From View From The Top, 21 October 2026
Everyone is getting the same data, the same tools and moving towards the same cost structure. Which means none of them is has the edge anymore. The final part of this series is about what is left — and who ends up on the other side. In other words, who wins?
Last week I sat through a presentation meeting being given by one of the big data providers. Comparables, rents, yields, planning pipelines, ownership, transaction history, heat maps — everything, instantly, beautifully presented. The market for data has improved significantly from only a few years ago and it’s awesome. Genuinely impressive. At first I was excited to see the possibilities as we are building in many of these tools and platforms ourselves. They are very good at what they do. But I keep coming away with an uncomfortable feeling which I have been slowly writing about. What’s the impact of all this?
What Strikes Me
Every person in our industry is now downloading, exporting, and being handed exactly the same information. And each of us will walk out and use it to compete with each other, on the same assets, with the same arithmetic, arriving at roughly the same number. And just as quickly.
Whether its Wall Street or real estate today, twenty years ago that information was the business. Knowing a street, knowing which landlord was tired, knowing what the last building traded for because you were in the room when it did — that was the edge, local knowledge built over time and it was worth a great deal of money. It is now a subscription. Paid for. We all have access. Some will have it quicker than others, but we will all have it soon.
This View series has spent four issues on the machinery underneath a lending platform: what it costs, what becomes software, what cannot, and what that does to the people. This issue is the one readers have been writing in about since Part 1. Fine — so who wins? Let’s break it down.
THE THREE THINGS THAT STOP BEING THE “EDGE”
Let’s start with what will NOT separate anybody, because this is where most of the industry is still looking.
1. Information. Covered above. If you can buy it with a licence fee, so can the guy bidding against you. The only information edge left is the kind that was never written down — I’ll come back to that later.
2. Cost. Part 2 argued that sixty to seventy per cent of a lending platform’s cost is information work and that now it is about to become software. I believe that. But notice what happens next- everyone gets the same reduction. Within three or four years the low-cost base is not an advantage, it is the price of being allowed to compete for the same thing. The firms moving now buy themselves a few good years, but they do not buy a moat. And if you think you have propriety data today, my guess is in a few years that won’t be the case. Think Bloomberg.
3. Speed. Same story. Faster credit decisions are a wonderful thing to sell for about eighteen months, after which a borrower simply expects them and the firm that cannot deliver them is not in the conversation.
Which produces a very uncomfortable sentence, and I would rather write it plainly than try and dress it up:
If your advantage can be purchased in the market, it is not an advantage. It is an upgrade, and everyone is upgrading.
WHAT CANNOT BE BOUGHT
Three things, and only three that I can think of:
1. The relationships that originate. The broker who calls you first because of fifteen years rather than because of your interface. No software produces that and no amount of capital accelerates it. Relationships are Alpha. I have written before that the most expensive part of this business stays analogue; it is also the part that generates every pound of revenue.
2. Judgement at the edges. Part 3’s territory. The month-fourteen call when things go wrong. The sponsor you back or do not back. The moment at step three when the machine’s assumption is wrong and someone has to notice before it is carried confidently to the end. What does the renegotiation or work out entail. That judgement is becoming scarcer at exactly the moment it becomes more valuable — which was Part 4’s problem, and nobody has solved it.
3. Access to what nobody else can afford to serve. This is the one I keep returning to, and I think it is the largest-
Britain is short roughly the same number of homes as America while carrying a fifth of the population. A 5x difference. The schemes where most of those missing homes live — the £3m to £30m developers — are served in the United States by a purpose-built mid-market lending industry that runs like a machine and in the UK by almost nobody. Not because the deal is bad. Because it costs nearly as much to underwrite the small loan as the large one, and so the large one gets priority.
THE CONCLUSION I DIDN’T EXPECT TO REACH
When I started this series I assumed the answer to “who wins” would be a better lender. A sharper, faster more streamlined credit process, a cleverer structure, a lower cost of funds (yes some of this is true).
But working through it every night for a very long time I have changed my mind, and I will say what I think I’ve determined:
The lending itself is becoming plumbing
Necessary, skilled, unglamorous plumbing — but plumbing. Or maybe a more financially elegant word, “a clearing system.”
When the cost of originating, underwriting, documenting and monitoring collapses toward a common floor for everybody involved, the act of writing the loan stops being where the value sits. It becomes the pipework between a pool of capital and a scheme that needs it.
Which means the winners are not the best lenders with 30 years experience. They are the ones who own something at the other end of the metal pipe-
The relationship that brings the deal in. The judgement that prices what the model cannot see. The position in a part of the market that nobody else can reach economically (think Wal Mart). Possibly a share of the thing being built rather than only a coupon on it (think partnerships).
