure has usually done nothing wrong.

That gap is the most misunderstood thing in mortgage pricing. Best-buy tables and lender press releases quote the floor of the market, then the quote lands and it starts with a five, and the natural conclusion is that you have been treated badly or that the deal was bait. Usually neither is true. The advertised rate is real, it is just conditional, and four specific conditions decide whether you reach it.

What the market actually looks like right now

Both averages slipped by 0.01 percentage points week on week, and Nationwide cut selected two, three and five-year fixes by up to 0.15 points with effect from 18 August, taking its lowest fixed rate to 4.48%. So the direction of travel is gently downward, even though inflation rose in July and a Reuters poll found nearly 90% of economists expect Bank Rate to stay at 3.75% through 2026.

Those figures are advertised rates rather than quotes, and they are averages across the whole market. Note the last row in particular: a first-time buyer with a 5% deposit is looking at 5.34%, almost a full percentage point above the headline 4.39%. Same week, same market, same lenders.

The four things that decide where you land

Loan to value does most of the work. Rates step down in bands, typically at 90%, 85%, 80%, 75% and 60%. The very best rates in the market almost always require 40% equity. A borrower at 95% and a borrower at 60% are not being offered slightly different versions of the same deal; they are shopping in different price tiers, and the spread between the top and bottom band is routinely a full point or more.

Fees convert a cheap rate into an expensive one. A market-leading rate frequently carries a product fee around £999, sometimes more. On a small balance that fee can swamp the interest saving entirely, which is why the true comparison is total cost over the deal period rather than the rate. We ran the arithmetic on this when the gap between two and five-year fixes had almost vanished, and the shape of it holds here: the cheaper rate can leave you worse off.

Affordability decides whether the tier is available at all. Income multiples, outgoings, dependants, existing credit commitments and the lender’s stress test determine the maximum loan. A borrower who fits comfortably at 4.5 times income has the whole market; a borrower stretching to the edge finds the cheapest lenders decline before the rate is ever discussed.

Your credit file sets the floor. Missed payments, defaults and CCJs move you out of mainstream pricing altogether, and the specialist lenders who will consider the case price for the risk. For those borrowers the entire best-buy table is theoretical, which is the point we made in what a broker actually said about mortgages with missed payments: the first question is who will lend at all, and the rate menu follows from that answer rather than leading it.

The one lever most borrowers can actually pull

Three of those four are largely fixed by the time you apply. Loan to value is the exception, and it is worth checking whether you are sitting just above a band boundary.

If your loan is 81% of the property value, finding the extra 1% either through savings or through a higher valuation moves you into the 80% band and onto a visibly better rate for the whole term. On a £200,000 property, moving from 81% to 80% means finding about £2,000. Against a rate improvement that might be worth £20 to £40 a month across a five-year deal, that arithmetic often works comfortably in your favour.

It is worth checking at every renewal too, not just at purchase. Between capital repaid and any price growth, borrowers routinely cross a band without noticing and stay in the tier they were in five years earlier because nobody re-ran the number.

Frequently asked questions

Is the advertised rate a con? No. It is a real product that real borrowers get, and lenders are required to be clear about eligibility. It just describes the best-case borrower: large deposit, clean file, comfortable affordability, and usually a product fee attached.

Why did two lenders quote me differently on the same day? Because criteria differ more than pricing does. Lenders take different views on bonus and overtime income, self-employed accounts, recent job changes, existing debt and property type. The lender whose rules happen to fit your circumstances will look cheapest, and which one that is changes case by case.

Should I wait for rates to fall further? Nearly 90% of economists expect Bank Rate to hold at 3.75% through 2026, and fixed rates move on swap rates and lender competition rather than Bank Rate directly, so waiting is a bet rather than a plan. The reliable move is to book a deal early, since most lenders let you secure one three to six months ahead and many allow a switch if pricing improves before completion.

Does a broker get better rates than I can? Not usually a better version of the same product. What a broker does is know which lender’s criteria fit your case, which is what determines whether you reach the good tier at all. That matters most for exactly the borrowers the best-buy tables serve worst.

The useful reframe is to stop reading the headline rate as a price and start reading it as a qualification standard. It tells you what the market charges someone with a big deposit, a clean file and room to spare. Work out honestly which of those four you fall short on, and you will know both why your quote looked different and which one is worth doing something about before you apply.

The cash ISA as most people know it is on borrowed time. Under draft regulations published by HMRC on 16 July 2026, anyone aged 64 or under will only be able to pay £12,000 a year into cash ISAs from 6 April 2027. The overall £20,000 ISA allowance survives, but for younger savers the remaining £8,000 will only fit inside a stocks and shares or innovative finance ISA. Savers aged 65 and over keep the full £20,000 cash allowance.

