Reusable quantitative models
Four models covering sovereign debt, housing affordability, inflation passthrough and the industrial metals cycle. Each runs in the browser, exposes every parameter rather than embedding it in a fitted curve, and is designed for repeated use as assumptions change rather than delivered as a fixed result.
Four cycles that determine industrial and property outcomes
Debt sets the cost of capital. Housing converts that cost into transaction volumes and construction activity. Inflation determines the real return on both. The metals cycle is where all three land as physical demand, which is why it leads industrial earnings and lags monetary policy. The four are points on a single cycle rather than separate subjects, and the models are constructed so that an assumption revised in one can be carried consistently into the others.
Debt Sustainability Model
Debt dynamics, the snowball term and the balance that holds the ratio flat. Project a public debt ratio forward from an interest rate, a growth rate and a primary balance, decompose what is moving it, and read off the fiscal adjustment required to stabilise. Includes a rate stress path.
Housing Affordability and Capacity Model
Borrowing capacity, payment burden and the gap between price and what income supports. Turn a price, an income and a mortgage rate into a payment burden, an affordability-implied price and the gap between the two. The rate sensitivity curve prices what each 100 basis points does to buying power.
Energy Passthrough Model
A distributed lag from an input cost shock to headline and core inflation. Push an energy price shock through three channels running on different lags and watch what reaches headline and core inflation quarter by quarter. The second-round setting is what separates a temporary shock from a persistent one.
Industrial Metals Cycle Indicator
A weighted composite that places the metals complex on a cycle clock and tests for supercycle conditions. Six drivers, weighted and direction-corrected, resolve into a level and a momentum reading that place the complex in one of four cycle phases. A separate structural score tests whether an upswing is cyclical or a supercycle.
Applications by role
Each model was developed for client engagements before publication. The applications below reflect the most common uses.
Developers, investors and mortgage analysts
Size the addressable buyer pool at a given price point and test how far price has to fall, or income rise, before the affordability constraint clears. Borrowing capacity is convex in the mortgage rate, so rate risk in a pipeline cannot be applied as a constant.
Top priorities
Macroeconomics and applied policy
Every model is an explicit identity or a stated lag structure rather than a fitted black box, so the arithmetic can be reproduced, criticised and recalibrated. Parameters are exposed as controls instead of buried in an appendix.
Top priorities
Mining, industrials and rates coverage
The cycle phase is the input most often missing from a bottom-up earnings model. Slowdown is where consensus estimates are most reliably too high, because spot prices have not yet confirmed what the forward-looking drivers are already saying.
Top priorities
Planning, pricing and procurement
All four models are built to be run repeatedly against changing assumptions rather than delivered once as a static answer. Copy the assumption set out, keep the vintage, and mark it against outturn when the data arrives.
Top priorities