For someone watching a mortgage book
The policy rate is the number everyone quotes. It is not where you cut.
When rates rise, the commentary follows the print. A portfolio manager still has to decide which balances to watch. The rate reaches a household only if that loan’s coupon can reprice. The monthly payment goes up. What is left after essentials and the mortgage — the buffer — is the cash that can absorb the hit.
Fixed-rate borrowers do not feel this hike in their payment. Floating-rate borrowers do. Two households with the same coupon shock can end in different places, because one started with room and the other did not.
The picture that follows is a modelled book of 2,000 mortgages, built so the path can be replayed. It is a way to see where pressure concentrates. It is not a forecast, not a census of the euro area, and not credit advice. Euro-area figures come after the model, as a check on whether the direction is real.
- Buffer
- Monthly income left after essentials and the mortgage payment.
- Floating
- A coupon that can reprice when market rates move. A fixed coupon stays where it is.
- Thin
- Residual income under 6% of income. A teaching line for this story, not a regulatory definition.
A portfolio decision
Where do you cut when rates rise?
Scroll follows one floating mortgage, then every loan in the book. The cut is the group whose payment can still rise and whose cash is already thin.
Where you cut
Not the overnight rate.
The overnight rate is common to the whole book, so it cannot tell you where to look. 79.0% of balances sit outside the sleeve. The cut starts with the 21.0% that are floating and already thin.
From there the work is ordinary credit work: when those coupons reset, what the buffer is after the new payment, and which names were already in the weakest third before the hike.
€672left this month
Your turn
Move the shock. Change who can reprice.
The film held the hike at 300 basis points and the floating share at 35%. Both are assumptions. A smaller hike, or a book with fewer loans that can reprice, changes how much balance ends up in the sleeve.
A 300 bp shock on a book that is 35% floating puts 21.0% of unpaid balance in the sleeve.
- Thin share
- 27.8% → 33.3%
- Sleeve
- 21.0%
Still the same modelled book: seed 42, 2,000 loans. Dragging recomputes every payment. Raising the floating share assigns floating status to the largest balances until that share is filled, which is why the sleeve grows faster than a random mix would. That assignment is a model choice, not how euro-area mortgages are actually distributed. Thin still means residual income under 6%.
Violet in the thin zone is the watchlist
Reality check
The average improved.The tail did not.
The book you just watched is illustrative. These figures are not. They are published euro-area numbers from the same hiking cycle, and they do not all point the same way.
If you managed the portfolio off the average, the household sector looked safer: debt fell relative to income. If you looked at mortgagors under payment stress, the stressed share rose. Housing costs for people with mortgages also outran general prices.
92.8→87.0
Percent of income. The euro-area household sector delevered.
ECB sector accounts, 2022 Q4–2023 Q426→33
Share of mortgagors with debt service above 40% of income. A published simulation — the same direction as this book, not the same metric.
ECB WP 3053Observed · ECB Consumer Expectations Survey, Jul 2022–Jan 2024
- Mortgagor housing costs≈12%
- Housing costs, all households+10.2%
- HICP prices+5.5%
Housing costs for people with mortgages outran general prices. Both figures are from the CES focus box.
Read the two big figures as a disagreement, not as a tie-break. Debt to income is an observed sector average. The 26 to 33 is a published simulation of borrowers whose debt service exceeds 40% of income. It is not this book’s 6% buffer, and it is not loan-level European data.
The rhyme is the direction. In the modelled book, thin residual income moves 27.8% → 33.3% of balances, the same way the published stressed share rose. The decision quantity is narrower than either average: 21.0% is floating and already thin. A manager watching only the debt-to-income ratio would have called this period a relief.
The decision
Size the watchlist from the sleeve, not the policy rate.
In this modelled book the sleeve is 21.0% of unpaid principal: floating-rate balances with less than 6% of income left after essentials and the mortgage. The other 79.0% either cannot reprice, or still has more room than that line.
Cutting on the overnight rate treats every loan as the same exposure. Cutting on “thin” alone includes fixed-rate names whose payment will not move because of this hike. The sleeve is the intersection: the coupon can still change, and the household is already close to the wall.
What to watch inside that sleeve
- 01Reset dates. A floating loan that does not reprice for years is not the same problem as one that resets this quarter.
- 02The buffer after the next payment change. The policy print is upstream. The residual is the quantity that got smaller.
- 03Names that were already short of room. 90.4% of the accounts that became thin were already in the weakest third of buffers before the shock. The hike mostly found households that had little slack.
Modelled example, seed 42, 2,000 loans. Not a real portfolio and not credit advice.
What this picture is
The cloud is an illustrative mortgage book: 2,000 amortising loans, seed 42, about 35% floating by unpaid balance. Floating status is given to the largest balances until that share is met. A +300 basis point shock reprices floating coupons only. Thin means residual income — after essentials and the mortgage — under 6% of income. That cutoff was chosen so the thin share rhymes with ECB Working Paper 3053’s rise in borrowers above 40% debt service (26% to 33%). It is not that metric, and it is not a regulatory definition.
Scroll blends the calm snapshot into the shocked snapshot. The sliders recompute each payment from the amortising formula. Loan 724 is one floating name in that book: income €4,792, buffer €672 before the shock and €65 after.
Observed figures are the ECB sector accounts and the Consumer Expectations Survey. They are not outputs of this book. Nothing here is loan-level European data.