That is the headline, and it has produced a predictable mix of outrage and panic. Both are worth examining, because the details of the draft say something different from the headlines: most savers will not hit the new cap at all, the people who will are exactly the ones with the best reasons to hold cash, and there is still a full tax year to plan around it.

What do the draft rules actually say?

The proposal is a technical consultation on The Individual Savings Account (Amendment) Regulations 2026, published on 16 July and now closed. It is draft secondary legislation, which means it is not yet law: the regulations still have to be made before any of this takes effect. The policy direction was announced at Autumn Budget 2025; the July draft supplies the mechanics.

Two rows of that table matter more than the headline. The transfer ban means an under-65 cannot route around the cap by subscribing to a stocks and shares ISA and moving the money across, and it also removes a genuinely useful option: shifting invested money into cash when your circumstances change. And the proposed charge on interest earned on cash held inside investment ISAs is there to stop the obvious dodge of leaving the money uninvested. The drafters have thought about the workarounds.

Who actually hits the £12,000 cap?

Fewer people than the outrage suggests. The parliamentary petition against the change makes the point itself: the average annual cash ISA subscription is around £7,000, comfortably under the new cap. If you drip £300 a month into a cash ISA, nothing about April 2027 affects you, and your existing balances are untouched either way.

The people who do hit it are a specific group: savers moving £15,000 to £20,000 a year into cash. That describes someone building a house deposit over two or three years, someone in their fifties and sixties de-risking ahead of retirement, and anyone parking the proceeds of a house sale or inheritance while they decide what to do. Critics, including building societies, have argued this is precisely the money that should not be pushed towards investment risk, because it has a short time horizon and a fixed purpose. The age split has drawn particular fire: a 65-year-old keeps the £20,000 cash allowance while a 65th-birthday-in-May saver does not, which is hard to defend as anything but arbitrary.

The Treasury’s counterargument is that Britain holds too much long-term wealth in cash, and that £12,000 a year of new cash saving is still generous. Both things can be true. The policy is aimed at money that sits in cash for decades; the collateral damage lands on money that sits in cash for three years for a good reason.

What should savers do before April 2027?

The current tax year and the next one, 2026/27 running to 5 April 2027, are the last under the old rules if the draft goes through unchanged. That makes the planning straightforward rather than clever.

If you are a heavy cash saver, use the full £20,000 cash allowance while it exists, and remember the cap restricts new subscriptions, not balances: money already inside cash ISAs stays there, keeps its tax wrapper, and can still be transferred between cash ISAs for better rates. The rates themselves are also cooperating for now. According to Which?, the best one-year fixed cash ISA paid 4.91% in early August, with longer fixes touching 5%, and NS&I lifted its fixed bonds to between 4.82% and 4.85% on 19 August. With the Bank of England holding Bank Rate at 3.75% in July by six votes to three, and the three dissenters voting for a rise, nobody should assume today’s rates survive the winter, a dynamic we covered when we looked at why rates moved the way they did this year.

Two cautions before anyone stuffs every spare pound into cash. First, this is still a draft: consultations produce amendments, and the final regulations could soften the cap, change the age rule or slip the date. Acting on the parts that benefit you anyway, like using an allowance you were going to use, is sensible; reorganising your finances around an unmade law is not. Second, filling a cash ISA is only the right move if cash is the right home for that money. If it is a ten-year pot, the argument for investing it existed before this policy and exists after it. The same logic we applied to the two-year versus five-year fix decision applies here: match the product to your actual time horizon, not to a rule change.

Frequently asked questions

Does the £12,000 cap affect money already in my cash ISA? No. The cap applies to new subscriptions from 6 April 2027. Existing balances keep their tax-free status in full, and transfers between cash ISAs remain allowed, so you can still chase better rates on old money.

Can I put £12,000 in cash and £8,000 in stocks and shares? Yes. The overall £20,000 allowance is unchanged; only the cash portion is capped for under-65s. The draft blocks the reverse route, moving invested ISA money back into cash, for under-65s.

Is this definitely happening? Not yet. The regulations were published in draft on 16 July 2026 and the consultation has closed, but they have not been made law. The direction is clearly signalled, the details could still move.

Why are over-65s exempt? The government’s logic is that older savers legitimately need capital security. Critics call the cliff-edge arbitrary, and it is one of the most challenged features of the draft.

The cash ISA cap is that unusual thing: a policy that will genuinely affect only a minority of savers, wrapped in a headline that alarms all of them. Work out which side of the £12,000 line your actual saving habits fall on, and most of the anxiety resolves itself. If you are under it, carry on. If you are over it, you have until April 2027 and one full allowance year to arrange things on your own terms, which is more notice than savers usually get.

