Book workshop tickets: www.tinyurl.com/pbwoct26
"Budgets are not merely affairs of arithmetic, but in a thousand ways go to the root of prosperity of individuals, the relation of classes and the strength of kingdoms." - William Ewart Gladstone, 1859
The quote pertains to a week in which a Chancellor stood up in Coventry and pledged fiscal restraint, and the bond market spent the following four days taking the thirty-year gilt to its highest yield since 1998. Gladstone understood something that gets forgotten in the run-up to every Budget: the arithmetic is the easy part. What a Budget signals about the relation of classes, and about the strength of the kingdom, is what the people lending us the money are actually pricing. We are seven weeks out from the red box, and they have started pricing already.
As we run into the last quarter of a year that has had more in it than most, our next Property Business Workshop is live and tickets are moving. Thursday 1st October, the Hyatt Regency in Manchester, with myself and Rod Turner, powered by Roma Finance, who do bridging, development and term lending and who know what they are talking about. Early Bird is still on. The VIP dinner has two seats left at the time of writing, and I would not hang about on that one. Book in at www.tinyurl.com/pbwoct26
Welcome back to Trumpwatch. Last week we had Warsh's first Jackson Hole and a central banker telling us, in his own understated way, that there was work to do. This week we got the other kind of monetary policy. The kind that arrives by tweet.
Five thousand dollars. Per adult citizen. If the Republicans hold both houses in November.
Let me be careful with the language here, because the word that wants to come out is "bribe" and I would rather earn it than just throw it. A government spending money on its citizens is not a bribe. A government spending money on its citizens conditional on how they vote, announced as a conditional, is something else. That is the bit I keep coming back to. It is not "we will do this because it is right". It is "we will do this if you do this". Stated out loud. In public. With a number attached.
The costings, and these are other people's numbers rather than mine: Kent Smetters at Penn Wharton puts the hole at around $1.15tn. Others are quoting $1.3tn. Gross federal debt is somewhere around $40tn, and the Congressional Budget Office has this year's deficit at $2tn before anybody promises anybody anything. So the cheque, if it is ever written, is roughly 2.9 per cent of the entire national debt, or 55-65% of the annual deficit, added in an afternoon. Let's be as charitable as possible - as usual. The Treasury will get some cashback on that money - maybe 30% or even 40%. Should that change our attitude towards it though? Probably not.
Put that in the frame that matters to us. The Tax Policy Center reckons tariffs raise about $1.7tn across fiscal 2026 to 2036. Good luck with forecasting those ten minutes out, let alone ten years out, would be my take. The dividend is three quarters of eleven years of highly theoretical tariff revenue, spent in one.
Now. Is this more overt than political promises have ever been? Yes, I think that’s undeniable, and I do not think it is close. Governments have always bought votes. They have generally had the decency to launder it through a policy document first. A winter fuel payment, a tax threshold, a capital programme in a marginal seat. The dressing-up was the tell that somebody, somewhere, still thought the electorate would object to the naked version.
But is it therefore worse than the laundered kind? I genuinely do not know, and I have been turning it over all week. There is an argument, and it is not a stupid one, that the honest bribe is the better bribe. At least you can see it. At least it is priced. The billions buried in annex C of a spending review are the same money doing the same job with none of the scrutiny.
What I do know is this: politics needs less lying. If the choice is between being bought openly and being bought quietly, I will take openly, and then argue about whether I want to be bought at all. Which I do not. With two chunky donations to Reform announced within 48 hours, there is a concern that the UK is going the same way as the US - but our citizens just don't roll the same way, in my view - we’ve seen the divergence when Farage attempted to sound Putin-sympathetic, and immediately dropped like a stone in the polls. At the very least that was a little geography lesson for him - at best it was a reminder that we are the United Kingdom, NOT the United States.
The reason any of this sits here in the Supplement rather than a US politics one is the same reason it always is. Bond markets do not respect borders.
A trillion-dollar+ promise is a trillion dollars+ of issuance, or it is inflation, or it is both. Markets do not wait to find out which. The US ten-year and thirty-year moved on it, and by the time a US long-end move has finished being a US long-end move, it has already arrived here.
And it arrived in a week that was already ugly for other reasons. Inflation, expected to remain high in the US, did. There's now officially a half-percent gap between UK and US inflation, the other way around than usual (traditionally we “run hotter” than the US, and have a higher inflation figure - and we have done very much so over recent years).
How about the oil price, which is doing the actual work as we've been discussing (or some of it anyway)? Well, the show goes on. The United States and Iran resumed strikes. Brent closed at $101.21 on Wednesday, its highest close since 22 May, and went through $108 dollars on Thursday. It's $104.60 as I write this. It is up more than 8% in September alone, and it's still early.
I want to be careful about sequencing here, because it matters for what you conclude. There is a temptation, and I felt it myself when I saw Friday's growth print, to tell this week's story as "strong economy, higher yields". That is not the order it happened in.
The thirty-year gilt touched 5.948 on Tuesday, which is being reported as the highest since 1998. The ten-year got to 5.38, the highest since 2007. Sixty-nine of the seventy conventional gilts in issue saw yields rise. All of that was before Friday morning, and all of it was happening while Brent was climbing towards $108.
So: the energy shock is the engine. The growth surprise is the accelerant. Getting that the right way round is not pedantry, because the two resolve differently. An oil-driven inflation expectation unwinds if Hormuz calms down. A growth-driven one does not.
Goldman is warning Brent could go above $120 in 2027 if Gulf output stays four million barrels a day below pre-war levels. I would file that under plausible but unproven, which is where I file most 2027 oil calls, including my own - I don't make them, because it isn't my thing - I just speculate as a layman and inject my usual dose of common sense. In the same way that I look at the yields when they are at 30+ year highs and decide today isn't the day I speculate on the future of bond yields with authority - ranges are ranges, but also the market can stay irrational longer than you, or I, can stay solvent. Remember that. The market can stay irrational longer than the US Government can remain solvent - just because Soros broke the Bank of England in the 90s doesn't mean someone today couldn't break the Fed, and they'd be arrogant enough to presume it couldn't happen - until it did - in my view.
Our own Chancellor gave his first big speech on Monday, in Coventry, and I am going to spend proper time on it in the Deep Dive because it deserves it. But the Trumpwatch-relevant part is the framing. John Healey stood up, pledged fiscal restraint, said explicitly that he was building on his predecessor's work to re-establish credibility in the bond markets, and talked about growth in every postcode.
The bond market's reply, over the following four days, was to take the thirty-year to a twenty-eight-year high. I am not trying to ascribe that to him. The oil is the oil and the Americans are the Americans, and a new Chancellor gets about ninety seconds of grace in a world where Brent is doing what Brent is doing. But if you are looking for the single most uncomfortable chart in British politics this morning, it is a Chancellor's credibility speech with four days of gilt yields plotted after it.
Standard Fare, in one sense. Every Chancellor discovers the bond market eventually. This one has discovered it in week one of parliament being back to business.
Phew. Stepping away from the transatlantic weather, let us go and look at what all this is doing on the ground - how about our comfort blanket of the real time UK Property Market?
As is customary, Chris Watkin has been relentlessly crunching the portal numbers. No stats show this week because he has been away, and quite right too, so what we have instead is his regional pack for Week 35 of 2026, TwentyEA data crunched by Denton House. Eleven regions. And I will be honest, in a week like this one the regional cut is more useful than the national headline anyway.
Chris flagged the caveat before I could: last week was a Bank Holiday, so the numbers will be down. They are. But they are down in a specific and quite interesting shape, and the shape is the story.
Across the eleven regions Chris sent me, 28,916 new homes came to market against a combined 2026 weekly average of 32,995. That is 87.6% of a normal week. Under, then, but only by about an eighth - and when you lose a fifth of a week, that can't be too bad. This is a week where we expect stock to come to market, however.
Every single region came in below its own average bar one. Hold that thought, because it is Inner London, and we are coming back to it.
