Chapter 7 of my book opens with an internal bank memo. The memo does not exist. I constructed it, and the book says so plainly. But every figure underneath it is real, drawn from PwC, ONS, and Bank of England public reporting, and that is the point of the exercise. The scary part is not invented. Only the paperwork is.
It reads like this:
STRICTLY CONFIDENTIAL. DRAFT. NOT FOR EXTERNAL DISTRIBUTION.
Subject: Emerging Correlated Default Risk in Professional-Class Mortgage Portfolio, South East England
Recent monitoring of income verification data across our professional-class segment has identified patterns that fall outside the parameters of our standard stress testing framework.
I will give you one finding from it, because one is enough. Preliminary modelling in the memo puts professional-class mortgage exposure at £540 billion to £750 billion. Hold that number next to the current reality: total UK mortgage balances in arrears stand at £20.6 billion, about 1.2 per cent of the book, and that figure has been reassuringly stable for years. The gap between those two numbers is the blind spot.
Now, the banks stress test for a lot of things. The Bank of England’s stress test models deep recessions, falling asset prices, higher interest rates, and all seven major UK lenders passed the most recent one. But those scenarios assume defaults arrive one household at a time. Individual borrowers experiencing individual hardship, spread nicely across the country. What the scenarios do not model, or not properly, I think, is what happens when financial services, consulting, legal, and technology roles in the South East all shed income at the same time, for the same reason, in the same postcodes where the £400,000-plus mortgages cluster.
That is not a recession pattern. It is a correlation pattern, and correlation is the thing mortgage models are worst at.
I spend my weeks advising the kinds of firms that employ these borrowers, and I build the tools doing some of the automating, so I am not writing this from a safe distance. The hiring freezes and the quiet headcount reductions are not hypothetical to me. They are my Tuesday meetings.
The memo is only the first few pages of the chapter. What follows it is the part I actually care about: a translation of what the memo’s language means at a kitchen table, because “emerging correlated default risk” is a phrase designed to make a frightening thing sound manageable, and one family’s version of the whole event, month by month, decision by decision, down to the line items they cut first. And then why the 2008 playbook, forbearance and deferral and waiting for incomes to come back, is the wrong playbook when the incomes are not coming back.
I will not reproduce that here. It took a chapter to do properly and it is the closest thing in the book to the reason I wrote it.
If you have a large mortgage justified by a professional salary, the chapter is about you. I mean that kindly, but I do mean it.
This is from Chapter 7 of The Next Rung, my book on what AI displacement actually looks like from inside. It’s open for pre-order now: