Macro Regime Intelligence
Know the regime.
Prove the call.
Markets do not move continuously. They sit in regimes — periods with their own rules — and then they break. The money is made and lost at the break.
GXPOL tells you which regime you are in, how firmly, and what it means for the book you actually hold.
The read
One number, five states, updated daily across seven economies. Behind it: a decade of engineering, and a standing refusal to publish a number we cannot defend.
The problem
The regime turned in February. Your book found out in April.
Most macro research is written after the fact and read for reassurance. It is fluent, it is plausible, and it is unfalsifiable — by the time a view is stated clearly enough to act on, the move has happened.
The alternative on offer is a black box: a vendor score with no stated basis, which you are invited to trust because the vendor is expensive.
Neither is a process. Neither survives a bad quarter and an investment committee asking what you knew and when.
What changes
Four things you can do on Monday that you cannot do today.
See the break coming
Stress that is happening now is measured separately from conditions that merely look benign. When the two diverge, that is the warning — and it is the sequence that has preceded most regime breaks we can measure.
Know where you are exposed
Not "equities are risky". Which sectors, which industries, which positions in the book you hold — ranked, with the model's own confidence attached to each.
Find the precedent
Today's conditions matched against two decades of history — not by price level, but by the shape of the stress. What those environments did next is a better guide than an opinion.
Defend the decision
Every call is timestamped with the conditions that produced it. When the committee asks why you moved — or why you didn't — the answer is a record, not a recollection.
Before you ask
Is this another black box?
You can see what each engine measures and how the headline score is built: it is the plain average of all ten.
We tried to beat equal weights three times, with a velocity blend, volatility-scaled thresholds and per-country weights. None of them held up out of sample, so equal weights is what runs. The results are in the methodology paper.
Coverage is weaker for Canada and India, and every reading shows how much of it is backed by live data.
Start here
Send us your book. We will tell you what it is short.
A benchmark or a holdings list is enough. You get back a regime-conditioned exposure read: where the book breaks, what the closest historical precedents did next, and where the overlay would move you. No call required.