The ready reckoner, for those who have followed it through its various humiliations this year, is parked this week. You cannot read a supply signal off a holiday week and I am not going to pretend otherwise.
16,529 sales agreed against a combined average of 21,830. That is 75.7% of a normal week. Down by a quarter.
And there it is. Listings down an eighth on the yearly average, sales down a quarter. The Bank Holiday took twice as much out of the selling as it took out of the listing.
Which gives us the number I actually care about this week - so, how about the friction arithmetic?
In a normal week across these eleven regions, the market produces about 1.51 new listings for every sale agreed, which is 32,995 divided by 21,830. In Week 35 it produced 1.75, which is 28,916 divided by 16,529. Arguably - it might be "that sort of week" when back from the Summer Holidays - but it's enough to flag it.
That is roughly a 16 per cent deterioration in said ratio in a single week. Not because anything changed about the market. Because everybody took Monday off, and the bit of the process that needs two solicitors, a chain and a mortgage broker took the whole week off, while the bit that needs one homeowner and a photographer did not. Or because holidays are over and it's listing time for those who've made a decision over the summer (or been swayed by a big, juicy overinstruction possibly - too cynical?)
So the reckoning is this: a Bank Holiday week is a stock-building week. Roughly 5,300 sales that would normally have been agreed were not agreed, while roughly 4,100 listings that would normally have appeared appeared anyway. Net, the market put on stock it will now have to work off in September, in a month where the mortgage pricing has just got worse. More on that in the gilts section. Rough working, and I would not put it in front of a statistician, but the direction is not in doubt.
Here is the bit I did not expect to find when I opened the pack. Scotland sold 1,964 against an average of 2,010. That is 97.7 per cent of a normal week. Barely a scratch.
Every other region: between 66 and 78 per cent of average.
Scotland takes its summer Bank Holiday on the first Monday in August. England, Wales and Northern Ireland took theirs on 31 August, which sits squarely inside Week 35. Scotland worked a full week. Everybody else worked four days. So that's an accurate gauge of the temperature - but they've also been back to school for weeks. I might stop looking for something that isn't there, one day - but not quite yet.
Strip Scotland out and England and Wales ran at 73.5% of average. Scotland ran at 97.7%. I cannot remember the last time the portal data handed us a natural experiment that clean, and I want to be careful about what it spells out and what it does not. It does not tell us the market is strong. It tells us that the entirety of this week's national shortfall is explained by one Monday. If you see a headline in the next fortnight about a collapse in sales agreed at the end of August, you now know exactly what it is looking at, and you know the person who wrote it did not check the calendar. We will talk more on this next week, I’m sure.
Scotland's listings-to-sales ratio, incidentally, actually improved slightly: 1.09 against a normal 1.12. Everybody else's got worse. The control group behaves like a control group. Who'd have thought it? How about Inner London, which went the other way? Inner London sold 819 against an average of 1,239. That is 66.1%, the weakest of the eleven alongside the North East. And Inner London listed 3,124 against an average of 3,107. The only region in the pack that listed above its own average, in a four-day week. Again, though, we do see this at this point in the calendar every year - and if Chris wasn't away, he would be confirming it himself, I have no doubt!
Let’s run the ratio anyway though. How overweight is it? Inner London normally produces about 2.51 listings per sale. In Week 35 it produced 3.81. That is roughly a 52% deterioration in a week, against a national 16. The rest of the Inner London card reads accordingly. Chance of selling 34.63% against a national average of 55.5. Average days to sell 94 against a national 75. And the price comparison ought to stop anybody who still thinks of central London as the safe end of the market: minus 5.0% against 2019, and minus 5.7% against 2025. Not real terms. Nominal. Seven years, and the number in front of it is a minus sign. Do you want me to tell you just how far above 30% RPI inflation has been in that time? Probably not, if you own anything there.
I have been saying for a while that the London discount would show up in the flow data before it showed up anywhere anyone was willing to print it. I might well be wrong about the cause, and I am fairly sure I was early on the timing. I often am (please don't snigger, it does not become you). I do not think I am wrong about the direction.
Chris's pack gives us the average asking price on new listings and the average asking price of what went sold subject to contract. The spread between them, across the eleven:
North East 7.5%, at £224,958 listed against £208,162 SSTC.
East Midlands 8.3%
Outer London 10.3%
Wales 10.6%
West Midlands 12.0%
Scotland 12.3%
North West 14.3%
South West 14.6%
South East 17.5%
Inner London 18.6%
East Anglia 20.5%, at £499,763 listing against £397,427 SSTC.
Two health warnings, and I am going to give them properly rather than in a footnote, because this number gets misused constantly, including by me.
First, these are different baskets. The listing figure is what came to market this week. The SSTC figure is what got agreed this week, which is mostly stock that has been sitting for weeks or months. This is not the discount on a given house. It is the gap between what sellers are asking today and what the market is actually clearing.
Second, and this follows: do not cross-compare this with the old intergalactic series. Different metric, different construction. I said a fortnight ago that we needed to settle on one reader-facing definition and stop switching, and I have still not done it. That is on me, and it is going in the diary. I'll get there - I just like to think it through. A lot, sometimes!
What the spread does tell you, cleanly, is where the optimism lives. It rises almost monotonically with price. The North East asks £225k and gets £208k. East Anglia asks £500k and clears at £397k. The expensive end of this market is where the fantasy is concentrated, and East Anglia at 20.5% is now carrying more of it than Inner London.
The full spread, against a national average of 55.5 per cent and 75 days: Scotland 80.84 per cent and 47 days. North East 61.99 and 62. North West 60.75 and 68. West Midlands 59.31 and 74. Wales 58.92 and 88. East Midlands 55.98 and 81. South West 53.20 and 84. East Anglia 50.96 and 78. South East 49.80 and 81. Outer London 46.82 and 80. Inner London 34.63 and 94.
Wales is the one worth a second look. Nearly 59% chance of selling, comfortably above national, but 88 days to do it, which is thirteen days slower than national. Welsh houses sell. They just take their time about it, and a seller who prices for a fast market there will spend the autumn wondering why nothing is happening. Is that a Welsh problem or a rural problem? I suspect the latter, but I would want the sub-regional cut before I said so with any confidence.
And Scotland again: 80.84% and 47 days. Even allowing that the Scottish system is structurally faster, that is a different market from the one the rest of us are operating in.
The £64k question. Which is this. If a four-day week costs the English and Welsh market a quarter of its sales and only an eighth of its listings, what does that market then do with the stock it has just accumulated, in a month where the five-year swap has moved 31 basis points against it?
I think it reduces. I think September's price reduction count is going to be a big one, as the withdrawal numbers will be as the agents clear out the dead wood in the runup to Xmas (yes, I said it. I have to) and I think the asking gaps above are going to widen before they narrow. I would not put real money on the size of it. I would put real money on the direction.
Chris - this is my weekly appreciation paragraph. Thanks for what you do! If you want some help positioning yourself as a local market expert - as an estate agent or any form of property professional - give Chris a shout! Either way give his channel www.youtube.com/@christopherwatkin a follow and some love, please! If you google (or AI search) "Chris Watkin Property Market Stats Show [YOUR INVESTMENT AREA]" there will be an informative one on your area of interest, somewhere along the line - you can bank on it. Worth an hour of your time.
Stop. Drop the mic. Macro time. We have the Lloyds house price index, which has just done something it has not done since 2023. We have the RICS residential survey, which has done something rather strange in the rental section that has been one way traffic for half a decade. We have July's growth figure, which is the reason half the City spent Friday rewriting its notes. And bringing up the rear, as the contract requires, the gilts and the swaps, which this week are less of a closing item and more of the thing the rest of this newsletter has been circling since the first paragraph.
Lloyds: the first annual fall since November 2023. August: down 0.2% on the month, down 0.4% on the year, average price £298,468. That year-on-year number is the one that got the news coverage, and rightly. It is the first annual fall since November 2023. Nearly three years of the index grinding out positives, and it has stopped.
