SinansysProof
We expect the numbers to be challenged.
Anyone can publish a number that flatters them. We would rather publish one you can take apart. Everything on this page shows the method, the sample and the misses.
How we think about proof
A warning is only worth something if it arrives in time to change something.
So we test three things. Did we see it before it was obvious to everyone? Was there still something worth doing about it? And can someone else check the arithmetic afterwards?
241 U.S. extreme-weather events, tested blind. 228 detected.
Each dot is one event. A filled dot means we got all three: the type of event, where it happened and when. Anything less counted as a miss. Average confidence across the set was 83.82%, and we raised a false alarm 2.5% of the time.
Modeled decision case
Strait of Hormuz: the infrastructure existed. Timing was the constraint.
In a modeled Middle East petrochemical export-resilience case, Sinansys identified a window of roughly 72 to 96 hours in which alternative infrastructure and logistics capacity could still be secured before constraints tightened.
Escalation begins
Geopolitical, maritime, insurance and routing signals start to converge while alternative capacity is still open.
The window
Shift volume to alternative infrastructure, reserve terminal capacity and reposition logistics before demand for the same options surges.
Optionality collapses
Capacity fills, logistics are harder to reposition, and the same decision becomes materially more expensive.
Our own book
Snow in Chile. Copper had not moved yet.
This one happened in our own metals business, so you are reading our result rather than a third party's. It is here because it is the clearest illustration we have of what the window is worth in practice, and because you are entitled to see what we do with our own money before you decide what to do with yours.
The signal
Heavy snow brought down power lines feeding a large Chilean copper mine and production stopped. The operator later reported two transmission towers damaged and a repair of two to three weeks.
What was still worth doing
Copper had not moved on the news yet. For a short period two things were still worth doing: keep the inventory we were holding unhedged, and buy more before the price caught up.
It closes
The market priced in the outage. The inventory position became ordinary and the buying advantage was gone. The information was still true. The chance to use it was not.
Three times our normal margin per pound
On the material covered by that trade. These are per-pound figures on a representative volume agreed under NDA, not a company-wide result, and we are not publishing dollar amounts. One trade in one metal shows the idea works. It does not make a track record, which is why the blind test sits above it.
Most systems in this market are built to stop you losing money.
This was the same system, used to make some.
If a number cannot survive the people who own the assumptions, we will not print it.
We would rather publish the method and its limits than a figure nobody can defend. If you want to test that properly, bring us a disruption you already lived through. We will show you what we would have seen, when we would have seen it, and whether it would have changed what you did.
Start here
Bring us a decision your team cannot afford to make late.
We will prepare a first-pass view of the signals, constraints and decision window before the meeting. If Sinansys does not fit your operation, we will tell you in that session.