"The Government are very keen on amassing statistics. They collect them, add them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But you must never forget that every one of these figures comes in the first instance from the village watchman, who just puts down what he damn pleases." - Sir Josiah Stamp
The quote pertains to the deep dive, as ever - and to rather a lot of this week besides. The ONS published its Blue Book impact article on Thursday and moved the date Britain recovered from the 2008 crash. The labour force survey remains, by the ONS's own label, "official statistics in development". Two landlord surveys landed within days of each other pointing in different directions on tenant demand. And the housing benefit system has spent five years paying rents as they were, not as they are. Every one of these figures comes from somewhere, and this week we go looking for the village watchman.
As the summer winds down and the Budget countdown clock ticks louder - 66 days to the red box, for those keeping score - our next Property Business Workshop is live and tickets are selling. This is the really juicy one for taking big steps forward in your property businesses - subject matter JVs and M&A - joint ventures done properly, and buying and selling property businesses, portfolios and companies, not just houses. If you are already in one that isn’t working as well as it could - this workshop will help, and I know you are out there - I’ve seen two live examples just this week. Rod and I have been deep in the material for this one over recent weeks, and the timing could hardly be better: when margins compress and exits get harder, the deals move up a level, and the people who understand structure eat first. Book in on the next Property Business Workshop with myself and Rod Turner - Thursday 1st October - Manchester - https://tinyurl.com/pbwoct26 - there’s only ONE VIP dinner ticket left to get that extra time in with Rod and myself. I’m also delighted to break the news that this workshop will be powered by Roma Finance - my experience with Roma has been that they really do think differently about bridging and development lending, and have a refreshing take on the entire sector. They will also be represented at the workshop and Rod and I are delighted to be working with them. Watch this space!
Welcome back to Trumpwatch. If the last few weeks were the locked-loaded-cancelled cycle and the "pause", this week the pause hardened into something with a name: economic warfare. The President has stepped back from the negotiating table with Tehran, announced what he called the most crushing sanctions programme yet, and set about squeezing Iran's remaining oil revenue at source. Meanwhile, back home, a state house seat in deep-red Pennsylvania fell to the Democrats by 88 votes, and the commentariat spent the back half of the week arguing over whether Tuesday-into-Wednesday was the worst 24 hours of the second term so far. Let me take these in order, because the first one is the engine under half the numbers in the rest of this Supplement, as it has been since February.
The war, then. Where are we, honestly? The talks are paused - Trump's word, not mine - and the diplomatic track that briefly flickered over the summer has gone dark again. In its place: sanctions on Hezbollah, threats of "crushing" economic measures against Tehran, and a blockade that is visibly biting. Newswire reporting this week had Iranian oil offers to Chinese buyers falling away as the US squeeze took hold, which matters because China has been the buyer of last resort for sanctioned Iranian crude for the best part of a decade. The Strait of Hormuz remains the deadlock it has been for months - attempts to truly reopen it have gone nowhere - and the Iranian foreign minister, Abbas Araghchi, dismissed the sanctions threat as a "diversion" that "will only bring further defeat". The IRGC, for its part, warned that if the shooting restarts, the weapons used "will be completely different from the past in every respect". Encouraging stuff all round, I’m sure you won’t agree!
The military picture tells you something the rhetoric doesn't. The USS George Washington arrived in the region this week to relieve the USS Abraham Lincoln, and the reporting around that rotation was notably candid about the strain this war has put on the US Navy - carriers and crews are being run hard, the strain is no longer being hidden, and Washington's frustration is reportedly spilling over onto partners like Oman who have tried to keep the mediation channels open. An unpopular war that cannot force a surrender, cannot reopen the strait, and cannot easily be ended - I would suggest that is roughly where we are, although I would be delighted to be wrong. End in sight? Nope. Iranian chat says no way any solutions are coming until after the midterms - so here comes your lesson, Mr Trump (they hope).
And the oil price noticed. Brent pushed towards $94 on Thursday - the highest since late July - rising more than 2% on the day of the economic warfare announcement, and is back up there as this goes to print. Remember the shape of this year: $60 or so kicked us off, $120 very briefly at the peak, down to the low-$70s on the summer's peace hopes, and now grinding back up as those hopes deflate. The market keeps trying to price the end of this war, and the war keeps declining to end. For UK readers the transmission is the one we have traced all year: oil into the Ofgem cap, the cap into the CPI, the CPI into the Bank's nerves, the Bank's nerves into your mortgage pricing. Fuel prices visible at the pumps, but also the second order effects on pricing of your weekly shopping. More on every link in that chain below - this week's inflation print is essentially the March war arriving in your gas bill, five months later, right on schedule.
Now, the domestic story, which is the one Washington itself spent the week chewing on. On Tuesday, the 12th district of the Pennsylvania state house - Cranberry Township and surrounds, held by Republicans since 1999, carried by Trump by 18 points in 2024 - elected a Democrat, Brandon Dukes, by fewer than 100 votes. A TD Bank loan officer beat an Air Force veteran in Daryl Metcalfe's old seat. On its own, a state house special with turnout under a third means very little, and the Republican leadership duly said so. But it does not sit on its own: Democrats have now flipped around 30 legislative seats nationwide since January 2025, a Franklin & Marshall poll has just 29% of Pennsylvania voters rating the President's performance as good or excellent, and the Washington Post reported this week that multiple Republican Senate candidates are now running adverts critical of their own party. With the midterms in November, the pattern is the story, not the seat. Can we blame the White House for wanting to talk about anything else? Not really - hence, perhaps, the volume of the sanctions announcement two days later. I am not saying the two are connected (well, OK, I am). I am most definitely saying that in an election year, they never aren't.
Then there is the Fed, which delivered the week's quietest big story. The minutes of the July meeting landed on Wednesday and confirmed what we knew - a 9-3 hold at 3.50-3.75%, with Hammack, Kashkari and Logan all dissenting in favour of an immediate hike, the most lopsided set of one-way dissents since 2016. But buried a few paragraphs down was the genuinely new bit: Chair Warsh wants to cut the number of FOMC meetings from eight a year to six. Gold and silver both jumped roughly 4% on the combined news, and the US long end - already at highs not seen in 19 years - stayed heavy. That was all the chatter in podcastistan as far as rates went this week - including those who can foresee a 6% US 10-year yield in the relatively near future. Concerning stuff, folks, I’m not going to lie (but then when things are high, there are always those who will see higher highs). Fewer meetings, less forward guidance, a chair who has explicitly told markets that their pricing does not bind him. Whatever you think of the policy, the communication regime that ran from Bernanke through Powell is being dismantled in real time, and markets that grew up on it are having to relearn how to read a central bank that prefers not to be read.
Which brings us to Jackson Hole. The symposium runs Thursday to Saturday this coming week, and Warsh delivers his first keynote as Chair on Friday morning. The official theme is "Financial Innovation: Implications for Payments and Policy", and precisely nobody will be listening for the payments content. Markets price roughly a 39% chance of a September hike; the September meeting is three weeks after the speech; and the last several Jackson Hole keynotes have moved markets regardless of the printed agenda. One more delicious detail while we are here: Jerome Powell did not leave the building when he handed over the chair in May. He remains on the Board as a governor, term running to 2028, sitting at the same table, voting on the same decisions. Standard Fare, somehow, in 2026. It’s one of the very hardest for markets to attempt to price accurately, because September hikes before midterms start to look political and the Fed avoids that wherever possible (or it did, although the Warsh/Trump link can only be ignored by fools).