The underlying detail is where it gets more interesting than the headline, as it usually is. Regionally: Northern Ireland up 6.9%, the South East down 1.6%, Greater London down 1.5%. So the index is not falling. Bits of it are falling, hard, and other bits are still running near 7%, and the average of those two things happens to land slightly below zero. That is a very different market from a market that is going down in a 2009 slump, for example.
And look at how neatly it lines up with Chris's flow data above. The regions with the worst price comparisons in the TwentyEA cut are Inner London, Outer London, the South East and the South West. The regions holding up are the North East, the North West and the West Midlands. Two entirely different datasets, different methodologies, one week apart, telling the same story about the same fault line. A story we've been talking about for years now in the Supplement.
I have been saying for years that the national house price index is the least useful number in British property, and this month is about as clean a demonstration as I could have asked for. Minus 0.4% describes precisely nobody's house, tells a contrary story to Nationwide and the ONS, and is good for headlines-only I think.
Worth noting for the pedants, especially as I am the PresiPedant of all: this is the Lloyds index, formerly Halifax, rebranded from July. Same series, new badge. The Howard Brown era is properly over, as I said. No longer does he give you extra.
How about the RICS residential market report? Less bad on almost everything, and then the rental section..... The August survey landed on Thursday and the activity picture is genuinely improving. Not good. Less bad, consistently, across nearly every measure.
New buyer enquiries minus 19, the least negative reading since January. Agreed sales minus 17, which sounds grim until you remember it was minus 38 in April. New instructions at zero, up from minus 2. Market appraisals minus 17. Sales expectations at three months minus 3, from minus 13, and at twelve months plus 6, from plus 3.
Prices, though, are still going the other way. House price balance minus 28, essentially unchanged from minus 29 in July. More surveyors seeing agreed prices fall than rise, and expecting that to continue over three months before steadying out over a year. That combination is the market I recognise. Activity slowly returning, prices still under pressure. It is what happens when the binding constraint is affordability rather than confidence: people come back to the market, they just cannot pay what the last lot paid. Affordability is down 9% since January, and will get a bit worse, thanks to the bond yields (this is almost the exact reverse of the improvement in affordability that came in 2025 when Reeves told the lenders to read the FCA stress test rules more closely).
And then the rental numbers, which made me go back and check I had read the right month. Tenant demand plus 18. Landlord instructions minus 14. Rent expectations plus 44. I flagged in the middle of August that the July survey had tenant demand turning negative for the first time since 2020, and I said at the time that it was either the most important number of the year or a blip. One month later it is plus 18 and rent expectations have jumped to plus 44. One more reminder, if you need one (surely I say it every week?) that we don't make a decision on one month's data.
I will be honest: I do not believe a diffusion index swings from negative to plus 18 in a month because the underlying reality swung. Diffusion indices are moody. Something in there is noise. But I said in August that I would not build a thesis on one print, and I am not going to build the opposite thesis on the next one either. What I would say is that the July reading now looks a good deal more like an outlier than a turning point, and anybody who wrote "the rental crisis is over" off the back of it, and I am fairly sure a few did, should be feeling slightly warm around the collar. Business as usual? Or August is always busy, and this one is no different? Our own voids are down, and that's also the case for the Boardroom Club members too - we enjoyed an excellent face to face meeting this week.
Let the pedants, like me, argue about the methodology. The landlord instructions number at minus 14 has been negative for so long it has stopped being news, and it is the one I would actually watch. Supply is still leaving the agents' books or coming in-house/sitting with the bigger operators.
Onto GDP and Growth: the Burnham Bounce, presuming that's what it is.... July growth came in at plus 0.4% on the month, against a consensus of zero. Not a small miss. A complete one. The biggest I can remember for some time. Three months to July plus 0.4, the eighth consecutive three-month-on-three-month gain. Annual growth plus 1.6%. Sad to say that's HUGE in today's world, even though we need to aim higher.
Now. Where do we want this figure to be? Obviously up. I am not going to be the man who complains about growth. But the reason yields went up on Friday rather than down is that this is the wrong kind of beat for a Chancellor three days into pledging fiscal restraint and seven weeks out from a Budget, and everybody in the gilt market worked that out inside about four minutes.
The composition is where it lives. On the month: services plus 0.4, production plus 0.2, construction plus 0.1. On the three months: services plus 0.6, while production and construction each fell 0.5.
Read those two lines together. Services are carrying the whole thing - as they do every time. Great for services? Great for the overall figures. The making-things part of the economy and the building-things part of the economy are both shrinking over the quarter, and the only reason the headline is positive is that we are very good at the main one (which, arguably, is why it IS the main one).
The construction detail, though, which is the bit that matters most to everybody reading this? Construction output in July was plus 0.1% on the month, and the statisticians say the increase came solely from repair and maintenance. Solely. New work contributed nothing.
So the roaring economy, in the sector that is supposed to solve the housing shortage, consists entirely of fixing things that already exist. We are not building our way out of anything. We are patching.
Is the Burnham Bounce real? Something is clearly going on in services, and it would be churlish to pretend otherwise. Whether it survives a winter of $100 oil and a Budget on 28th October, I would not like to say. My honest read is that this print is stronger than the economy underneath it, and that the revisions will be interesting - as will the ebbing backwards from the bounce. I have said that about enough prints to be wrong occasionally.
Moving forward - brace yourselves - Gilty, or not Gilty. To the item that has been sitting in the corner of this newsletter since paragraph one. The five-year gilt closed the week at 4.902%. Last week's close was 4.630%. That is plus 27.2 basis points in a week, and it leaves the five-year sitting 5.5 basis points below its fifty-two-week high of 4.957% (which happened midweek). That's way outside of the normal range of movements, and uncomfortable, and in the wrong direction. I believe that justifies the Black Armbands. Don them accordingly.
The thirty-year is at 5.910%, against 5.775% last week. Plus 13.5 basis points, and 6.9 basis points below its fifty-two-week high of 5.979% - also seen this week. Nearly 6%, for 30 years, year on year - for doing nothing. Are we in the right game, we ask ourselves? Depends - because this drift could go further.
The intra-week story is worse than the closes. 30-year at a 28-year high. The 10-year got to 5.38%, the highest since 2007, which is to say the highest since before the financial crisis. Sixty-nine of the seventy conventional gilts in issue rose on the week. That is not a rotation within the curve. That is the whole curve.
How about the swaps? If you haven't been watching, I bet you can't wait...... Here is the number I would put in front of anybody who still thinks the gilt market is somebody else's problem.
The five-year SONIA swap, which is the rate lenders price five-year fixed mortgages off, closed Thursday at 4.67%. Last week: 4.36. A month ago: 4.22. A year ago: 3.82. I wondered why I hadn't quite slept so well this week.......
That is plus 31 basis points in a week and plus 85 basis points in a year. And note what it did relative to the gilt: the swap moved 31 basis points while the five-year gilt moved 27.2. The swap outran the underlying. When that happens it is usually the market pricing more uncertainty about the path rather than simply a higher level, and it is the swap that prices your debt, not the gilt that gets the headlines. We are still relying on the 25-30 basis point discount on the swap that we've been seeing for years now - and indeed we need it to get bigger, not smaller!
So what for the mortgage rates? That's always the biggie after a week like this. I said a sensible limited-company five-year fix was pricing around 6.25% with stress tests running near 7.5%, the highest since Truss. My best guess on where five-year limited-company fixed debt sits in six months' time is 6.5%, after this week. That is a guess, it is mine, I have written it down so you can hold me to it, and I would put it somewhere around a coin toss with a lean. It’s rare that I revise it on the back of one week’s activity but at this stage, and at the higher highs we are seeing - it is prudence. Any vendors coming up to a remortgage will have a rude and painful surprise next week - and opportunity knocks on the back of that.
