From View From The Top, Issue #37
The August Stocktake
Grading eight months of opinions — then closing the laptop until September
I’ve always held August as sacrosanct.
Most of you know that I spent the majority of my career, thirty years to be exact, building business verticals across six countries, and this one month has stayed protected through all of it. To pause. To look at the water. To swim, eat well, and let the year catch up with me. just to breath and take stock.
And I found that the best decisions of my career didn’t happen at a desk. They happened on the beach or my ranch; most often they happened somewhere around the third week of August, unbidden, while I was thinking about nothing at all. I learned a long time ago not to fight that.
So I think this will be the the last issue until September. I’m heading off and need a break. But before I go, there’s housekeeping I owe you.
I’ve now written thirty-six issues of this newsletter. Eight months of calls and positions taken in public. I’ve shared some stories and experiences and have made a few calls some controversial. But anyone can publish opinions weekly. The discipline is going back and checking them.
That’s what we demand from a loan book. So I think it’s fair you demand it from me.
So — lets look behind the curtain; the stocktake.
WHAT HELD UP
1. The institutional exit from rental.
I wrote in Issue 29 that funds citing depreciation for leaving UK residential were mostly describing something else: they overpaid at peak and their cost of capital moved against them. We funded at the top being optimistic. Depreciation was the language. But overpayment was the problem. Eight months later the exits continue, the portfolios keep coming to market, and the buyers turning up are exactly who I said they’d be — operational owners at reset prices, not the next round of financial engineering.
2. The valuation spiral.
Same issue: the KPMG effect pushing auditors to conservatism, banks re-running numbers, covenant pressure building quietly. That’s only intensified. And here’s the part I didn’t fully anticipate — it turned into an advantage. A 2026 Red Book valuation is the most honest number this market has produced in a decade. Painful as that is. I now build every LTV on it gladly. But it’s also created more regulation and scrutiny over lenders introducing Annex 1 regulations which now everyone needs to comply with.
3. The capital coming back changed.
I told you in the spring the constraint was moving from capital to process, and more recently that the returning money looks nothing like the money that left. Both confirmed at my own desk, in real time. The back leverage conversations happened as described. And the allocator questions got harder — several taken almost word for word from the due diligence checklist I published in Issue 33. When your own homework gets used to examine you, the market is working. I passed, incidentally.
4. And 1947.
The oldest call in the catalogue — that Britain’s housing shortage is a deliberate seventy-eight-year-old political settlement, not a market failure — has taken no damage at all. It won’t. That’s what structural arguments are for.
WHAT DIDN’T WORK OUT — OR HASN’T YET
A scorecard with no misses is marketing. Here’s the other column.
I got the rate story wrong before I got it right.
Issue 25 was built around rate cuts reigniting demand — didn’t happen. And I had to reframe it before posting. The structural point survived: no rate cycle fixes a supply-constrained market. But I’d reached for a cyclical hook that wasn’t there. Thirty years in and I still occasionally write the argument before checking the weather. Noted.
I wrote for an audience I hadn’t built.
One Monday post this year — technology and data disrupting real estate — did 285 impressions. It was my worst number in two years. The argument was right. I believe it more now than when I wrote it. But I served it to an audience of UK credit and property people I hadn’t yet given a reason to care. Maybe people aren’t paying attention to what’s happening, maybe they don’t care or don’t want to acknowledge the shift taking place. You earn the right to a new argument one issue at a time. But I feel this is a big one in the making. It’s just that it has been covered as much as I thought it would be. Keep watching because agentic AI and tools is about the be a game changer.
The vintage call is conviction, not evidence.
I said the loans being written now may be the best of the cycle. I believe it as strongly as anything I’ve published — and it can’t be verified until 2028, when this vintage reports. So it sits in the unproven column, and I’d rather put it there myself than have someone else do it for me. Mark your own positions. It’s the entire theme of this newsletter.
And I promised a deal walkthrough for six straight issues before delivering something else.