Why does all this matter to your portfolio in Birmingham or Leeds rather than a trading desk in Manhattan? Because the contagion channel never closed. US CPI for July came in at 3.4% with core at 2.5% - cooler than feared, but still miles from target with a divided committee and a war premium in the energy complex. They will swallow 3, for a bit, with deflationary chat around AI and similar. 3.x - harder to ignore. When the US long end sells off, gilts get dragged along in its wake, and UK swap pricing - the thing that actually sets your five-year fixed - moves with it. A hawkish surprise from Warsh on Friday would reach your mortgage broker by the following Tuesday. A dovish one would too, to be fair, and I know which one I would rather have. We'll see. This is exactly why there is conversation and action between Warsh and Bessent about buying longer-term bonds, controlling the yield curve if you will (I’ve written extensively about this before, but not for a few years - it might well be time to revisit that in the near future), and issuing more short-term paper. It works if you are really going to fix the underlying problems - but with the amount of pressure on the debt pile thanks to the medical, welfare and defence commitments, none of which are mathematically showing ANY signs of subsiding and indeed are growing at a frightening rate in nominal terms - what’s the answer? There isn’t one without a genuine crisis. “Never waste a good crisis” - remember that.
Phew - stepping away from the transatlantic macro-storm, let's get back to the reassuringly measurable reality of the real-time UK property market.
As is customary, Chris Watkin has been relentlessly crunching the portal numbers and then publishing them at Property Industry Eye. His analysis for Week 32 of 2026 - the week ending 16th August - is where it is at, as always. If you want to know how the macroeconomic gridlock translates to the local high street, and the REAL property market on the ground, look no further. And the headline this week is a relief: after last week's unexpected summer wobble, the market bounced back. Exactly why we don’t reach for the valium on a rough week, folks.
Supply first. 32.6k new listings this week, up from 31.5k last week, against a 2026 weekly average of 36.4k and a ten-year average for this point of the year of around 33k. So listings are running fractionally below the long-term norm for mid-August, which is what you would expect with half the country on a sun lounger thanks to the excessive heat in the week in question. The year-to-date picture is the more interesting one: 1.166m new listings so far in 2026, which is 0.3% BELOW 2025's pace, 3.2% ahead of 2024, and 10.5% above the 2017-19 pre-Covid average. Regular readers will remember my year-plus adage of "10% more stock than a normal market" as the ready reckoner - after drifting above 12% earlier in the summer, the gap has settled back to almost exactly the reckoner level. Remember though, it is the bathtub that is fuller than it would normally be - very close to the overflow - but the rate of water is now decidedly slower than normal. I wouldn’t call that mean reversion after one reading, but the supply glut story of 2025 and early 2026 is clearly not accelerating any further now. We’ve a couple of months of data to support that at this point in time. The enthusiasm to sell is still there; the frenzy is not.
Demand, then - and this is the bounce. 24.4k homes went sold subject to contract in Week 32, up from around 21.7k the week before, and back within touching distance of the ten-year week-32 average of 25.2k and the 2026 weekly average of 24.6k. Chris's read, which I share, is that last week's drop looks like a hot-weather-and-holidays hangover rather than anything more sinister - and I flagged as much last week when I said reserve judgement. One week does not make a wobble, and one week does not unmake one either, but the recovery to trend is exactly what a "nothing to see here" explanation would predict. The Five Ds - death, debt, divorce, downsizing and the diddy ones - do not take August off, after all.
The year-to-date demand picture is less cheerful, and it is worth being honest about it: 786k gross sales YTD is 7.1% below 2025's pace and 0.7% below even 2024, although still 6% above the pre-Covid 2017-19 average and almost exactly on the decade average of 784k. Translation: 2026 is a normal year for transactions following an exceptional 2025, which pulled forward a chunk of activity ahead of the stamp duty reset and the Budget. Functional, not frothy. Net sales tell the same story: 18.2k for the week (up from 16.9k), 609k YTD, 5.7% below 2025 but 4.2% above the pre-Covid average and 13.5% ahead of the limp lettuce that was 2023.
Now the friction, which is where this market continues to earn its "brutally price sensitive" description. In July, 79.6k homes exchanged and completed while 76.3k withdrew from agents' books unsold - both figures will rise as late reporting comes through, but on the current cut only 51% of homes leaving the books in July actually sold. Call it half, against a seven-year average of 57.6%. Half the homes leaving estate agents' books are leaving without a sale. I keep quoting this stat weekly and I will keep doing so, because it is the single best corrective to the "market is fine" and "market is dead" headlines alike - the market is fine FOR CORRECTLY PRICED STOCK and dead for the rest. Those who have actively lost money - a list growing by the day with a falling London market - either don’t take the losses unless they absolutely have to or are cannon fodder for agents who know no other way other than to overprice stock to get instructions.
And the pricing evidence backs all of this up with unusual clarity this week. Chris reports that four out of five homes listed and sold so far in 2026 achieved that sale without needing a price reduction - while 22.5k homes were reduced this week alone, with 13.7% of the entire stock of 767k homes for sale reduced during July (against a six-year average of 11.2%). Two markets, one country: priced right first time and selling, or priced wrong and beginning the long salami-slice descent. The gap between the average asking price of all listings (£392k) and the average asking price of the homes actually going sale agreed (£357k) is running at 9.9% on Chris's current measure, against a longer-term average he puts at 16-17% - this is a typical August anomaly but will soon completely blow out the other way when the first week of September’s figures are reported. Don’t conclude too much at this point.
Here is this week's back-of-the-envelope, and I will hedge it as rough working as ever. The sell-through rate - the percentage of homes on agents' books going sale agreed in a month - was 14.2% in June, against a pre-Covid average of 15.5%. Compound 14.2% a month over a typical 12-week sole agency term and you get roughly a 37% chance of your home going under offer within the contract period (1 minus 0.858 cubed, for the fellow spreadsheet obsessives, although that should be a 13-week period or even more technically just over a quarter of a day longer than that!). Now overlay Hamptons' research from Monday: 63% of offers from cash-backed landlords in July were pitched at least 10% below asking, versus 25% of offers from first-time buyers and 27% from home movers. So the coldest, most data-driven money in the market is systematically bidding a tenth below asking while retail buyers bid close to it - and roughly half of everything that leaves the books leaves unsold. If you are buying, that is your negotiating context in one paragraph. If you are selling, it is your pricing memo. Rough numbers, generously rounded, but the direction is not in doubt.
The fall-through rate came in at 25.7% against a decade average of 24.5% - slightly elevated, all far too high of course, but all numbers have to be looked at in the context of the norm - while only 5.07% of homes sold STC actually fell through in June, below both the 2025 average and the ten-year average. Pipelines stood at 487k on 1st August, down from 508k a year earlier, with total stock at 767k against 763k twelve months ago. Stock flat year on year, pipeline thinner: the machine is processing what it is fed, it is just being fed slightly less.
On pricing itself, July's agreed sales averaged £345.41 per square foot - 1.2% ahead of twelve months ago and 11.9% ahead of five years ago, and down from June's record £350.22. Sub-inflation nominal growth, which is to say falling real prices, which is to say improving affordability. Don't tell the Guardian - there's no ragebait here, after all. Worth noting Rightmove's August index landed on Monday singing from the same sheet: new seller asking prices down 2.0% on the month, the largest August drop since 2018, with the number of homes for sale at a 12-year high for the time of year. Asking prices in one month doing what the salami slicer usually takes three to do - sellers, or at least their agents, appear to be learning.
