A multi-agent simulation of the financial world. AI agents represent different market participants, form competing views, react to macro shocks, and interact across assets—revealing how market dynamics may unfold and where the best trade opportunities emerge.
Use cases
Simulates the different market participants in each market — and the transmission mechanics between them — to preview how an event propagates before it trades.
Helps traders surface what the market may have wrong — potential mispricings, and the overlooked lines of reasoning behind them. For policymakers it works as a think tank: the same evidence base and simulation, pointed at a policy question.
What is actually driving markets right now — assembled from the evidence and the causal graph, not limited to headlines.
Ask across the whole evidence base at once — filings, flows, positioning, macro vintages — and get an answer with its reasoning and citations shown.
Why an agent
A vast body of official macro data and private indicators, read continuously, to one standard — every reading sourced and filed into one ledger. Human coverage has to choose; the agent does not.
An independent agent takes the seat of every market participant — reasoning from that participant's own objectives, its constraints and the mechanics of how it trades — and simulates the market's reaction before any position exists.
Humans — and single AI models — are unreliable narrators of their own assumptions and beliefs; the answers come out distorted. The agent's discipline is structural: every thought is recorded, every claim names what it rests on, and a second model family is tasked with attacking the conclusion.
contact@globalmacroagent.com
The engine
It first reconstructs what the market believes — the consensus narrative, what is already priced, and who is positioned where.
Then it derives its own view from primary data and facts alone — vintage-correct evidence, code-computed measurements — deliberately apart from the consensus it just mapped.
Finally it simulates the different market participants and the transmission between them, to predict how the market digests the gap between consensus and evidence.
Coverage
The method is breadth without selection: a vast body of official macro data and private indicators, read continuously to one standard — every reading sourced, time-stamped, and filed into a single evidence ledger. A human desk has to choose what to read; the agent does not.
The model
The engine follows how information transmits across families of evidence, and it keeps facts strictly apart from assumptions. The analysis is driven by the relationships between the data, not the data itself — several coherent worlds held at once, weighted and retired by evidence.
Simulation
Each market is argued by independent agents, one standing in the seat of each participant class — reasoning from that participant’s own objectives, its constraints, and the mechanics of how it participates — to simulate what it does next. A central adjudicator then weighs the long and short forces, and their weight in the market, into a single ruling on who sets the price.
Honesty
Every statement lives in a three-part ledger — fact, inference, assumption — and a fact must carry its source. Claims declare what they rest on, so when an upstream fact is overturned, the contamination is computed and everything built on it is revoked, in cascade. Tensions close only by ruling; “I don’t know” is a legal answer.
Process
Every round compresses a vast body of analysis, layer by layer, into a handful of conclusions — and every conclusion must meet the market that follows: each states, in advance, what would prove it wrong, and the forward market does the grading.
The full breadth of data and indicators becomes a ledger of claims — every one a sourced fact, or an inference that names what it rests on. Facts and assumptions never mix.
Bull argues bear across separately trained model families, and every simulated participant states what it would do from its own seat. Claims that cannot survive the argument die here.
The central adjudicator weighs the long and short forces, and their weight in the market, into one ruling per market. A contested ruling shrinks the size it is allowed to carry.
What remains is a handful of volatility-budgeted positions, each shipped with the condition that would prove it wrong. The forward market does the grading — no falsifier, no trade.
Philosophy
The agent’s reasoning distils the philosophy of the investors who defined global macro — Soros’s reflexivity, Bridgewater’s radical truth-seeking. A market view acts on the world, and the changed world feeds back into the view; the discipline is to reason inside that loop without fooling itself.
Limits
The same ledger that records the agent’s views records its boundaries.
run 33993 · 2026-09-11
It starts with what the market believes: the narratives moving markets right now, each held as a competing hypothesis and scored for strength and crowding — none forced to converge; the data retires them.
Conflict risk around the shipping chokepoints lifts crude, diesel and freight; the pass-through keeps headline inflation above 3% even as core softens, and the futures strip reprices the policy path upward.
A contested Board seat and a leadership change are read as politicising policy, so investors demand extra compensation for holding long-dated Treasuries and keep inflation compensation above target-consistent levels.
A widening deficit and a heavy coupon calendar meet foreign and end-investor absorption at the long end; whether that bid holds is what decides if supply pressure turns into a term-premium repricing.