The teasers ran ahead of my readiness. What finally arrived — the workout case study — was better than what I’d promised and became the most-forwarded thing we’ve published. But the lesson stands. Don’t tease what isn’t written. The market forgives changed plans; it notices serial promising.
WHAT I’VE CHANGED MY MIND ABOUT
Volume.
I started the year believing more posts meant more reach. The data spent six months politely disagreeing. I now publish less than I did in the spring and everything performs better. Same rule as lending: origination volume is vanity. Quality is survival.
The personal writing.
I treated the personal material — the stories behind the building — as a side dish to the market commentary. The numbers reversed that verdict, and so did the messages that reached me privately. The piece I wrote in June about the empty room travelled further than any market call I’ve ever made. I’ve stopped being surprised. You can get market analysis in a hundred places. There are very few places to hear what the building actually costs. There will be more of this from me — in a form I’ll say more about after the summer. I think my experience is broadly more interesting when told from the drivers seat.
And technology — from observation to conviction.
This one deserves its own section. Because it’s most of what I’ll be writing about this autumn.
WHERE I’M POINTING THIS AUTUMN
When I come back in September, this newsletter is going to concentrate on one big question, approached from several directions:
How does this industry actually improve its delivery?
I’ve spent eight months documenting the problems — the fee stack that consumes 3-4% before the asset does anything, the process friction that turns four-week closes into twelve, the fragmentation that makes European real estate run on spreadsheets and reconciliation. I’ve lived the cost of it through a workout where the hardest enemy wasn’t the market. It was stale information moving slowly between five parties.
The autumn is about the answer. And the answer, increasingly, has a name the industry is only starting to take seriously: agentic AI.
Not chatbots. Not copilots drafting emails. Systems that execute work — underwriting analysis, data reconciliation, monitoring, reporting, covenant tracking — end to end, at machine speed, at a fraction of today’s cost.
Here’s why this matters to my world specifically. Investment management is a cost-structure business wearing an asset-management costume. The 2-and-20, the fund administration, the armies of analysts rebuilding the same spreadsheet — all of it exists because the work has historically required expensive people doing slow things. When agentic systems take on that work, the cost structure of an investment firm doesn’t improve by 10%. It changes shape entirely.
And in a market where — as I argued in the vintage issues — gross returns are set by discipline and entry price, the firms that restructure their costs first will deliver net returns the others simply cannot match. Same assets. Same market. Different machine underneath.
That’s the thread for autumn: what it means for lenders, for fund managers, for the fee stack, for the people — and for who wins the next decade of this industry. Regular readers will have noticed I keep saying “more on this soon.”
Soon starts in September when I’m back at my desk.
SEPTEMBER
The runway, because the autumn won’t be quiet.
On 9-10 September I’m in Paris, co-chairing the Build-to-Rent discussion at Europe GRI — can the model still deliver across regions? If you are there maybe I’ll run into you. The week before, this newsletter publishes my full answer, so the room can argue back with the text in hand. The week after, I’ll tell you what the room actually said.
Beyond Paris, the autumn holds more than I can currently write about. Regular readers will have noticed the breadcrumbs. I’ll leave them unswept for now.
HARD LANDING
The stocktake is the easy part. Any lender can tally a scorecard in a quiet month.
The harder question is the one August exists for. Not “what did I get right?” but “am I still pointed at the right things?”
I’ll be asking it near the water, with my birthday somewhere in the middle, the way I have every August for thirty years. Some years the answer comes back: keep going, same direction. Some years it comes back different.
This year, I suspect, the answer is already waiting for me. It’s been forming all through this compressed, back-and-forth summer — through the workouts, the vintage arguments, the conversations with capital that looks nothing like the old capital, and a technology thesis that stopped being a theory and started being a plan.
But I’m going to let August say it in its own time.
See you in September. It’s going to be a big one.
BACK IN SEPTEMBER
View From The Top returns the first week of September — the Build-to-Rent question in full, ahead of Europe GRI in Paris. Then: the autumn series on delivery, cost structure, and what agentic AI does to investment management.