The rental side: £1,780 per month was the average asking rent in week 32, with August 2026 averaging £1,805 against £1,800 in August 2025 - so essentially flat year on year at the national level on Chris's portal measure, and remember this is asking rents on new listings, the most real-time cut there is. Against that, £1,394 in August 2021 tells you what the last five years did. Availability continues its quiet improvement: 323k rental properties available in July against 319k a year ago, with 135,928 new rental listings in July - up from 128,821 last July and massively up from the 111,080 of July 2022. The ONS's own rental index, published Wednesday, has rent growth accelerating even as house price inflation slows, with the average London rent at £2,317 and the North East at £783 - the ONS measures all tenancies including renewals (the data sample from Dataloft is c. 550k tenancies strong), Chris measures new listings, so the two can happily disagree for months at a time. The stock-shortage era is easing at the margin on the portal data; the ONS is still catching up with the rises already banked. Both things are true. More on what landlords themselves say about all this in the Deep Dive, where the surveys this week started arguing with each other.
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! This week Chris is joined by the vastly experienced Iain White to go through the numbers and then run the data ruler over Oxford's estate agents - well worth your time.
Dust off the Macroscope, then. Unemployment and the labour market report. Inflation. The flash PMIs, with a side of retail sales. This is "meat week", folks. Enjoy. Bringing up the rear the contract states that we have to talk about the gilts and swaps, and will do so for many years to come, I'm sure.
The labour market report opens - as it will do for some time yet - with the payroll count, and the count is still shrinking. 30.3 million payrolled employees, down 78,000 on the year to June, with the April-to-June quarter (the period matching the survey data) down 86,000 on the year and 37,000 on the quarter. The early estimate for July shows another 94,000 down on the year but essentially flat on the month - and these early estimates are notoriously bad, so reserve judgement until next month, as ever. The direction has been unmistakable for two years now; what has changed is the pace, which has slowed from evaporation to seepage. The month-on-month numbers have been hovering around flat since the spring. Hopefully the worst is over, although I have typed that sentence before. The assumption at Bank of England level was that the employment market would take 2 full years to absorb the pain inflicted in the November 2024 budget on employers when it came to Employers’ National Insurance Contributions - and we aren’t there yet, after all.
Then we get into the three segments, in the order we always take them, because unemployment alone tells you almost nothing. The employment rate - the share of 16 to 64 year olds with a job - was 75.1% in April to June, down 0.2% on the year but up 0.1% on the quarter. The unemployment rate - those aged 16 and OVER who are looking - was 4.9%, up 0.2% on the year but down 0.1% on the quarter. And economic inactivity - 16 to 64s neither working nor looking - held at 20.9%, with 9.11 million people in that category, up 55,000 on the year. Note the denominator anomaly that persists in this data: the unemployment rate has no upper age cap, so a 70-year-old seeking work counts as unemployed while never appearing in the employment or inactivity rates, which is why the three numbers refuse to add up neatly and why anyone quoting only one of them is - deliberately or not - telling you a fraction of the story.
Where does the arc stand? Under this government the employment rate has gone from 74.6% at the start to 75.1% now, having touched 75.3% along the way - progress, then a stall, then this quarter's tentative uptick. Pre-Covid we had hit 76.5%, and that remains the number to aspire to; at the current pace we would get there sometime next decade, which is not a forecast so much as an arithmetic complaint. Vacancies fell again to 707,000 in May to July, below pre-pandemic levels, and the ratio of unemployed people to vacancies has climbed to around 2.5 - the highest since the middle of the pandemic. That ratio is the single best summary of bargaining power in the jobs market, and it has been moving the employer's way for two years straight.
Wages are where the squeeze shows its other face. Annual growth of 4.1% including bonuses, 3.5% regular - which, against this week's inflation print, means real pay growth of 1.3% and 0.7% respectively. Positive, but thinning, and thinning at exactly the moment inflation has turned back up. A labour market that is loosening while real wage growth fades is the textbook picture of an economy that has had demand squeezed out of it - and yet, as we shall see in a moment, the activity surveys are perking up. One caveat before we move on, and it connects directly to this week's Deep Dive: the Labour Force Survey behind these rates still carries the ONS's own "official statistics in development" label, with sample sizes recovering but volatility acknowledged. We work with the data we have, while remembering who the village watchman is.
OK. Inflation. The CPI rose 2.9% in the twelve months to July, up from 2.6% in June - the first increase in the annual rate since March - with CPIH at 3.1% and RPI, for the contracts and student loans still chained to it, at 3.2%. Was this a surprise? Not remotely. The Bank's own August projections had the peak around 3.2% in the fourth quarter, and July was always going to be the month the war arrived in the household energy bill. And so it proved: the single largest driver was housing and household services, up from 2.7% to 4.1% annually, courtesy of the July change in the Ofgem price cap - a rise of £221 to £1,862 for the typical dual-fuel direct debit household. Gas prices rose 14.7% in the month, the largest monthly rise since October 2022, taking gas to its highest level since March 2024. The mechanism is worth spelling out because it explains the timing: Ofgem's assessment window for the July-to-September cap ran from 18th February to 18th May - the first window to capture the outbreak of the war on 28th February. Your July gas bill is the March oil market, laundered through a regulatory formula with a five-month delay. And the next cap announcement, covering October to December, is due this coming week, with an assessment window that caught the June spike AND the July round trip. Place your bets.
Now the underlying detail, which is considerably more interesting than the headline and - whisper it - considerably more encouraging. Core CPI was UNCHANGED at 2.6%. Services inflation, the Bank's obsession and rightly so, actually EASED from 3.6% to 3.4%. The entire acceleration is a goods story: goods inflation rose from 1.7% to 2.2%, all of it energy. Food inflation fell to 1.3%, its lowest since September 2021, and its contribution to the index is the smallest in nearly five years. So the domestically generated, wage-driven, hard-to-kill portion of inflation is still grinding lower while the imported, war-driven portion pushes the headline up. If you sit on the MPC, which number do you set policy on? The hawks will say headline drives expectations and expectations are the whole game after a shock like this. The doves will say you cannot hike your way out of a gas price. Both are right, which is precisely why the committee split 6-3 last time out and why September's meeting - the 17th, mark your diaries - is live in both directions and probably neither. 6-3 or 5-4 to hold look like the favourites, with 5-4 to hike not being out of the running at this point in time.
Some texture from the divisions, because the texture is telling. Transport went the other way entirely, slowing from 5.7% to 3.6% as motor fuels dropped out - diesel fell 8.8p in the month to 167.6p a litre, petrol 3.1p to 152.2p. Airfares produced my favourite footnote of the release: European fares FELL 4.3% in July against a 38% rise a year earlier, while long-haul fares rose 31.7% - because the war has closed Middle Eastern airspace and airports to many carriers, crushing long-haul capacity while short-haul demand softens. The 10% or however many book late/last minute also didn’t bother, of course, because of the very clement weather available in the UK, and staycations have done well on the back of this. There is a whole geopolitics lesson in one row of the CPI basket. Furniture recorded its smallest July price fall since 1989, and clothing its smallest July fall since 2020 after retailers started discounting earlier in June - which is a neat bridge to the real economy data in a moment. For international context, our 2.9% sits above Germany's flash 2.8% and France's 2.4%. Not an outlier, not a triumph. The path from here, per the Bank, is a grind up towards 3.2% or so in Q4 and then a slow descent through 2027 - this certainly does not feel like a runaway inflation problem, but I remain unconvinced of any orderly glide back to target while the war premium sits in the energy complex and the Q4 cap looms. What has surprised me a little, and we have to factor in going forward (that the Bank actively don’t, because it is political) is Burnham’s fixation on the cost of living situation - expect more measures that will lower the cost of living, and have a direct impact downwards on inflation (we’ve already seen some of it as some, including me, would have expected more inflation earlier than we’ve seen it so far) - likewise, we cannot possibly predict the oil price over the next 12 weeks, let alone the next 12 months (if you can, send me a DM!). We'll see, quite literally, this week as far as the Q4 pricing for the cap goes.