The Bank of England's own quoted rate data for August already shows the two-year fix at 75 per cent loan to value moving to 4.92 per cent from 4.79, and the revert-to-rate at 6.58. Market expectations for Bank Rate in a year's time were sitting at 4.63 per cent as of 9 September. Nobody is pricing a rescue - nor should they.
So, on stress testing. Some of you reading this will already be modelling at 7.5 per cent and will not need telling. For anybody who is not: it is worth doing this week rather than in November, and it is worth doing on the whole book rather than on the deal in front of you. Not because I think 7.5 is where we land. I do not. But the gap between where you fix and where you can survive is the only number that has ever really mattered in a leveraged business, and this is the first autumn in three years where I have watched that gap close rather than open. Get it done.
The optimism inside the realism, and there is some: the activity data is improving, the north is still putting on price, Scotland is selling in 47 days, and the reason yields are where they are is mostly a strait in the Persian Gulf rather than anything structural about Britain. Straits reopen. Gilt curves are not destiny. Affordability for buying houses in income terms is improving, although for leveraged buyers it is deteriorating faster than it improves, sadly.
Black armbands away, folks - Keep Calm and Carry On.......
OK. Having ended the lesson on current rates - here comes the Deep Dive. Four this week, and the thread running through all of them is the one Gladstone put his finger on at the top: the difference between what a policy costs on paper and what it does in the world. We have a Chancellor's first growth speech, delivered on Monday and answered by the gilt market by Friday. We have the landlord database, which after two years of "soon" now has a launch date, a fee (thanks to my “friend” who sent me a meme of a calculator displaying 700 x £65, at least we can try and laugh it off eh?) and a regional sequence, with the West Midlands going first. We have the Minimum Income Standard, the closest thing this country has to a public definition of "enough", which has quietly given up on social housing. And we have the quarterly lending statistics from the FCA and the Bank, which on first reading describe a completely different country from the one in the Macro section above, and on second reading do not.
First up, the Chancellor. John Healey's first major speech in the job. Worth taking seriously for two reasons: it is the last full statement of intent we will get before the Budget on 28th October, and it was made three days before the bond market went to work on the long end. There were no tax or spending measures in it, those being held back for the red box. What there was, was a framing, and framing is what the next seven weeks will be about.
Theme 1: The Credibility Pledge
The Summary: In his first major speech as Chancellor, delivered at a manufacturing site in Coventry on Monday 7th September, John Healey named growth as his top priority and committed the Government to fiscal restraint. He framed that commitment explicitly as continuity, saying he would build on his predecessor's work to re-establish the United Kingdom's credibility in international bond markets. The speech was billed as a growth speech rather than a fiscal statement and contained no tax or spending measures, those being reserved for the Budget on 28th October. The credibility framing was the organising idea of the address: the growth measures were presented as the route to fiscal sustainability rather than as an alternative to it. The speech was delivered on the Monday of the first full week of parliament's return, and preceded by three days the ONS growth release covered in the Macro section above, which showed July growth of 0.4% against a consensus of zero.
The Propenomix Perspective: There is a version of this speech that works, and it is not far from the one he gave. Growth as the route to the numbers adding up is the correct answer, and it is a great deal more honest than pretending a spending review can be balanced with a scalpel. But credibility is not a thing you announce. It is a thing other people decide about you, and the people who decide it publish their verdict every day at about half past four in a yield. The verdict that week was a twenty-eight-year high on the thirty-year, and I am not going to pin that on him, as I said in Trumpwatch, because the oil did most of it and the five thousand dollar promise did a fair bit of the rest. But there is an uncomfortable asymmetry in this job that every Chancellor learns the hard way. You can only lose credibility with the bond market. You cannot win it. The best outcome available to him on 28th October is that nobody notices. What would I want from the speech that I did not get? A number. Any number. A debt-to-GDP path, a headroom figure, a date by which the current budget balances. "Building on my predecessor's work" is a sentence that reassures nobody who lends money for a living, because the predecessor's work is precisely what they were pricing at 5.9%. My hedge: it was a growth speech, he said it was a growth speech, and I am marking it against a fiscal test it did not sit. Fair enough. The Budget sits that test, and the market has already written the first question.
Theme 2: Growth in Every Postcode, and the £150 Million
The Summary: The speech set out a plan to drive growth "in every postcode", backing what the Chancellor described as Britain's extraordinary strengths to deliver benefits across the whole country. The headline regional measure was a new £150 million scale-up fund across the North of England, aimed at companies past the start-up stage that struggle to find growth capital domestically. The regional framing sat alongside the national ambition, and the Coventry venue was chosen to make the point geographically. Further detail on the fund's structure, its administering body, its investment criteria and its timetable was not set out in the speech itself, and no accompanying document was published giving it. The measure sits within a broader growth package that includes the unicorn target and the procedural reforms covered in the two themes below.
The Propenomix Perspective: I want to be fair here, because £150 million is a real number and the scale-up gap in the North is a real problem. Companies that get to £2 million of revenue and then cannot find the next £5 million are the single most under-discussed drag on regional productivity in this country, and somebody noticing is better than nobody noticing. But let us keep the arithmetic honest. £150 million is, on rough working, about one seven-thousandth of the sum the American president promised in a single post this week, and set against the phrase "growth in every postcode" - I hate this choice of words because it fails before it starts, even if nearly everyone is doing really well (which is unlikely) - it is a rounding error with a press release attached. And for those of us who buy houses, the thing to watch is not the fund. It is whether anything in this agenda changes the wage line in the North, because that, and not the fund, is what moves the value of a two-bed terrace in Oldham. Wages move house prices; funds move headlines. I am not convinced yet. I would like to be. Where does this leave us? Watching the Budget for the bit that has a number and a mechanism attached, and treating the rest as intent.
Theme 3: Doubling the Unicorns
The Summary: The Government used the speech to set out an ambition to double the number of UK unicorn firms, meaning privately held companies valued at over one billion dollars, describing it as a clear signal of ambition for the British economy. The measure sits alongside the scale-up fund as part of a package aimed at the capital-formation end of the growth problem rather than the demand end. No target date for the doubling was attached in the material published alongside the speech, and the baseline figure from which the doubling would be measured was not stated.
The Propenomix Perspective: A target with no date and no baseline is not a target. It is a mood. I do not want to be cynical for sport, and ambition is not a vice in a Chancellor, but this is the kind of line that gets written by somebody who needs a number in the speech and has not been given one, and everybody in the room knows it. More substantively: the unicorn count is a strange thing to aim at. A unicorn is a valuation event, not an output. You can double the number of billion-dollar private companies in Britain without adding a single job or a single unit of productivity, simply by having a good year for late-stage funding rounds. We have spent the last decade discovering that the British problem is not creating companies worth a billion dollars. It is keeping them here once they are, and keeping the listing, the head office and the payroll here with them. Will they be building houses, providing care, and paying wages in Sunderland? I am not convinced of that at all. I suspect the room in Mayfair gets more of the attention than the firm in Bolton, because it makes the better headline. Too cynical? Perhaps. Check back in a year.
Theme 4: The Axe to Consultation, Litigation and Administration
The Summary: The Government said it is "taking an axe to the thicket of unnecessary consultation, litigation and administration holding back growth". The framing targets the procedural layer that sits between a decision to build and the building itself, encompassing statutory consultee requirements, judicial review exposure and administrative process. The measure follows the National Planning Policy Framework rewrite covered in this newsletter on 30th August, which proposed a clear rules-based planning system with a default yes near stations, and the consultation response on statutory consultees published alongside it. Specific legislative vehicles and timings for the consultation and litigation reforms were not detailed in the speech. Separately this week, the Home Builders Federation published an analysis finding that Section 38 highways adoption agreements take just over four years on average from submission to adoption, with some taking up to twelve years.