For years the fixed-rate decision came with a price signal attached. Committing for five years cost meaningfully more, or meaningfully less, than committing for two, and that gap told you what the market believed about where rates were heading. You could disagree with the market, but at least it had an opinion, and it charged you for taking the other side.

Right now, it barely has one. On Moneyfacts averages from late August, a typical two-year fixed rate sits at about 5.61% and a typical five-year at about 5.64%. Three basis points. On a £200,000 mortgage that is roughly £5 a month. Those are averages of advertised rates rather than quotes, but the shape is the point: the market is charging you almost nothing to choose either way.

That sounds like good news, and in one sense it is. But it also quietly removes the crutch most borrowers leaned on. When one option is clearly cheaper, you can tell yourself the decision made itself. When the pricing is flat, the decision is entirely yours, and it has to be made on something other than price.

Why the gap closed

We looked at why mortgage rates moved the way they did this year in a previous piece, and the short version bears repeating: fixed rates are not priced off the Bank of England base rate. They are priced off swap rates, which is the cost lenders pay to lock in money for a set period, and swaps reflect where the market expects rates to be over that period.

When two-year money and five-year money cost a lender about the same, the market is saying it expects rates over the next five years to average out at roughly the level of the next two. No steep cuts priced in, no fresh spike priced in. Base rate has now been held at 3.75% for five consecutive meetings, with the next decision due on 17 September. Flat expectations, flat pricing.

The market has been wrong before, on both sides, and anyone who fixed in early 2022 or refixed in late 2023 knows it. But “the market expects roughly nothing” is the honest starting position, and it means the useful question is no longer which fix is cheaper. It is which fix fits your life.

The bet you are making either way

Strip away the jargon and the two products are two different bets, and it helps to see them side by side.

The last row is the one people skip, and it is usually the one that ends up mattering. Early repayment charges on a five-year fix commonly start around 5% of the balance and step down each year. On a £200,000 loan that can mean £10,000 to leave in year one. Most five-year products are portable, meaning you can in principle carry the rate to a new property, but porting is an application, not a right. The lender reassesses you at the time, and if your circumstances have dipped, the port can be refused and the charge lands anyway.

So the five-year fix is not really a bet on interest rates. It is a bet on your own life staying still: same house, same relationship, same income shape, for five years. Some people can make that bet comfortably. A lot of people in their twenties and thirties honestly cannot, and a slightly cheaper monthly payment is poor compensation for a five-figure exit fee.

The choice that is always wrong

While the two fixes are finely balanced, there is a third option that is not balanced at all, and that is doing nothing. When a fixed deal ends and no new deal replaces it, the loan rolls onto the lender’s standard variable rate, and the average SVR is currently about 7.13%. Against either fix, that is roughly £180 a month more on a £200,000 repayment mortgage from the day it happens.

People lapse onto the SVR for two reasons. Some simply miss the date, which is fixable with a calendar reminder six months out, since most lenders let you book a new deal three to six months before the current one ends. Others freeze because they cannot decide between products, which is the expensive irony of this whole subject: agonising over a £5-a-month difference while paying £180 a month for the privilege of not choosing.

If your credit file is not clean

One honest caveat. Everything above assumes the market is open to you, and if the last couple of years left missed payments or worse on your file, the fix-length debate is a luxury that comes second. As we covered in what the broker said about mortgages with missed payments, the meaningful question for those borrowers is which lenders will have them at all, and the product menu follows from that answer rather than leading it. Get approved first, optimise fix length within whatever that lender offers, and treat the full market as something to come back to at the next renewal, when the file is two years cleaner.

Frequently asked questions

Is a three-year fix a sensible compromise? Sometimes. Fewer lenders offer them, so the pricing is often slightly worse than the market’s best two- and five-year deals, but if 2028 lands awkwardly for you, a 2029 renewal date can be worth paying a little for.

Should I wait for the September decision before fixing? You can usually have it both ways. Because offers can be booked months ahead, you can secure a deal now and, with many lenders, still switch if a better rate appears before completion. Waiting with nothing booked is the only version of this with real downside.

Do trackers make sense while the gap is this small? Trackers suit people who genuinely might repay or move at short notice, since many carry no early repayment charge. As a pure rate bet they only win if cuts come faster than the market expects, which is exactly the bet the flat fix pricing says the market is not making.

What actually decides it, then? Your honest five-year horizon. If you can say with a straight face that you will be in this property, with this mortgage, in 2031, the five-year fix buys certainty at almost no premium. If you hesitated while reading that sentence, the two-year fix is the price of keeping your options open, and right now that price is about £5 a month.

The vanishing gap between two and five-year money will not last forever. When it reopens, the market will go back to nudging you toward one answer. While it lasts, the decision is unusually pure: not a rates call, just an honest look at how still your life is likely to stand. That is a harder question than reading a comparison table. It is also the only one worth answering.