Onto the PMIs, my darlings of the real time economy - the flash readings for August, hot off Friday's press, with July's retail sales riding shotgun. And the darlings delivered an upside surprise: the composite output index rose to 52.5, a four-month high, against expectations of a fall to 51.6. Services did the lifting at 52.8 (consensus 51.8, up from 52.1), while manufacturing cooled to a five-month low of 51.5 as the precautionary stockpiling that flattered the spring readings unwinds. Chris Williamson - whose commentary I have quoted for years, and who like me tends to look past the headline to the composition - reads this as consistent with GDP growth of around 0.3% in the third quarter, crediting sunny weather and the technology investment boom, the same AI capex story that showed up in the official business investment figures we covered last week. Two consecutive months above 50 after the spring contraction; the economy that was supposed to be flattened by the war has instead been merely dented.
But underneath, as ever, is where the meat is. Service sector employment fell AGAIN - the 23rd consecutive month of decline, which S&P Global notes is the longest such run since the survey began in 1996. Twenty-three months. Firms are not sacking; they are simply not replacing leavers, citing the cost of employment - which is the National Insurance and minimum wage story of the last two Budgets showing up as a slow puncture rather than a blowout, exactly as I suggested it would when the measures landed. This is also the same story you hear from companies in Europe that have implemented a heavy amount of AI. Not shedding - generally - but implementing and then not REPLACING leavers. And cost pressures re-accelerated in August from July's five-month low, driven by energy - the survey was collected 12th to 19th August, so this is the war premium hitting input costs in real time, before it reaches any official index. Williamson's summary for the Bank: a hawkish bias, but caution, with no hike until the trajectories clear. That is also, for what it is worth, roughly my read of the September meeting - and why my current forecast falls between 6-3 and 5-4 voting to hold, with a lean on 6-3.
Retail sales, briefly, as the third leg of the real-economy stool: volumes fell 0.5% in July after rising 0.7% in June and 1.3% in May, with the three months to July still up a healthy 1.1%. The July dip looks mechanical - retailers pulled demand forward into June with early discounting, and the heatwave-and-World-Cup spending burst (fans, barbecues, beer - the great British summer portfolio) was never going to repeat monthly. Consumer confidence, per the commentary around the release, is at a two-year high. A consumer with thin but positive real income growth, spending selectively, confidence recovering: not a boom, not a bust, a grind. Which has rather been the theme of this entire Macro section, now I look back at it. Treacle, I first said, back in early 2023 - and treacle is still very much on the menu as a main course, not a dessert option (sadly).
Gilty, or not Gilty? The court reconvenes, and this week the evidence arrives mostly from across the Atlantic. The 5-year gilt closed the week at 4.595% from 4.547% on Monday, with the 30-year at 5.8104% up from 5.765% at the start of the week. Twelve months ago the 5-year stood at 4.094%, so we’ve lost 50 bps the wrong way here. The week's shape: heavy early, following the US long end - which touched levels not seen in nearly two decades around the FOMC minutes - then a partial retracement as the flash PMIs and the in-line US CPI calmed nerves into the weekend. The UK curve continues to trade as a high-beta passenger on the Treasury bus: we did not vote for the driver, we cannot reach the pedals, and the bus is currently being driven by a man who has reduced the number of scheduled stops.
Swaps: the 5-year SONIA at 4.32% and the 3-year at 4.25%, hardly moving on the month, but up from the 3.7s 12 months ago (so we are 53 and 54 bps adrift respectively from 12 months ago). The mortgage market evidence this week points one way: Rightmove's weekly mortgage tracker has average fixed rates back above the psychologically important 5% mark, and the sub-5% five-year money that briefly reappeared in the early summer has been quietly repriced or withdrawn. Bank Rate, for the avoidance of all doubt, remains at 3.75% - the market noise about 4.00% is the forward curve and three dissenting MPC votes talking, not the current rate - and the gap between where the Bank IS and where the swap market says it is GOING remains the entire story of mortgage pricing in 2026. September 17th brings the next MPC decision alongside the annual QT envelope decision, with Jackson Hole's tone-setting five days after this edition lands. If Warsh hawks it up on Friday, expect your broker's rate sheet to notice by mid-week. If he doesn't, the relief rally has room to run - a couple of quiet weeks on the geopolitical front would do more for five-year money than anything the MPC says. Neither is guaranteed, and I would not put real money on either. What this means in practice: if you have completions between now and November, the case for securing rates rather than surfing the curve remains what it was last week and the week before. Boring advice ages well. As alluded to in Trumpwatch - one world sees rates go DOWN alongside Yield Curve Control measures, that will be heavily scrutinized. Don’t mess with the Bond Markets though - they are as undefeated as Rocky Marciano was.
OK. Here endeth the lesson on current rates - here comes the Deep Dive. The through-line this week, as trailed at the top, is ground truth: what the numbers actually say underneath what the headlines say they say. We have the landlords' own survey data from Pegasus Insight and Paragon, which cheerfully contradicts the surveyors. We have the NRLA holding the housing benefit system's feet to the fire ahead of the Budget. We have Knight Frank quantifying Prime Central London's decade of decline in terms polite company avoids. And we have the ONS itself confessing, in the annual Blue Book exercise, that economic history is a first draft subject to revision. The village watchman has been busy.
First up, the landlords' own ground truth. Pegasus Insight's Landlord Trends research for the second quarter of 2026 landed this week, published through Paragon Bank, whose long-running landlord research programme it powers - you will have seen the findings written up twice under two names in the trade press this week, but it is one survey, and it is the best panel data on what actual landlords are actually doing that exists in this country. Four themes.
Theme 1: Tenant Demand Turns Back Up
The Summary: The proportion of landlords reporting strong or very strong tenant demand rose to 63% in the second quarter of 2026, a five percentage point increase on the previous quarter and the first rise in two years. Demand perception had declined steadily from the exceptional levels recorded in 2024, falling to 61% by the fourth quarter of 2025 and reaching a low of 58% in the first quarter of 2026 before this quarter's recovery. Pegasus Insight notes that tenant demand has been gradually normalising from the extreme conditions of 2024, and characterises a five-point quarterly rise as evidence that the underlying need for rental accommodation remains considerable. The firm also highlights that the recovery in demand arrives while supply in the private rented sector remains under sustained pressure, with its research consistently showing landlords are considerably more likely to sell rental properties than to purchase them over recent quarters, raising questions about the availability of homes should demand continue to strengthen.
The Propenomix Perspective: Here is where it gets interesting, because eight days ago the RICS survey told us tenant demand had turned NEGATIVE for the first time in the post-2020 era, and we discussed it at length. So which is it? Both, probably - and the difference is instructive rather than embarrassing. RICS polls surveyors and letting agents, who see applicant volumes at the front desk; Pegasus polls landlords, who feel occupancy, arrears and the speed at which a void fills. An agent can see fewer applicants per property (down from the deranged 20-plus of 2023) while a landlord still lets every property within days at full asking rent. More and more have defected to OpenRent, which RICS can never discuss of course as they are not helping their members with their businesses. Demand can normalise and remain considerable at the same time - the queue shortening from twelve to six is a collapse in one dataset and business as usual in the other. My read, hedged as ever: the froth has gone, the floor has not. And with 135,928 new rental listings in July on Chris's data, supply is finally responding at the margin, which is exactly when you want to know whether your local market has a floor. Mine do. Check yours.