The Propenomix Perspective: This is the bit of the speech that actually matters to this readership, and it got the least coverage, which is roughly how these things always go. If you want one number that shows why, take the HBF's finding this week that Section 38 agreements are taking just over four years on average, with some at twelve. Twelve years to adopt a road. Not to build it. To adopt it. That is what the thicket actually looks like on the ground, and no amount of scale-up funding touches it. Every month of that delay is carried cost on a site that is already built, and it lands in the price of the house at the end and in the service charge after that. So: good. I would rather have this than the fund, and I would rather have it than the unicorns by a distance. My hedge is the one I have earned. We have had "axe to red tape" speeches from roughly every government of my adult life, and the thicket has grown through all of them, because the thicket is made of individually reasonable things. Somebody has to be willing to be blamed when the reasonable thing gets removed and something goes wrong again, and that is a test of nerve rather than drafting. We will not know for two years whether he passes it. Taken with the 30th August NPPF and the consultee response, it is the most coherent statement of intent on process I have seen from the Treasury. Intent is not a road, though. The potential nonsense bit - this is simply rewording of Starmer and Reeves promising to cut regulation, and then not cutting it at all apart from in the financial services sector. So someone has looked at this and gone - well, they were right last time but they didn’t do it. So how can we reword it? Will they now actually DO it? I do get a feeling they are learning from what went wrong last time out, but we need to measure action not intention.
Second, the register. On Wednesday 9th September the Government published the Private Rented Sector Database Regulations 2026 and, with them, the thing this readership has been asking me about weekly since the Renters' Rights Bill was first published: when. The "Register your rental property" service opens on 15th December 2026, the duty lands region by region over the following eight months, and it costs £65 per property per year. I have been asked more or less every week for two years when the database was actually going to arrive, and my answer has always been a shrug and a hedge, because the timetable kept moving. It has stopped moving.
Theme 1: The Dates, and Who Goes First
The Summary: The Private Rented Sector Database Regulations 2026, published on 9th September, set out when landlords must register and what information they must supply. The "Register your rental property" service opens on 15th December 2026 for landlords anywhere in England, but the duty to register applies regionally, with a three-month window from the date each region is called forward. The West Midlands goes first, with regulations commencing on 15th December 2026 and a registration deadline of 14th March 2027. The East of England follows on 15th January 2027, the East Midlands on 15th February, the South East on 15th March, Yorkshire and the Humber on 15th April, the North West on 15th May, the North East on 15th June, London on 15th July and the South West on 15th August 2027, with deadlines three months after each. By 14th November 2027 every in-scope rental property in England must be registered. Landlords with properties in several regions may register all of them from 15th December rather than waiting for each deadline. Only properties currently let, or let during the roll-out, need registering at this stage; unoccupied properties will need registering before marketing once the public-facing side of the service launches.
The Propenomix Perspective: Put 15th December 2026 in your diary now, and if you own anything in the West Midlands put 14th March 2027 next to it in red. The sequence starts with the region a great many of the people reading this actually buy in. The regional stagger is, I think, sensible administration and terrible communications. Sensible, because a national switch-on would have collapsed under the load on day one. Terrible, because the practical effect is that most landlords will hear "December 2026", check that they are not in the West Midlands, and forget about it until a letter arrives, and the letter will not arrive, because nobody has your address until you register. Three months is not long when the thing you need is documentation you have not looked at since you bought the place. One practical note. The early-registration option is worth more than it looks: if you have stock in four regions, register the lot on 15th December and have one renewal date rather than four. Is the timetable going to slip? Perhaps at the back end. I would be surprised if the West Midlands date moves now, because moving the first date is the one thing that would tell everybody the rest are movable too.
Theme 2: The £65
The Summary: Registration carries an annual fee of £65 per property, renewable each year. The fee is per property rather than per landlord, so the cost scales linearly with portfolio size: £65 for a single-property landlord, £650 a year on ten properties, £3,250 a year on fifty, in each case before administrative time and separate from any selective, additional or HMO licensing fees that already apply. The fee will be pro-rated during the roll-out so that landlords registering early are not charged more than those registering later, after which charges reset annually. Registration requires a GOV.UK One Login. Landlords must create a landlord entry and a separate dwelling entry for each property, and the information required includes the landlord's date of birth, address, phone number and email, the rent charged, the number of occupants, and copies of the gas safety, electrical safety and energy performance certificates. The fee is set to fund the operation of the database and local authority enforcement activity associated with it.
The Propenomix Perspective: £65 a property a year is not, on its own, a portfolio-breaking number. On a property renting at £900 a month it is about 0.6% of gross rent, and I have seen people get more worked up about less. But I would make two observations, and the second is the one that matters. First, it is annual and it is per unit, which means it is not a compliance cost so much as a tax on holding rental property, collected by a database, and it will be uprated. Nothing set at £65 in 2026 is still £65 in 2030. Second, and this is the bit I keep coming back to: it lands on exactly the people who least need another fixed cost. HMRC's own landlord statistics, which we covered a fortnight ago, show 2.88 million unincorporated landlords with nearly half of them declaring under £10,000 of property income. For a landlord with one flat and £6,500 of income, £65 is a per cent of the lot, stacked on top of everything else that has been added since 2016. That is not the rogue landlord this register is aimed at. The same year an extra 10% goes on the tax (from 20% to 22% at basic level) if you have the temerity to make a profit. This hits the accidental landlord, and the accidental ones have been leaving for four years. The document list is the other thing to look at. Gas, electrical and EPC uploaded, rent stated, occupants counted: that is the first time a central body has held all of that for every let in England in one place, and I would think about what a future Chancellor could do with a dataset that tells him the rent on every private tenancy in the country. Nothing this Budget, I suspect. Something, eventually.
Theme 3: The Possession Sanction
The Summary: Under the Renters' Rights Act, a landlord in breach of the duty to register will not be able to obtain a possession order except where the ground relied upon is Ground 7A or Ground 14. Ground 7A concerns serious anti-social behaviour resulting in a conviction, breach of an injunction or closure order, and Ground 14 concerns nuisance and anti-social behaviour more broadly. In practical terms an unregistered landlord loses access to the standard grounds for possession, including those relating to rent arrears, the landlord's intention to sell, and the landlord's or a family member's intention to move in, while retaining access only to the anti-social behaviour route. Separately, landlords who fail to register may face civil penalties from the local authority, and continued or repeated failure may attract higher penalties. The sanction operates automatically through the court process rather than requiring enforcement action to be initiated by a local authority.
The Propenomix Perspective: This is the part of the announcement that should get everybody's attention, and it is buried under the fee in every piece of coverage I have read, which tells you something about how we all read policy. The fine is the headline. The possession sanction is the actual enforcement, and it is considerably sharper than a fine, because it is not discretionary and it does not require a council to do anything. It just sits there, and the day you need it you discover you have not got it. Think about the sequence. A tenant stops paying in month four. You go to serve notice. Your registration lapsed because the renewal email went to an address you stopped using in 2024. You now cannot get a possession order on arrears. You can get one for anti-social behaviour, which is no use to you, because your problem is not noise, it is money. £65 a year suddenly looks like the cheapest insurance policy in the portfolio. I would set a calendar reminder rather than rely on a direct debit, because direct debits fail quietly and calendars do not. Is this proportionate? For the landlord who forgot, no, and there will be a hard case in the press within a year of the first deadline. For the landlord who never intended to register, it is the only sanction that has ever worked in this sector, because it costs the state nothing to apply. In theory. However - as I said on an intelligent LinkedIn post on this subject (not on my LinkedIn, to be clear!) - the rogues won’t be stopped, because the rogues don’t use the court system for evictions - they have other methods. I have said for years that regulation here punishes the disorganised more than the dishonest. This is the clearest example yet.