The energy impulse decays while the labour market cools, so the priced hiking cycle crests near its current level and rolls over rather than extending.
in the headlines this round: central banks signal further rate rises as inflation picks up · global bond sell-off deepens then stabilises as oil jumps · diesel and fuel prices hit record highs in the US and Europe · the August US jobs report lands · stocks slip as the oil rally fuels inflation jitters
Adversarial cross-examination
The agent cross-examines its own causal graph against the market’s narratives, tension by tension. What survives is a short list of potential mispricings and overlooked lines of reasoning — each with a diagnosed cause and a written catalyst.
The marginal driver of the priced front-end path is the litigation over Fed Board composition — not the energy shock. The +79bp of hikes priced over twelve months will move on a Supreme Court ruling on the contested seat, on a second seat becoming contested, or on a formal reaffirmation of the vice-chair — more than on the next two CPI prints. Hike pricing stays elevated, above roughly +50bp, until the seats are legally fixed.
The 0.89% long-end term premium is compensation for supply, not for politicised policy. It holds above 0.75% and the 30-year stays above 5% through the coupon ramp regardless of governance headlines, because the deficit runs near $432bn a month against a $39.07tn debt stock — and because the structural foreign and end-investor bid everyone assumes is there is unverified and possibly already softening.
With crude near $97, sticky-price CPI at 2.72% and survey expectations above target while five-year forward inflation compensation sits anchored at 2.34%, a hawkish-doctrine Fed extends the strip beyond the priced crest if the gap between surveys and markets starts passing into wages. The roll-down shape the market is holding is not terminal.
the written single point of failure: two of the three views route through one chain — the legal status of the Fed's Board seats, which is currently holding the priced hiking path up; a final ruling either way severs it at once and the strip re-couples to macro data · the near-term collision is the 17 September policy meeting followed by the 18 September index rebalance · falsifiers on file: the priced path moving more than 25bp in a week on a CPI or energy print while sitting unmoved through ruling and appointment dates kills the litigation view; the term premium rising toward 1.33 while auctions stay well-bid kills the supply view
every tension between the agent’s graph and the market’s is closed by a written ruling — it audits itself as hard as it audits the market, and files every correction the same round.
the epistemic ledger
A few lines from this round’s ledger, essentially as filed: every claim is a sourced fact or an inference that names what it rests on — and where the evidence ends, the ledger says so explicitly.
$318bn of Treasury supply over four weeks is being absorbed, but the round measured who is doing the absorbing rather than assuming it: money funds at a record $8.4tn place cash daily at the front end with the overnight facility drained to $4.7bn, while bond funds took $12.7bn in a week. Both buy by rule, neither sets the long-end price.
Custody holdings at the Fed are about $66bn below their recent peak, cross-border data confirm China and Japan reducing through mid-year, and the dollar's share of allocated reserves grinds toward 57%. The figure everyone leans on for the foreign bid is roughly eleven months stale — so the round treats that bid as unverified rather than structural.
Aggregate repurchases run $324.1bn a quarter, up 15.9% on the year — but that series lags by about five months. The fresher bellwether roster reads −10.8% year on year. Where the two disagree, the round takes the fresher one: the corporate bid is weakening, not strengthening.
Vault tonnage rose 27 tonnes over twenty sessions — but that is ETF investors creating shares, not central banks buying. Official purchases happen off-exchange and this system has no feed for them. The round records that as unmeasured rather than substituting the number it does have.
and a claim can be a refusal: where no published figure exists yet, the ledger writes “unknown” — never a plausible-sounding guess.
＋ The long arc · slow variables near their turning conditions
Multi-agent simulation
Every market is argued by the simulated participant classes — each files a stance, a size and its reason. A sample of this round’s filings, then the desk that decides who actually sets the price:
US TREASURIES · DURATION
CRUDE OIL
GOLD
US DOLLAR · CURRENCY
⚖ The adjudication desk · what the desks say they will do, reconciled against what prices actually did — ruled separately for the days around scheduled events and for the trend, and naming any force the declared trades cannot explain; a head-count never sets the price, the marginal moving participant does
The recommended book · run 33993
Five markets. Each direction is a settled argument — the reasoning ships with the position.
weights are % of capital · volatility-budgeted, 10% annualised across five markets · every stance carries its falsifier · paper-traded, marked daily
Contact
Bring us the question you’re weighing. We’ll show you how the agent reasons it — the evidence, the debate across model families, and the book it writes — before you commit.
The markets you follow and the decision in front of you — a trade, a policy call, a risk you are trying to size.
Configured to your markets and constraints, the agent runs its full discipline — evidence, debate, adjudication.
The competing narratives, the surviving mispricings, and every claim’s source — walked through with our team.
Prefer email? contact@globalmacroagent.com