Theme 2: Arrears at a Record Low
The Summary: The proportion of landlords reporting rent arrears during the preceding twelve months fell to 26% in the second quarter, down four percentage points from 30% in the first quarter and the lowest reading in the history of the research programme. The report frames this as evidence of a resilient sector despite continued market uncertainty, and situates it within a longer-run pattern in which buy-to-let credit performance has consistently outperformed the owner-occupied sector: mortgages three or more months in arrears have been proportionately lower in buy-to-let than in owner-occupation in all but one of the past 26 years. Alongside the arrears data, 86% of landlords reported making a profit from their lettings activity in the quarter, up two percentage points on the previous quarter, indicating that the overwhelming majority of portfolios remain cash-generative despite the higher interest rate environment of recent years.
The Propenomix Perspective: A record low in arrears, in year two of a war-driven cost squeeze, with real wage growth at 0.7% - I will admit that surprised me, and I try to let the data surprise me rather than argue with it. Two mechanisms suggest themselves. The first is selection: after ten years (some partial, sure) of Section 24, higher rates and regulatory attrition, the tenancies that exist are the ones that survived underwriting by increasingly careful landlords - we have collectively become better at referencing because we can no longer afford not to be. The second is prioritisation: when the essentials get expensive, the roof gets paid first, and the arrears show up in the discretionary economy instead - which is precisely what the retail data's "selective consumer" is telling us. Either way, the 26-year credit record deserves more airtime than it gets. The narrative says buy-to-let is the risky end of mortgage lending; a quarter century of arrears data says it has been the SAFER end in 25 years out of 26. I doubt that makes next year's risk weightings any kinder, but it should. Cui bono? The banks with their “risk premia” have no justification at all that’s grounded in fact, in this case.
Theme 3: Yields Above Seven
The Summary: Paragon's accompanying rental yields analysis found average gross rental yields across landlord portfolios reached 7.02% in the second quarter of 2026. The yield figure sits alongside the profitability data - 86% of landlords in profit - and reflects the combined effect of several years of strong rental growth against a backdrop of broadly flat nominal property prices across much of the market. The research positions current yields at or near the strongest levels recorded in the programme's recent history, with the bank characterising the underlying health of the buy-to-let sector as considerably stronger than public commentary tends to suggest, a point made explicitly in its presentation of the findings against what it describes as ongoing market uncertainty.
The Propenomix Perspective: Seven point zero two. Run the clock back to 2021 and the equivalent national average was scraping along in the fives - what has happened since is five years of double-digit cumulative rental growth meeting a sales market that has gone sideways in nominal terms and backwards in real terms. This is the quiet repricing I have banged on about for years: the market never crashed, it just stood still while the income caught up, and the result is that the entry maths for a professional buyer today is better than at any point since the mid-2010s - on paper. The hedge, and it is a load-bearing hedge: gross yield is a vanity metric if your debt costs 5%-plus and your compliance bill keeps compounding. The spread over five-year money is perhaps 200 basis points gross for the average portfolio (so this would mean the average weighted debt cost that the average portfolio is carrying is around 5%, of course - not that today’s price is 5%, because it isn’t); the spread over the 14 years to 2022 was routinely 50% better, or 1.5x that that against far cheaper debt - more like 300 basis points. Better than 2023, worse than 2015, and entirely dependent on buying the right stock at the right discount - which, per Hamptons and the withdrawal data earlier, is currently very much on offer for those with the patience to insist on it. If you are in the (just about) majority of encumbered landlords, as many of my readers and my peers are (alongside me), then you’ll be fast to point out that 5.5% when debt was 2.75% is superior to 7.02% when new debt is 6% - and, of course, you aren’t far wrong. If you are unencumbered, you’ve enjoyed the normalisation of interest rates although you might still be thinking hang on - those 30 year bonds at 5.81% look pretty incredible and I’d suggest you aren’t far wrong.
Theme 4: Sellers Still Outnumber Buyers
The Summary: Despite the improving demand, arrears and yield picture, the research continues to find landlords materially more likely to report an intention to sell property over the coming year than to purchase, extending a pattern that has persisted across recent quarters of the programme. Pegasus Insight explicitly flags the tension between the two findings: strengthening tenant demand set against a supply base that its own intentions data suggests will continue to contract, with the firm raising concerns about the availability of rental homes if demand continues to build. The intentions data does not distinguish in the headline between full portfolio exits and partial disposals, and the research notes that stated intentions have historically overstated eventual sales activity, though the direction of the gap between intended sales and intended purchases has been consistent.
The Propenomix Perspective: The gap between what landlords say and what landlords do is one of my favourite datasets that doesn't quite exist. Intentions surveys have predicted a mass exodus every year since 2016; the actual attrition has been real but gradual - the 834k-property decade exodus we covered a few weeks back, not a stampede (and please remember that measures the outs, not the ins). But note what happens when you stack this quarter's four findings on top of each other: demand up, arrears at record lows, yields above seven, and the marginal landlord STILL wants out. That is not a business problem, because the business has rarely looked better on the operating line. It is a regime problem - tax, regulation, and the sheer administrative weight of the Renters' Rights era pushing people out of an asset class their own numbers say is performing - and of course, alternative investments that aren’t deemed “too risky” by the average UK punter - bond yields are high and that translates through into nice returns in savings accounts, the likes of which were invisible between 2009 and 2022. For those of us staying, the arithmetic is almost uncomfortably favourable: every departing amateur is supply removed from my competition and stock added to my buying funnel, at a 10% discount if Hamptons' offer data is any guide (those who know me know I buy cheaper than that - the business model dictates that I have to). I have said before that this is the professionalisation of the sector by attrition, and nothing in this quarter's data changes my mind - although I hold the view a little more lightly than the confident version of me from 2021 would have.
Second, then, to the NRLA, who chose Budget-countdown season to publish analysis on the least glamorous, most consequential number in the rental market: the Local Housing Allowance. Their argument, compressed: unfreezing housing benefit will not blow up rents, and freezing it is quietly blowing up everything else.
Theme 1: The Freeze and What It Has Done
The Summary: Local Housing Allowance determines the maximum housing support available to private renters claiming benefits. Introduced in April 2008 to cover the cheapest 50% of rents in a local area, it was cut to the 30th percentile in April 2011, repegged to the 30th percentile in April 2024 after a multi-year freeze (frozen 2014-2020), and then quickly refrozen from April 2025, with ministers now deciding whether the freeze continues from next April. Because frozen cash rates meet rising market rents, coverage erodes automatically: the homelessness charity Crisis estimates that fewer than 2% of private rented properties currently listed are affordable to those relying on the benefit, and that almost half of the 1.6 million Universal Credit households renting privately face a shortfall between their housing support and their actual rent. The NRLA's analysis lands amid a co-ordinated push, with some forty organisations writing jointly to Government calling for rates to be unfrozen ahead of the Budget.
