An economy is never fully seen, and it changes in ways no one has written down.
This is an Open World Digital Twin built with Polytrion. Its goal is to show how the loops of the world economy connect: sixty-one machines, each one a declared piece of the economy, linked by 282 connections, each machine reacting to what the others publish. No correlation between assets is declared anywhere; whatever co-movement appears comes from the connections.
Each machine is a small declared state machine with its own rules. It publishes values and reads the values other machines publish. The twin is the whole network, stepped one week at a time.
Machines and connections as published on fibchannels.com/model/loops, where the full wiring is laid out.
A conventional economic model starts from statistics: assets that move together are declared to move together. This twin starts from the other end. Every machine states what it reads, what it publishes and how it responds, and the 282 connections carry those values between machines. If two assets move together in a run, it is because the connections made them.
That is what makes it an Open World Digital Twin: the structure is written as declared rules that can be read, checked and corrected one machine at a time, and the model is explicit about what it does not yet contain.
The 61 machines are organised into loops that feed one another. They are described here as mechanisms, not as forecasts.
Growth drives employment and prices, the central bank responds, and yields build from real rates, inflation expectations and risk premiums. Higher yields raise the cost of the deficit, which feeds back into a supply premium.
Oil and gas prices weaken the euro through the terms of trade, energy reaches euro-area inflation faster than US inflation, and the central bank’s response partly closes the yield gap.
Yield differentials encourage borrowing in yen; as positions grow, so does the exposure to forced liquidation.
Trade surpluses bring in dollars, which are parked in long-dated US assets, while credit stimulus shapes global commodity demand.
Competing claims on copper, power and capital-market funding.
Fragmentation, energy, fiscal confidence, weather and refining enter as exogenous variables that a scenario can move and a baseline leaves alone. Commodities carry them into prices.
The full picture, with the connections drawn, is on fibchannels.com/model/loops.
One run steps all 61 machines and their roughly 280 connections, week by week. A world-week is one run of the whole network for one week.
| Part | Runs × weeks | World-weeks |
|---|---|---|
| Ensemble | 9 scenarios × 400 runs × 26 weeks, plus a first stopping pass | ~150,000 |
| Sequencing study | 4 orderings × 250 runs × 104 weeks | ~100,000 |
| Correlation study | 5 paths × 250 runs × 104 weeks | ~130,000 |
| About 40 studies | typically 4–9 arms × 60 runs × 26–104 weeks | ~600,000 |
| Total | ~1 million |
Simulated time, as reported by the engine. Figures are approximate.
The model captures the monetary and capital-account loops. It does not yet contain China’s domestic economy or trade flows between regions. Recent additions bring Chinese export prices into US and euro-area core inflation, and let European household saving respond to local conditions, with excess saving flowing to US assets. A twin that lists its own gaps can be corrected where it is wrong.