A platform that owns none of those and competes purely on the quality of its lending is competing on the one input that is about to be available to everyone at the same price.
The generational wall of money heading into private credit — Apollo’s phrase, not mine — has to pick its platforms, and I wrote in the spring about what capital is picking: evidence over marketing, operators over AUM. The characteristics below aren’t just how you win borrowers. They’re how you get chosen by the capital. The bigger the AUM doesn’t mean you’re solid these days. In fact, today it simply may mean you have more problems right now.
MORE SPECIFICALLY
There are four characteristics I would expect the firms that take on the next decade in this market to have most of them.
They rebuilt the machine rather than bolting software onto the old shape. Every org chart in this industry is a fossil of what the work used to require. Adding agents to a structure designed for manual processing produces a faster version of the wrong thing. The ones who win drew it from scratch. Rewrite the work flows.
They went down-market on purpose. Everyone chases the £50m+ ticket because the economics worked that way for thirty years. When the cost per deal falls far enough, the fifteen-unit developer becomes worth serving — and that is where the volume, and the national need, actually is. 4,000 middle-market developers all looking for capital.
They kept building judgement while everyone else cut it. Part 4’s argument. The firms that treat the saving purely as a cost reduction will discover in a decade that they removed the apprenticeship and kept the org chart. The ones that deliberately rebuild the career ladder — juniors in the room for the difficult calls, not just the clean approvals — will own the only input that stayed scarce.
They own more than the loan. Equity alongside the debt. The operator relationship. The platform the borrower keeps coming back to. Something with a claim on the upside rather than a fixed return on somebody else’s. This one is more controversial I suppose.
And the honest counterpoint: most firms will do this slowly esp if they are well established and steaming ahead. Most will take the cost saving, run the same book with fewer people, enjoy two or three comfortable years, and then wonder in 2031 why the spread went and the volume never appeared.
That is not a prediction about an industry. It is what I watched happen to another one.
HARD LANDING
I have now spent five issues now describing a transition I am also sitting inside, and it would be dishonest to finish without saying the obvious thing.
Writing the thesis is the easy part. Talk is cheap
I tried to build a version of agentic agents three years ago. I knew exactly what I wanted it to do, the models could not do it, and we spent real money finding that out. Too early is the same as wrong, only more expensive — I wrote that in Part 2 and I meant it about myself. My timing was wrong on execution.
I have also watched balance sheets I thought were permanent disappear inside five years. Size is not safety. It’s just more messy when it goes wrong. And I predict by the end of next year we will see a few big names exit the private equity world. Nobody in this industry is as clever as their assets under management suggest, and I include everyone I have ever raised money from in that.
So I am not going to end a series about who wins by implying I know that my own platform is one of them. What I know is which questions are the right ones, because I have been on the wrong side of this transition once before, standing on a trading floor watching an entire cost structure evaporate around people who thought the floor was permanent and being inside big machines who’s names only a few remember. I’ve also seen big names come and go. The market disregards hubris.
The floor was the cost. That was the whole lesson, and it took most of us fifteen years to learn it.
When I started years ago on Wall Street real arbitrage existed and information was key. When I built grocery anchored shopping centers in post Soviet countries there were none, but the demand was there lurking. I built the demand. When I built the first five-star hotels in CEE, there were none and everyone said there was no demand. I built the demand. When we launched an organic salad concept in London years ago that sector didn’t exist. In each case we didn’t invent anything new, we just adapted to the markets latent. And that’s what we need to do now and it’s easier said than done.
WHAT I WOULD LIKE FROM YOU
This series has produced more correspondence than anything I have written, and the best of it has come from people who disagreed with me.
So two things I’d like to ask-
If you run a platform and you have done the cost work — cost per deal, cost per pound lent, loans per head — I would still like to see it, even roughly and in confidence. Nobody publishes these, which is precisely why nobody knows whether their own is any good. I’d love to hear from you.
Also I am going to start convening some small, closed conversations on this — I’m thinking eight or ten people, no slides, nothing for sale, nobody being quoted. So if you want to brain storm and figure things out and where the opportunities are reply to me on LinkedIn — and say so. I will come back to you with some dates.
THE MACHINE UNDERNEATH · THE FULL SERIES
Part 1, The Sandwich Problem · Part 2, The Cost Structure Will Fall · Part 3, What the Machine Can’t Do · Part 4, The Human Cost · Part 5, Who Wins.
SUBSCRIBE
If you would like to find out more about this topic, or get this newsletter each week and all the back issues please subscribe. Or if you know someone who would argue with this please share – pauloberschneider.com