The Propenomix Perspective: Strip the politics out and look at the mechanism, because it is a beautiful piece of bad design. The state sets a maximum rent it will support, based on rents as they were at some frozen point in the past, in a market where rents have since risen 20%-plus. The tenant cannot bridge the gap, the landlord cannot absorb it, and so the benefit-dependent tenant is simply priced out of 98% of the advertised market - not by malice, but by arithmetic nobody updated. Where do they go? Temporary accommodation, at nightly rates that would make a Mayfair concierge blush, paid by councils already teetering - we covered the HRA debt mountain and the temporary accommodation bill earlier in the summer, and this is the same story wearing a different hat. A system designed to help people rent has become a system that decides who cannot. I would suggest that if you designed this from scratch you would be sacked, but of course nobody designed it - it froze, which in Whitehall is a decision that never has to be announced. Pushed around on a spreadsheet, at a much higher cost, mostly blinded by ideology here and Rayner is as guilty as anyone, and you get the sense that Burnham agrees with her. Hopeless. When Shelter and the NRLA agree, and agree violently - people should be listening. They aren’t. Frankly you might as well abolish the system and rebuild it from the ground up. Put it back to the 20th percentile if that’s what you need to do, and inflation-protect it as you should have done in the first place - but at least be honest.
Theme 2: Does Benefit Chase Rent?
The Summary: The central empirical claim of the NRLA's analysis is that there is no clear historical link between the level of housing benefit and the pace of rent increases. The association examines the period from 2008/09 to 2015/16, when LHA rates initially tracked local rents, and reports that average weekly rents across England rose by roughly 2.5% a year during that era - from £153 to £184 on the English Housing Survey measure - rates of growth well below those recorded during subsequent freeze periods. The analysis directly challenges the Treasury view, recently echoed in the Prime Minister's remark that the benefit system is being "forced to chase rents" in the private rented sector, that raising LHA feeds through into higher rents. The NRLA's chief executive Ben Beadle argues rents are set by tax, mortgage costs, tenant demand and regulatory costs, with separate analysis suggesting restoring rates to the 50th percentile could lift 130,000 children and 215,000 adults out of poverty.
The Propenomix Perspective: Is the NRLA right that benefit does not chase rent? Broadly, I think yes - with a caveat they would perhaps rather I skipped. In a market where LHA claimants are 2% of effective demand for advertised stock, raising the benefit ceiling cannot move the market price; it merely determines whether that slice of tenants can participate at all. The incidence argument that sank Help to Buy - subsidy capitalises into price when supply is fixed - applies weakly here precisely because the LHA cohort is no longer the marginal bidder in most markets; the working tenant is. The caveat: in the specific submarkets where benefit tenants ARE the marginal demand - parts of the North East at £783 average rents, certain coastal towns - some pass-through is plausible, and pretending otherwise is advocacy rather than analysis. But as a national Budget question? The cost of unfreezing is real money, low single-digit billions; the cost of the freeze is homelessness statistics, temporary accommodation invoices and council insolvency, which is also real money wearing a disguise. Fiscal drag's shabbier cousin, and every Chancellor since 2016 has found it equally irresistible. It costs MORE when you look at the total costs - and I’ll prove it in PropenomAIx in an upcoming deep dive, without hesitation or doubt.
Theme 3: The Landlord's Interest, Declared
The Summary: The NRLA is the largest membership body for private landlords in the UK, and its analysis is explicit about the commercial dimension of its case: frozen LHA rates, it argues, lock benefit-dependent tenants out of the private rented sector altogether, narrowing the pool of viable tenancies for its members while pushing vulnerable households towards homelessness services. Beadle's framing calls on Government to set support against "housing costs as they actually are, not as they were in the past". The Institute for Fiscal Studies is cited within the surrounding debate, having warned that a tight fiscal environment is no excuse for a system that creates uncertainty for renters and unfairness between local areas, while the Renters' Reform Coalition and allied tenant groups argue the priority should instead be limiting rent increases directly rather than raising the subsidy that meets them.
The Propenomix Perspective: Let me be careful here, because I am a landlord reading a landlord lobby's paper and nodding, which is exactly the moment to check one's working. Yes, the NRLA's members benefit from unfreezing - a paying tenant beats an empty property and a council beats a chancy guarantor. Does the self-interest invalidate the analysis? No - but it does explain the framing, and the counter-position deserves its airing: the tenant groups' argument is that raising LHA subsidises rents landlords set, and the cleaner fix is more social housing so the state stops renting privately at retail prices altogether. I’m a huge supporter of funding social housing properly, not only because in the long run it makes economic sense for similar reasons as to why housing benefit freezes COST money to the state in a net-net proper calculation including the fallout costs. This is about the length of the time horizon - over the long term, sorting out social housing is the ONLY answer and I have said so before - the 119-years-to-clear-the-queue arithmetic has not improved since June when we discussed it. But Budgets are set on the short horizon, and on the short horizon the choice is not "LHA or social homes"; it is "LHA or temporary accommodation at treble the price". Some choices are only hard if you refuse to add up both columns. Whether this Chancellor adds up both columns on 28th October is another matter - the fiscal arithmetic we keep hearing about does not leave much room for compassion, however cost-effective.
Theme 4: The Budget Politics of It
The Summary: The timing of the intervention is explicitly pre-fiscal-event: the government must decide whether LHA rates remain frozen from April next year, making this an active Budget question rather than an abstract policy debate. The Prime Minister's own past statements feature prominently in the coverage - Andy Burnham argued during his leadership campaign that freezing LHA "makes families homeless and places unfunded pressures on councils when they have to pay for temporary accommodation", and that the housing crisis is having "a ruinous impact" on the public finances. The NRLA's analysis, alongside the forty-organisation joint letter and pressure from Crisis, Shelter and the IFS's commentary, effectively asks the government to honour in office the position its leader took on the campaign trail, with the estimated Exchequer cost of unfreezing set against projected savings in homelessness and temporary accommodation spending.
The Propenomix Perspective: So the Prime Minister is on the record, in terms, agreeing with the NRLA - which makes this less a lobbying exercise than a public presentation of the invoice. Will he pay it? The £64k question, or more precisely the low-billions question, in a Budget where every billion is spoken for twice. My suspicion - and it is only that - is a partial thaw: an unfreeze dressed as targeted support, phased, with the 30th percentile promise deferred to "when fiscal conditions allow", that phrase which does so much heavy lifting in modern Treasury prose. For landlords the practical point sits regardless of the politics: the benefit-adjacent end of the market is where the state's dysfunction meets your rent roll, and whether that segment reopens as viable demand next April or stays frozen out is worth several percentage points of effective demand in the markets many of us operate in. Watch this one on Budget day; it will not make the headlines, and it will matter more than most things that do.
Third this week: Knight Frank published its latest read on the prime London sales and lettings markets, and buried in the release is a number that deserves to be far more famous than it is. Regular readers know I am not a Prime Central London buyer and never have been - my money lives where the spreadsheets work - but PCL is the canary in this country's capital-taxation coal mine, and the canary has now been coughing for eleven years.
Theme 1: Thirty-Nine Months and Counting
The Summary: Average prices in Prime Central London fell 3.3% in the year to July 2026, which Knight Frank records as the 39th consecutive month of annual declines. The firm attributes the sustained fall to a succession of tax increases and political uncertainty rather than any single event. The longer arc is striking: prices rose modestly over the two years to April 2023, but that brief recovery followed a 59-month period of decline running from June 2016, and average values now sit 23% below their last peak in mid-2015. Prime outer London has held up considerably better, with steadier pricing and rents rising 3.2% in the year to July. On the lettings side, Rightmove data cited within the research shows new rental listings across prime central and outer London running 14% below the five-year average during the first half of 2026, sustaining upward pressure on prime rents even as the sales market languishes.
