Calibrated models for decisions where capital, physics, regulation and competing interests collide.
Expose the assumption. Calibrate the model. See where the decision turns.
A disputed decision goes in. A calibrated decision model comes out.
Assumptions are stated. Constraints are explicit. Competing positions run through the same model. The result shows what changes the decision, and what does not.
The mechanism behind the decision.
The numbers that make it real.
What follows under each defensible assumption.
Parent-bank exposure, capital requirements and resilience in a crisis.
Europe certifies carbon removal without the energy it costs. A tonne on the ledger and a tonne in the air are no longer the same object, and the rule made them different before anyone optimised against it.
Over two billion dollars, private and public, went into direct air capture. The floor was published after. Which of your commitments still has one to check?
Every roadmap promises a learning curve. Physics sets where it ends. We compute the floor, and the headroom left above it.
Pressure- and temperature-swing units, designed from the cycle outward.
Point-source CO₂ with commodity sorbents. The gain is in the cycle, not the material.
How close a running plant sits to its nearest limit, and which limit that is. Unitless, so one reading holds across units and sites.
Developed in Einsiedeln SZ
Per cent of the exposure. Implied loss: the crisis loss at which each proposal would be the balancing share, at equal weights (working paper, Table 1).
Proposed CET1 backing ratios for UBS AG's foreign participations range from 50 % to 100 %. Floor1200 puts them into one model and asks the question the percentages alone cannot answer: what loss is the rule meant to withstand?
Calibrated backing at the Credit Suisse loss of 60 % reported by the Swiss Federal Department of Finance, depending on the weighting of taxpayer protection against capital capacity; USD 31.8–34.0 bn, against 15.8 bn held today. Above 92 % the parent breaches its requirement before any crisis.
Read the 5-page paper →Assumptions, calculations, sources and limits stay visible.
Four proposals. One balance sheet. The missing variable is the loss assumption.
Read →The energy per tonne is set by the dilution of the feed, not by the number of plants built.
Read →At an antagonistic corner every smooth scalar indicator hides one margin, and the geometry chooses which.
Read →Where no direction improves every binding margin at once, the geometry decides which margin binds next.
Read →If the problem has competing numbers, measurable constraints and a consequence worth getting right, send it.
A finished specification is not required. A disputed assumption, a proposed number or a model that no longer convinces you is enough to start.
info@floor1200.ch