Theme 4: What the Database Is For
The Summary: The stated purpose of the database is twofold. It gives local authorities access to information on the private rented sector to help them identify non-compliant landlords and take quicker enforcement action, addressing the long-standing problem that councils have had no reliable record of who owns and lets what in their area. In future it will also enable tenants to see whether a landlord, or a prospective landlord, has registered and is compliant with key legal requirements, though the Government has said the information available to tenants will balance landlords' privacy concerns against the need for tenants to make better informed rental choices, and has not yet published the list of what the public will be able to see. The tenant-facing side of the service is a later stage; the December launch is the landlord and local authority side. The database therefore functions first as an enforcement tool and later as a consumer-facing check at the point of choosing a property.
The Propenomix Perspective: The tenant-facing half is the one nobody is pricing, and I think it is the more consequential of the two. An enforcement database changes the odds of getting caught. A public register changes the market. Once a prospective tenant can check a landlord in ten seconds on a phone while standing in the hallway, being registered stops being a compliance question and starts being a marketing one. The professional operator gets a visible signal they did not have before. The amateur who has not got round to it discovers their void periods have got longer, and will not necessarily understand why. I have based my strategy for years on the view that regulation in this sector consistently rewards the professional and punishes the casual, and that most people underestimate how quickly that gap compounds. I would file this as more evidence for that, and I would say, gently, that if you are running properties in a way you would not want a tenant to look up, the window for sorting it out has a date on it now. The privacy question is the one I would watch. The Government has said it will "balance" what tenants can see, which is Whitehall for "we have not decided", and there is a real argument on both sides: a tenant is entitled to know their landlord is compliant, and a landlord with one flat is entitled not to have their home address on a public website. My guess, hedged, is that tenants get a compliance flag and a registration number and not much else, at least to begin with. Which would be about right.
Third, the benchmark. On Tuesday 8th September the Joseph Rowntree Foundation published A Minimum Income Standard for the United Kingdom in 2026, produced as ever by the Centre for Research in Social Policy at Loughborough. The Minimum Income Standard is the eighteen-year-old exercise in asking ordinary members of the public, in rooms, what a household actually needs for an acceptable standard of living, and then pricing it. It is not an advocacy number and it is not a poverty line, and that is precisely why this year's edition matters: for the first time since the series began in 2008, the public has concluded that social housing is no longer a realistic option for anybody, and every budget in the report now assumes a private rent.
Theme 1: Social Housing Is No Longer a Realistic Minimum
The Summary: The 2026 Minimum Income Standard contains a significant methodological change. Since 2008, the deliberative groups that construct the MIS budgets had agreed that social rented accommodation would meet the minimum housing needs of pensioners and households with children; working-age households without children have been assumed to rent privately since 2014. In the 2026 research, drawn from thirteen groups meeting between October 2025 and February 2026 in Nottingham, Bristol, Leeds, Sheffield, Edinburgh, Cardiff, Norwich, Leicester and Manchester, all household types are now placed in the private rented sector because participants felt social housing was no longer a realistic minimum. The report is explicit that groups still regarded social housing as adequate; the change reflects the view that a household could not access it within a reasonable time frame. It cites expected waits for a three-bedroom social home in England outside London ranging from 0.7 years in Tonbridge and Malling to 75.5 years in Mansfield, and Shelter's estimate that clearing current waiting lists at the present build rate would take 119 years. The rent assumption is a lower-quartile private rent for the relevant property size in the East Midlands.
The Propenomix Perspective: Stop and read that again, because it is one of the most quietly devastating sentences published in British housing this year and it will get a fraction of the coverage of a house price index. The Minimum Income Standard is built by asking people, not by modelling them. For eighteen years that conversation has assumed that a low-income family or a pensioner might reasonably get a social tenancy, and this year the people in the rooms concluded the assumption is no longer credible enough to build a budget on. That is not a housing statistic. It is a statement about what the state has stopped being able to offer, arrived at reluctantly, by researchers whose whole method is to reflect the public back at itself. I would file it alongside Shelter's 119 years to clear the queue, which we covered in June and which the report itself cites. Different researchers, different method, same conclusion reached from the other end. The Mansfield figure is the one that will stick with me. Seventy-five years for a three-bed, in a town where a three-bed terrace costs a good deal less than the national average. That is not a shortage of houses. It is a shortage of a particular tenure, in a place with no shortage of the bricks. Is there a reading of this that is good news for the private landlord? On the numbers, yes: the country has just formally conceded that we are the minimum. On any other reading, no, and I will come to why in theme three. I also want to comment on the “lower-quartile private rent”. When I looked at what comes to market, I concluded in several areas a couple of years back that that sits at about the 85th percentile of the VOA’s rents. Why? Surely it should be the middle, shouldn’t it? No - there’s no reason why it should when you think it through. However, 2 years ago was near the low for decades in terms of stock. This is a question that deserves a fuller answer, and will inspire some future work on my PropenomAIx newsletter and YouTube channel (check them out if you don’t already subscribe!). The conclusion I feel safe drawing before doing that piece of work though - the bottom quartile does not mean one quarter of the available housing stock on the rental market at any one time. It means nowhere near that much.
Theme 2: What It Does to the Arithmetic
The Summary: Because private rents are substantially higher than social rents, moving the housing assumption raises the income required to reach MIS for every household type that was previously assumed to be in social housing. The report calculates that a single pensioner now needs a gross income of £21,300 a year to reach MIS in the private rented sector against £15,900 in social housing, a difference of £5,400. A lone parent with two children in full-time work needs to earn £67,600, up from £60,800 in 2025, and £7,800 more than the £59,800 they would need in social housing. A couple with two children need £77,400 between them, £3,800 more than the £73,600 social-housing equivalent. The rent element of the weekly budget rises by 50.1% for households with children and 34.3% for pensioner couples as a result of the tenure change alone. A single working-age adult, already assumed to rent privately, needs £31,500, up from £30,500, and full-time work on the National Living Wage reaches 73% of that. The report notes that mean private rents were almost twice social rents in 2024-25, at £250 against £129 a week, and that because Local Housing Allowance has been frozen since 2024 it now assumes only 88% of rent is eligible for Universal Credit support, down from 95%.
The Propenomix Perspective: That is what "unaffordable" means when you price it rather than assert it. £5,000 a year for a single pensioner. £7,800 for a lone parent. And the mechanism is the one that matters for our side of the table, because the mechanism that has pushed the benchmark into the private rented sector is the same one that has been supporting private rents for a decade: there is not enough social housing, so the private rented sector absorbs the demand, so private rents rise, so the income needed to reach a minimum standard rises. We are on the receiving end of that as a business, and I am not going to pretend otherwise or perform embarrassment about it. But I would rather be clear-eyed about why the rent roll is what it is than tell myself a story about market fundamentals. The fundamental here is a policy failure, sustained across governments of both colours for thirty years, and we are the ones housing the consequences. A hedge on the report's own numbers, because it invites them: the rent used is a lower-quartile East Midlands rent, which is a sensible choice for a national benchmark and a useless one for anybody in the South, where the same exercise would produce a gap two or three times the size. The London version of this research, published in March, had already reached the same conclusion for the capital, and there the private-rent assumption takes nearly half the weekly budget for some pensioner couples. And the 88% LHA assumption is the sleeper: the report has quietly built the frozen allowance into what a low-income household can expect, which is the researchers being honest about the world as it is, and it makes the Budget point in theme four rather sharper.
Theme 3: The Landlord Instruction Problem
The Summary: The MIS finding sits alongside this week's RICS survey, covered in the Macro section, which recorded landlord instructions at a net balance of minus 14 and rent expectations at plus 44, and alongside HMRC's landlord statistics published on 30th August, which showed 2.88 million unincorporated landlords with partnership income the fastest-growing category and nearly half of individuals declaring under £10,000. Taken together the sources describe a sector being asked to absorb a growing share of housing need while the flow of new rental supply into it continues to contract. The report also records that private renters spent an average of 34% of their income on rent in 2024-25 against 28% for social renters, that private rents have risen in real terms since 2022 for the first time since the financial crisis, and that the Renters' Rights Act, in force since 1st May 2026, may ameliorate the insecurity participants raised while leaving the cost problem unaddressed. Separately, the report references an IPPR estimate that 2.4 million households face unaffordable housing costs.