The Propenomix Perspective: Thirty-nine consecutive months of annual price falls is not a correction; it is a policy outcome. Stack the timeline: the 2014 stamp duty reform, the 2016 additional-dwelling surcharge, Brexit, the non-dom squeeze, cladding and building safety costs on the flat-heavy PCL stock, the 2024-25 tax rounds - each one individually defensible, cumulatively a decade-long experiment in what happens when you tax the top of a market continuously for eleven years. The answer, now in: it goes down 23% and stays down. And before anyone reaches for the world's smallest violin, note who actually got hurt - not primarily the oligarchs, who repriced and moved on, but the leaseholders in the flats underneath, marked to a falling market with remediation bills attached. Would I buy the dip? No - I said the same when Savills' prime numbers crossed this desk three weeks ago, and consistency demands I say it again: falling knife, thin yield, and a Budget in 66 days with wealth taxes openly back on the discussion menu. But I watch it closely, because PCL is where this country tests its tax ideas before they travel north.
Theme 2: The Real-Terms Reckoning
The Summary: The more arresting arithmetic in the coverage concerns inflation adjustment. Cumulative UK CPI inflation between mid-2015 and mid-2026 amounts to roughly 42%, meaning a property purchased in Prime Central London for £1 million at the 2015 peak would need to be worth approximately £1.42 million today simply to have held its real value. Knight Frank's index implies that the same property is actually worth around £770,000 in nominal terms. The combination - a 23% nominal decline against a 42% inflation backdrop - leaves PCL values approximately 46% below their 2015 level in real terms. The analysis also notes the decline predates the current government by a wide margin: the 59-month falling stretch began in June 2016, and building safety costs, ground rent issues and quadrupling mortgage rates after September 2022 all contributed well before the most recent tax changes took effect.
The Propenomix Perspective: Forty-six per cent, real. Let that settle for a moment, because it is the largest quiet destruction of paper wealth in modern British property history, and it happened without a crash, without headlines, without a single dramatic quarter - just eleven years of nominal drift while inflation did the demolition silently. This is the mechanism I have been describing when I’ve been talking about how markets actually correct in this country since 2022: Britain does not do nominal crashes if it can possibly avoid them; it does real-terms erosion, because falling prices that never print a minus sign in the Daily Mail generate no panic, no forced selling, and no political response. PCL is simply the extreme edition of what the wider market has done since 2022 in miniature, although it started much earlier than that. The lesson for the rest of us is not about London - it is that "prices were flat" and "you lost money" are frequently the same sentence, and the only defence is income. A 7% yield compounding through a flat decade beats a trophy asset falling 4% a year in real terms by a margin that would embarrass both owners if they compared notes. They never do, of course. Different dinner parties.
Theme 3: The Prime Lettings Squeeze
The Summary: While prime sales markets have fallen for 39 straight months, the prime lettings market is moving in the opposite direction. New rental listings across prime London ran 14% below the five-year average in the first half of 2026 on the Rightmove data cited by Knight Frank, and prime outer London rents rose 3.2% in the year to July, accelerating with a 1.2% rise over the latest three months alone. The supply shortfall reflects several forces identified across the research: discretionary landlords exiting in response to the same tax and regulatory pressures weighing on the sales market, some owners selling rather than re-letting into a heavier compliance regime, and others withdrawing stock to sell into any sign of a sales recovery. The result is a prime rental market tightening even as the prime sales market softens, with the two markets fed by the same shrinking pool of stock.
The Propenomix Perspective: The same property, two markets, opposite directions - and the connective tissue is the departing landlord. Every prime flat that sells to an owner-occupier is a rental removed from a lettings market already 14% short of normal supply, so the sales weakness directly manufactures the rental strength. We have watched this exact loop run nationally for three years; prime London just runs it with fewer units and bigger numbers. What interests me is the equilibrium question nobody in the research quite asks: at what rent does prime lettings demand break? Corporate relocation budgets and international student housing allowances are deep, but they are not bottomless, and at some point the arbitrage - rent at a 3% gross yield on a building falling in real terms - stops making sense for the TENANT'S employer too. I suspect prime rents have further to run before that point, but I would not model it continuing past a couple more years without checking my assumptions quarterly. Meanwhile the practical read-through for the rest of the country is confirmation, at the luxury end, of the thesis in the Pegasus data at the volume end: exiting landlords are the single biggest force in the rental market, everywhere, at every price point.
Theme 4: Value Signal or Value Trap?
The Summary: The research and its surrounding commentary present the relative value case with some care: prices 23% below peak nominally and materially cheaper in real terms have begun drawing selective demand back into the market, with agents reporting improved enquiry levels through 2026 and industry data showing under-offer volumes recovering from their lows. Set against that, the report catalogues the reasons for continued caution - the 39-month falling streak has not yet ended, transaction volumes remain thin by historic standards, stamp duty on high-value purchases remains a substantial frictional cost in a falling market, and the political environment ahead of the October Budget continues to generate speculation about further taxation of high-value property. The value argument and the momentum argument, in short, point in opposite directions, and the research stops well short of calling a bottom.
The Propenomix Perspective: "Cheap" and "good value" are cousins who are frequently mistaken for twins. PCL is unambiguously cheap against its own history - 46% off in real terms clears any reasonable bar. Whether it is good value depends entirely on whether the forces that produced the fall are finished, and I count at least three that are not: the Budget (66 days), the building safety tail (years to run on the flat stock), and the global picture, in which the internationally mobile money that built this market now has Dubai, Milan and Singapore actively bidding for it with friendlier tax regimes. Catching the bottom requires those to resolve favourably more or less together. Could they? Perhaps - a benign Budget alone might mark the low, and the under-offer data suggests some smart money is positioning for exactly that. But my framework has never been "buy what has fallen most"; it is buy below replacement cost with income that services the debt, and a £2,500-per-square-foot flat yielding 3% gross fails both tests at any discount to peak. I will cheerfully be wrong about the bottom from the comfort of a portfolio that cash flows. Someone has to buy the canary; it will not be me.
Last up, and it is a properly nerdy one, which regular readers will know is my idea of a treat: on Thursday the ONS published its Blue Book 2026 impact article - the annual exercise in which the national accounts are rebuilt with new methods and better data, and history is formally revised. This year's edition rewrote a date every economics student thought they knew.
Theme 1: What the Blue Book Actually Is
The Summary: The Blue Book is the UK's annual national accounts compendium, and each year's edition incorporates methodological improvements and newly available survey and administrative data, with revisions reaching back decades where methods change. The 2026 exercise, whose full results will be published on 30th October alongside the Pink Book, spans changes across the period from 1997 to 2024. Headline components this year include the integration of the Annual Survey of Goods and Services into the production measure of GDP, an open-period balancing exercise incorporating new and revised data for 2022 to 2024, a review of energy products and industries covering 2017 to 2024, revised treatment of precious metals in trade statistics, improved consistency between the national accounts and the public sector finances, and the first-ever inclusion of estimates of households' own-account electricity production. Thursday's article quantifies the impacts on GDP and its main components ahead of the full October publication.
The Propenomix Perspective: I appreciate that "annual national accounts methodology compendium" is not a phrase that quickens most pulses, so let me translate why this matters to a property investor. Every number we argue about - the growth rate the Chancellor will build the Budget on, the fiscal headroom, the productivity puzzle, the "fastest/slowest in the G7" bragging rights we enjoyed only last week - is downstream of this document. When the Blue Book moves history, it moves the baseline for all of it, and the OBR's Budget-day forecast will be constructed on the revised past. Fewer than one reader in a hundred will open the thing; the other ninety-nine will spend the autumn arguing about numbers it quietly changed. Which is, of course, precisely why the village watchman made the quote of the week - the figures are indispensable AND provisional, both at once, and the trick is holding those two truths without dropping either. Most commentary drops one.
