From View From The Top, Issue #40
What I Heard in Paris
Are we asking the wrong question?
Last week I was at the InterContinental Paris Le Grand for Europe GRI — two days of closed-door discussions with the people who own, build, fund, and lend against European residential. I co-chaired the Build-to-Rent roundtable, the session I previewed in this newsletter two weeks ago.
GRI runs on Chatham House terms, so no names. But the substance travels fine without them. Here’s what I heard — and where I think the room was right, wrong, and most interestingly, ahead of itself.
AI OPENED THE CONFERENCE. THAT ITSELF IS THE NEWS
Before a word about housing: look at the agenda.
The discussion that kicked off the event — the C-suite session, the one everybody attends — was about AI. What it does to our businesses. Our costs. Our teams.
Two years ago, this conversation happened in the corridor, if it happened at all. A year ago it was a breakout for the curious. Now it opens the conference — ahead of rates, ahead of capital flows, ahead of the housing crisis itself.
The room broadly agreed on direction: this changes cost structures, and it changes them soon. Where it split was on depth. Most of the room is still in the first sandwich — deploying AI that drafts, summarises, answers. The distinction I published last week, between instruction and execution, hasn’t fully landed yet. A few operators are further along; you could tell who by the specificity of their questions. They weren’t asking whether. They were asking how to govern it — which, as regular readers know, is exactly the right question. The mistake at step three doesn’t care about your enthusiasm.
My read: the industry has agreed AI matters roughly eighteen months before it has understood what kind of AI matters. That gap is the opportunity. The Machine Underneath series continues next week accordingly.
THE BTR ROOM: CAUTIOUSLY POSITIVE, HEAVILY CAVEATED
Now the roundtable. The question on the card: can Build-to-Rent still deliver across different regions?
The room’s answer, fairly summarised: yes, cautiously — but only if we rethink how housing is treated.
The obstacles came fast and were familiar to any reader of this newsletter: high land values, construction costs stubbornly elevated, slow and uncertain planning systems, rising regulation, expensive debt. The viability equation for new development is getting harder to solve, not easier — across essentially every European market represented in the room. Nobody disputed the demand side. Nobody serious ever does.
So far, so consistent with what I published before the summer: two models wearing one hat, the financial version finished, the operational version selective.
But then the discussion took a turn — and it’s the reason this issue exists.
THE QUESTION UNDERNEATH THE QUESTION
Someone picked up the two-models argument and pushed it somewhere sharper. Maybe, they suggested, we’re all asking the wrong question entirely.
Not: can the model deliver?
But: is the capital structure right?
Should housing ever have been underwritten like opportunistic private equity real estate — short holds, high target IRRs, exit-driven business plans, value measured at the sale?
Or is housing essential infrastructure — long-duration, income-driven, funded by patient capital, inside a regulatory and planning framework aligned with the need to increase supply?
Sit with the arithmetic, because the room kept circling it. Take today’s stack — the land values, the build costs, the planning drag, the regulation, the debt — and ask it to produce private-equity returns. The equation doesn’t solve. It was never going to. That’s not a market failure; it’s a mismatch between the asset and the return expectation strapped to it.
Ask the same stack to produce infrastructure returns — and a great deal more of Europe’s housing pencils.
This reframing matters because it explains the last decade better than any other single idea. The 2015-2021 institutional wave didn’t fail because rental housing is a bad business. It failed because opportunistic capital bought an infrastructure asset and demanded opportunistic returns from it. The exit wave I’ve written about all year — the depreciation-that-means-overpayment funds — is that mismatch unwinding. The capital wasn’t wrong to come. It was wrong about what it was buying.
And it connects directly to last week’s issue, because the viability crisis has two jaws. One is the cost of the machinery — the process friction agentic AI is about to compress. The other is the return expectation itself. Attack both — cheaper machinery, honest return framing — and the set of schemes that pencil expands dramatically. Attack neither, and we keep holding conferences about why nothing gets built.
THE REGULATION POINT — AND ITS PRICE
One more thread from the week worth pulling: regulation, and its cumulative cost.
The timing was pointed — the draft regulations for England’s new PRS Database published while the conference was running. The constructive read, which I share: better standards, transparency, and professionalisation strengthen confidence in the sector, and institutional landlords should welcome being distinguishable from the worst of the private market.
But the room’s caveat was right too. Regulation has a cumulative cost, and it lands on the same viability equation as everything else. Every layer — building safety, energy standards, licensing, database compliance — is individually defensible and collectively expensive. If we want institutional capital to fund the homes Europe urgently needs, policy has to support investment and delivery with the same energy it regulates the finished product.
The infrastructure framing actually resolves this tension. You regulate infrastructure heavily and you support its funding deliberately — that’s the deal. What doesn’t work is regulating like infrastructure while expecting the capital to price like private equity. That’s the current settlement, and it’s the worst of both.
WHAT I TOOK HOME
Three conclusions, on the train back.
First, the argument I published before Paris survived contact with the room — but the room improved it. “Two models wearing one hat” describes the operators. “Is the capital structure right?” describes the money. The full picture needs both: operational ownership, funded by capital that knows it’s buying infrastructure.
Second, the capital shift I’ve been documenting all year now has a destination. The money returning to European residential — the patient family offices, the American institutions wanting operators, the back leverage buying discipline — is, whether it uses the word or not, infrastructure capital. The repricing isn’t just of assets. It’s of the asset class’s identity.
And third: the AI session and the BTR session were the same conversation, held in different rooms. One was about cutting the cost of the machinery. The other was about correcting the returns demanded of the asset. Both jaws of the same viability problem — and the platforms that work both at once will define the next decade of this market.
That, as it happens, is where this newsletter is heading all autumn.
HARD LANDING
Forty years of conferences teaches you to listen for one thing: the moment a room stops defending its old question and starts asking a better one.
Paris did that. “Can BTR deliver?” is the old question — it assumes the model and interrogates the market. “Is the capital structure right?” interrogates ourselves. What did we think we were buying? What did we demand it produce? And was the failure in the asset — or in the expectation?
Housing didn’t fail Europe’s investors.
The underwriting assumption did.
The room is starting to say it out loud. That’s how repricings begin — not on screens, but in sentences.