The Propenomix Perspective: So the squeeze, stated as plainly as I can manage. The country has just formally conceded that the private rented sector is where the minimum standard of living now lives. In the same week, the surveyors tell us landlord instructions are negative and rent expectations are at plus 44. You cannot hold both of those and also hold the position that rents are a problem to be solved by making it harder to be a landlord. The arithmetic does not permit it. If the sector is the safety net, the sector needs supply, and supply responds to returns, and returns have been going one way since 2016, with a two-point surcharge on the way in April 2027 and a £65 register fee before that. I have said many times that I am not asking for sympathy for landlords, because nobody is going to give it and we shouldn’t waste our time courting it either. I am asking for the policy to be internally consistent, and at the moment it is not. The people who pay for the inconsistency are tenants, in the form of a rent expectations balance at plus 44, which is what a market says when the buyers are arriving and the sellers are leaving. Can we blame the researchers for the conclusion? Not remotely. They have described the world. The inconsistency is upstream of both, in a housing policy that has spent a decade treating the private landlord as the problem and the private rented sector as the solution, and I do not think you can run both of those for much longer without the second one starting to fail. The RICS instructions number has been negative for so long that it has stopped being news, and it is the one line in all of this I would actually watch. Put it another way - the sector has been under sustained attack for the last 11 years, but now is the minimum standard for those with lower incomes. Tension much? Of course. Well done, successive Governments……
Theme 4: What This Means for 28th October
The Summary: The Minimum Income Standard is used as a reference point in policy debate on benefit adequacy, the National Living Wage and Local Housing Allowance. The 2026 edition's shift to a private-rent assumption arrives seven weeks before the Budget on 28th October, in the same period as the National Residential Landlords Association's analysis of Local Housing Allowance covered here on 23rd August, which found no evidence of a link between benefit levels and rent inflation. LHA rates, frozen since April 2024 at the 30th percentile of rents as they then stood, remain a live Budget decision, as does the benefit cap, which was last uprated in 2023. The report finds that out-of-work benefits after housing costs now provide 24% of MIS for working-age households without children, 36% for lone parents and 28% for couple parents, the latter two down from 43% and 36% in 2025 as a direct result of the tenure change. It concludes that the Prime Minister's stated intention to "pull every single lever" on the cost of living has so far produced measures unlikely to close the gap between incomes and costs.
The Propenomix Perspective: I flagged three weeks ago that the Local Housing Allowance decision was a Budget-day watch item, and I promised a proper piece proving the freeze net-costs the state. I have still not written it, and this report has just handed me half the argument. The shape of it is this. If the benchmark for an acceptable standard of living now assumes a private rented home, and the Local Housing Allowance is frozen below private rents, then the state has adopted a standard it is simultaneously refusing to fund. Everything downstream of that, the temporary accommodation bill, the homelessness applications, the children in hotels, is the cost of the gap. Whether the Treasury will look at it that way in the current fiscal weather, I very much doubt. A freeze is a saving this year and a cost in three, and Chancellors seven weeks from a Budget with the thirty-year at 5.9% are not famous for their three-year horizons. My honest read is that LHA gets an uprating in the Budget, because the Prime Minister has said in terms that he agrees with the NRLA's logic, and that it is a smaller uprating than the arithmetic wants, because the Chancellor's credibility pledge has to be paid for somewhere and this is the sort of line that pays for it. Where does this leave us? Watching the Budget and expecting to be half-satisfied, which is an improvement on where I have been on this since 2023, and I will take it. The full piece is still owed. I have said that before too. Save money today, cost 3 to 10 times as much money tomorrow. “Genius” - in fact, completely the reverse of the way the state coffers should be operated. But nonetheless it is the way they have been operated ever since around the second round of cuts in the austerity drive - and no-one has had the guts, or the common sense, to do anything about it.
Last up, the lending statistics. The FCA and the Bank of England published the Mortgage Lenders and Administrators Return for the second quarter of 2026 this week, which is the quarterly count of what was lent, what was agreed, what is in arrears and what was repossessed, and it reads on first glance like it was compiled in a different country from the one in the Macro section. Gross advances up 31.7% on the year. Arrears at their lowest since 2023. Possessions falling. And then you look at the pipeline figure, and the quarter it describes, and the different country turns out to be the same one three months ago.
Theme 1: The Lending Boom in the Rear-View Mirror
The Summary: The quarterly mortgage lending statistics for the second quarter of 2026 show the outstanding value of all residential mortgage loans at £1,760.6 billion, up 0.8% on the quarter and 3.1% on a year earlier. The value of gross mortgage advances rose 11.1% on the quarter to £77.4 billion and was 31.7% higher than a year earlier. The value of new mortgage commitments, which is lending agreed but not yet advanced, rose only 1.4% on the quarter to £79.2 billion and was 1.3% higher than a year earlier. The figures cover the period from April to June 2026, which preceded the rise in gilt yields and swap rates that began with the outbreak of the war in late February and accelerated through the summer.
The Propenomix Perspective: Those last two figures are the most important thing in this week's Deep Dive and almost nobody will put them next to each other, so let us do it here. Gross advances up 31.7% on the year. New commitments up 1.3% on the year. An advance is money going out of the door on a deal agreed weeks or months ago. A commitment is a deal agreed now that will be funded later. So the first number tells you what the market did last winter and spring, and the second tells you what it is doing now, and the thirty-point gap between them is the market changing direction with the change already made. Anybody writing "mortgage lending surges" off the 31.7 this week is reporting the weather from March. Read alongside Chris's flow data and the RICS activity balances, the story is consistent: the sales agreed in the spring, when the five-year swap was in the low 4s, are completing now, and the pipeline behind them is being agreed against a swap at 4.67 and a limited-company five-year fix at 6.25. I might be wrong about how quickly the flat pipeline shows up in the advances number, and I would not put a quarter on it. I do not think I am wrong that it will. The one thing I would add for balance: 1.3% up on the year is not a collapse. It is flat. A flat pipeline in a market where the price of money has just gone up 85 basis points in a year is, in its way, resilient, and I would rather have flat commitments than the alternative - but the flat spot has already been with us, whereas the 85 basis points are yet to bite (and the size of that bite mark will depend on how temporary, or not, the 85 basis points - or more - are).
Theme 2: Arrears, Which Are Behaving Beautifully
The Summary: The proportion of total mortgage loan balances in arrears stood at 1.1% at the end of the second quarter, unchanged on the quarter and 0.1 percentage points lower than a year earlier. Of the £19.7 billion of outstanding balances in arrears, the value of non-regulated mortgages in arrears, a category that includes buy-to-let loans and other residential lending where the property is not occupied by the borrower, fell 3.4% on the quarter to £4.2 billion, the lowest level since the second quarter of 2023 and 12.7% lower than a year earlier. The statistics count arrears by balance rather than by number of accounts, and treat a loan as in arrears once the shortfall exceeds 1.5% of the outstanding balance, so the figures understate the number of borrowers who are behind by small amounts.
The Propenomix Perspective: Buy-to-let arrears at their lowest since the second quarter of 2023, in a week when the five-year swap moved 31 basis points. If you wanted a single illustration of how long the fuse is on a rate shock in this asset class, that is it. The explanation is not mysterious and it is worth stating rather than assuming. Most leveraged landlords are on fixed rates. Most of them were fixed either before the shock or in the window when pricing improved. The arrears number therefore reflects the cost of debt taken out somewhere between 2021 and 2024, not the cost of debt available on Friday afternoon. Which means this number is at its most reassuring precisely when it is least informative. It is a rear-view mirror with a very long lens. I want to be careful not to overstate it in the other direction, because I made the opposite case in 2023 and I was right then: falling arrears are good news, the sector is more resilient than the commentary assumed. But I would not take £4.2 billion and a downward trend as evidence that a 6.5% five-year fix is survivable. The people it will not be survivable for have not refinanced yet, and the first of them refinance in the spring. We could do with some data on - for example - just how many took a chunky arrangement fee against a property that might not have gone up much in value - or, indeed, as per the Watkin section - has gone DOWN in value in the interim? My guess, and it is a guess, is that this series bottoms in the next two quarters and turns in 2027, gently, and that gently is the word, because the alternative to a fixed rate rolling off is the Mortgage Charter in theme four. One more thing to remember - the debt service cover rules don’t apply to product transfers - they can be transferred at 100% debt cover, or even worse - but that leaves people holding “assets” bleeding money every month - fine if they are going up in value (think 2005 as an example) but not fine if it is a flat with a lease approaching its 80-year birthday.