Theme 2: Your Solar Panels Are Now GDP
The Summary: Among the methodological changes, Blue Book 2026 introduces for the first time estimates of own-account electricity production by households - the economic value of electricity generated by households for their own consumption, predominantly through domestic solar panels. Under prior methods, a household consuming electricity it generated itself created no measured market transaction and therefore no recorded output, despite identical electricity purchased from the grid counting fully in GDP. The change brings the treatment of household generation closer to the long-standing treatment of other own-account production in the national accounts, of which imputed rent for owner-occupiers is the largest and most familiar example. With domestic solar installation having grown substantially over the past decade and a half, the revision adds previously invisible output across the historical series, with the effect growing over time in line with the installed base.
The Propenomix Perspective: I love this change beyond all reason. For fifteen years, every panel bolted to a semi in Solihull has been generating real electricity, displacing real bills, and contributing precisely nothing to measured GDP - the economy got bigger and the statistics got relatively smaller, a tiny systematic understatement compounding in the sunshine. Now it counts, and the historical growth rate tilts fractionally upward with it. The property angle is direct, mind: own-account production is the same conceptual family as imputed rent, the fiction-that-is-not-a-fiction by which the national accounts recognise that owner-occupiers pay rent to themselves - a line item worth well over a hundred billion a year and the reason housing looms so large in CPIH. Every EPC-driven retrofit, every panel and battery mandated onto new builds, now shows up twice: once in construction output, once in perpetuity as household production. I am not entirely sure why it took until 2026, but then the imputed rent methodology has been argued about since the 1940s, so by ONS standards this was practically a sprint.
Theme 3: The Recovery That Moved
The Summary: The article's most quotable revision concerns the financial crisis. Following the incorporation of the new methods and data, the ONS now estimates that UK GDP returned to its pre-downturn level in the second quarter of 2013, one quarter earlier than the third-quarter-2013 date previously recorded. The change results from the cumulative effect of the methodological improvements across the intervening series rather than any single new source. While a single quarter's shift in a 13-year-old milestone has no direct policy consequence, it illustrates the scale sensitivity of widely cited economic narratives to measurement: the length of the post-2008 recovery, the depth of the productivity slowdown and international comparisons of recovery speed have all been built on vintages of data that the annual Blue Book process continues to amend, nearly two decades after the events being measured.
The Propenomix Perspective: Thirteen years after the fact, the recovery arrived three months earlier than everyone who lived through it was told. No one's mortgage changes; every economic history of the period is now marginally wrong. And this is the point I want to leave with anyone who trades, invests or writes on the back of a single month's data print: if the settled past can move a quarter in 2026, what do we imagine the provisional present is doing? We watched it live this very week - May's monthly GDP was revised from growth to flat in the space of one release, and the July payroll flash will be revised next month as it always is. My working rule, offered before and repeated because it keeps being right: never make a decision on one month's data, treat every first estimate as a rumour from a usually reliable source, and reserve conviction for trends three prints deep. The professionals who blew up in my two decades around this market share one habit - they believed a number precisely.
Theme 4: What October's Full Edition Means for the Budget
The Summary: Thursday's article is the advance notice; the full Blue Book 2026, including the complete revised quarterly path and the GDP revisions analysis, will be published on 30th October - two days after the Budget on the 28th. The quarterly national accounts on 30th September will carry any additional revisions to data from 2025 onwards. The sequencing matters for the fiscal event: the OBR's Budget forecast will incorporate the revised historical picture, including the rebalanced 2022-to-2024 open period, meaning the output gap, productivity assumptions and nominal GDP base underpinning the public finance projections will all reflect this vintage. The article also notes improved coherence between the national accounts and the public sector finances through reclassification changes, tightening the statistical relationship between the growth data and the borrowing aggregates that frame the Chancellor's headroom arithmetic.
The Propenomix Perspective: Mark the sequence, because it is quietly hilarious: Budget on the 28th, full Blue Book on the 30th. The Chancellor will stand up and build a fiscal edifice on a statistical foundation that gets formally renovated 48 hours later - although in fairness the OBR sees the revisions in advance, so the foundation moves before the speech even if the public documentation follows it. What should we watch? The 2022-24 open period balancing. If the recent past gets revised up, the productivity story softens, nominal GDP rises, and a Chancellor scrabbling for headroom finds a few billion down the back of the statistical sofa; revised down, and the black hole rhetoric writes itself. I would not dream of predicting which - genuinely, revisions are the one series I refuse to forecast - but I would note that Chancellors have historically been luckier with Blue Book vintages than the law of averages suggests. Too cynical? Probably. We will find out on the 30th, two days too late to matter politically, which may be the most honest sentence in this Supplement.
The Week Ahead. A quieter diary at home, a louder one abroad - set your alerts accordingly. The main event is across the pond at Jackson Hole, Thursday to Saturday, with Warsh's first keynote as Fed Chair on Friday morning; the theme says payments, the market will hear only September, and your five-year swap will react either way. Around it, the US second estimate of Q2 GDP lands Wednesday, with the PCE inflation gauge expected Friday. At home, the one to circle is Ofgem's announcement of the October-to-December price cap, expected towards the end of the week - the assessment window caught June's oil spike, and the read-across to Q4 CPI, and therefore to the Bank's 3.2% peak, is direct. Zoopla's house price index is due late in the week, Nationwide's August index follows at the turn of the month, and the Bank's Money and Credit data - July mortgage approvals - arrives on 1st September. Note Monday 31st is the summer bank holiday, so the week compresses into four days and the diary spills into the following Tuesday. And keep half an eye on the war headlines around the cap announcement: the two stories are now the same story, five months apart. Budget countdown: 66 days at publication. Mark it all; I will have done.
As we get towards the end for this week - I can't wait for our next workshop in Manchester on 1st October. JVs and M&A is the topic, and if this week's Supplement has a message, it is that the conditions for exactly this kind of dealmaking are assembling in real time: record numbers of landlords heading for the door, cash buyers naming their price, portfolios changing hands at discounts, and the businesses behind them - sourcing operations, lettings books, whole companies - coming to market with them. Structuring those deals properly, on both sides of the table, is precisely what Rod and I will be working through, with the real-life case studies and the no-holds-barred Q+A as always. The VIP dinner returns too, in the smaller setting where the properly specific questions get answered - the feedback from the last one still makes me smile. Super Early Bird pricing, at more than 20% off, has not got long left around. Book your tickets for Thursday 1st October, Manchester at: https://tinyurl.com/pbwoct26 - powered by Roma Finance, now covering all of your needs from bridging and development to term finance.
Above all - please remember to Keep Calm, ALWAYS listen to or read the Supplement, and Carry On. It has been a week of numbers arguing with each other - inflation up but services easing, surveyors gloomy but landlords cheerful, history itself revised a quarter to the left - and the temptation in weeks like this is to pick the numbers that flatter your position and call it analysis. Resist it. The fundamentals have not changed: there remains a shortage of 2-3 bed terraces and semis, there was one fifteen years ago, and nothing in this week's data - not the 63% demand reading, not the seven per cent yields, not the exodus of the marginal landlord - does anything but deepen it. "Investing in property" is not a guaranteed win, and with the rates conversation still hawkish-with-caution, the discipline around leverage matters as much this month as last. But buying below replacement cost, strong on the spreadsheet, strong on yield, in a market repricing slowly in real terms rather than crashing in nominal ones - that remains as close to a guarantee as this game offers. The houses still aren't being built. The people still need somewhere to live. Everything else, as we established this week, is subject to revision. KCCO!