Theme 3: Possessions, at a Level That Keeps Surprising People
The Summary: The number of new possessions in the second quarter of 2026 fell 7.1% on the quarter to 2,058 and was 15.6% lower than a year earlier. This continues a multi-quarter downward trend and leaves possessions at a level far below historic norms for a period of elevated interest rates. The figure covers possessions across regulated and non-regulated lending. Lender forbearance arrangements, including those under the Mortgage Charter, remain a material influence on the number, as does the strength of the underlying sales market as an alternative disposal route for borrowers in difficulty, who can in most cases sell at a price above the outstanding balance rather than be repossessed. The FCA's conduct rules requiring lenders to treat possession as a last resort also remain in force.
The Propenomix Perspective: Two thousand and fifty-eight possessions in a quarter, in a country with millions of outstanding mortgages, with Bank Rate where it is and the five-year swap where it is. For anybody waiting for the distressed stock to arrive: it is not arriving. I have been saying this for two years and I am going to keep saying it, because the "wait for the crash and buy the repossessions" strategy is still being sold to people at the back of rooms, and it has now been wrong for three consecutive years. The reason is forbearance plus a functioning sales market. A lender's choice is not between possession and getting paid. It is between possession, which is slow, expensive and publicly awful, and a term extension, which costs them nothing today. They will take the extension every time, and the borrower will usually take it too, and the property that would have been your bargain instead becomes a thirty-five-year mortgage on somebody in their late forties. Add to that a market where, for all the friction Chris counts every week, a house priced properly still sells, and the borrower in trouble has an exit that is not a court. Where does this leave us? Exactly where we were, doing the work. Deals are made, not found, and they are made by talking to motivated sellers rather than waiting for a court list. The 1990s comparison everybody reaches for happened with negative equity across a large part of the South, and nothing in this week's Lloyds regionals, even the minus 1.6 in the South East, gets you anywhere near that.
Theme 4: The Mortgage Charter, and What 22,400 Households Tell Us
The Summary: Data published alongside the lending statistics shows that in the second quarter of 2026, 22,400 mortgages were switched to interest-only arrangements or had their terms extended under the Mortgage Charter. The Charter, agreed between the major lenders and the Treasury in June 2023, allows borrowers who are up to date with payments to make these changes for a period of six months without an affordability check and without impact on their credit file, with the option to reverse the change within that period. The figure represents the flow of such switches within the quarter rather than the cumulative stock of borrowers who have used the arrangements since the Charter began.
The Propenomix Perspective: Twenty-two thousand four hundred households, in three months, quietly changing the shape of their mortgage so they can keep paying it. And it does not show up in arrears, because that is precisely what it is designed to prevent. This is the number that makes theme two make sense. Arrears are at their lowest since 2023 partly because the sector is resilient and partly because we have built a very effective machine for converting would-be arrears into longer debt. I am not criticising it. As crisis responses go it has been one of the better ones, and it has kept a lot of people in their homes who would otherwise not be. But it is worth being honest about what it does rather than what it looks like. It moves the problem from this year's arrears statistic into the mid-2030s, as capital that should have been repaid and was not. Somebody is going to reach sixty with an interest-only balance they extended in 2026 and a plan that consists of hoping. For our purposes, the practical read is this: that is a supply pipeline. Not this year, not next, but a steady flow of owners in their late fifties and sixties, with a balance and no repayment vehicle, who will need to sell. It is one of the reasons I keep saying the 2030s look more interesting to me than 2027 does. Is 22,400 a big number? Against the whole mortgage book, no. Against 2,058 possessions, it is eleven times the size, and that ratio, forbearance to possession, is the real story of this cycle. Chris's flow data will tell you when it changes. I would give it a couple of years.
The Week Ahead. Clear your Thursday, because this is the week the Bank has to say something instead of speculating on the interest rate. Tuesday 15th brings the labour market report, and I would watch the payroll number rather than the unemployment rate, and watch whether the long run of services job cuts the surveyors keep reporting has finally turned up in the official data: it has been twenty-three months of it in the purchasing managers' figures now, the longest run since 1996, and at some point the two series have to meet. Wednesday 16th is the heavy one. Consumer price inflation, where the question is how much of Thursday's $108 Brent is already in the number and how much is still coming; private rents and house prices from the ONS on the same morning, which will give us the official read against the surveyors' plus 44 on rent expectations; and building materials, which after this week's construction detail is worth more attention than it usually gets. The Treasury's monthly collation of independent forecasts also lands on the 16th, the day before the Bank moves, which makes it the most useful thing published all week and the least read. Thursday 17th is the Monetary Policy Committee, and also median house prices and sales volumes from the ONS, which nobody will notice because of the Bank. Friday 18th, the annual mortgage lending statistics.
Off-calendar, three things. Oil, obviously, and specifically whether the strikes continue into a third week. The Budget countdown, now at six weeks, with every Sunday paper a market-moving event until the red box. And the Ofgem cap for January, announced on 25th November, which is a long way off but feeds directly into where inflation peaks. My call on Thursday, and I am hedging it harder than usual: still a hold, but the margin has narrowed ever more since last week - I’m at 5-4 hold (55%) with the second option being 5-4 hike (35%), against the grain - with a saver on 6-3 hold which I’d have at about 10% likely.
As we get towards the end for this week - I can't wait for our next workshop at the Hyatt Regency in Manchester on 1st October, with Rod Turner, powered by Roma Finance. JVs and M&A is the topic, and given everything in the Macro section above, the conversation in that room about the cost of money, who has it, and what it costs to share it, is going to be a considerably more useful one than it would have been a month ago. The opportunities it is creating in deals where someone is looking at refinancing are huge. A joint venture partner with cash is worth rather more at a 4.67 swap than at a 4.22 one, and the arithmetic of buying somebody else's portfolio changes entirely when the vendor's refinance lands at 6.5. Early Bird is still on and the VIP dinner has one or two seats left at the time of writing. Book in at www.tinyurl.com/pbwoct26
Above all - please remember to Keep Calm, ALWAYS listen to or read the Supplement, and Carry On. It has been an ugly week for anybody who borrows money, and I am not going to dress that up: the thirty-year at levels it has not seen since 1998, the five-year swap up 85 basis points on the year, and a Chancellor who has discovered the bond market rather faster than he would have liked. But hold the shape of the thing in your head rather than the week. The reason yields are here is mostly a strait in the Persian Gulf and a promise made on the other side of the Atlantic, and neither of those is a statement about the value of a three-bed semi in Dudley. The activity data is improving. Landlord instructions are still negative, which means supply is still leaving a sector the country has just formally admitted it depends on. The fundamentals have not changed: there remains a shortage of two and three bed terraces and semis, there was one fifteen years ago, and nothing in this week's construction figures suggests anybody is about to fix it. Investing in property is not a guaranteed win, and I say that every week because it is true and because the weeks when it is easiest to forget are weeks like this one. The people who do well out of the next three years will be the ones who made their arithmetic honest in September rather than in January, and who were still standing when the refinancing wave broke. Those ready to buy - those who were willing to be patient for the sake of building a strong balance sheet. Most of you reading this fit into that category, I know. Enjoy the rest of your Sunday.
KCCO!