When a market doesn't move ahead of a known event, it isn't neutral. It has a position.
USD/CAD sat at 1.3656 on Sunday, March 2 — two days before 25% tariffs on Canadian goods were scheduled to take effect. The market wasn't uncertain about March 4. It was long deal, short tariffs. That's a directional bet. It resolves Tuesday.
The source of the bet is what happened in January. Tariffs on Canada and Mexico were announced, threatened publicly, and then suspended within 30 days — after Canada and Mexico offered border security and fentanyl interdiction measures that let the administration declare partial victory. The market learned a pattern: Trump tariff announcements come with compliance escape hatches. The threat precedes the concession; the tariff itself rarely lands.
That pattern got encoded in price. When subsequent tariff deadlines approached, the market applied the January prior: there's some probability of a last-minute deal. The prior is reasonable — it's based on observed behavior. The market isn't wrong to have learned it. It might be wrong to apply it here.
The administration announced March 4 tariffs explicitly as revenue tariffs and trade deficit correction — not a behavior-change mechanism. January's escape hatch was specific: if Canada addresses the border and fentanyl, the tariff recedes. The rationale and the concession were linked. Pull the lever, get the relief.
March has no such link. There's no compliance move Canada can make that addresses a revenue problem or corrects a trade deficit. The border/fentanyl lever was already pulled in January. There is no second lever. If the stated rationale is the actual rationale, then there's no escape hatch — because you can't concede your way out of being asked to pay a tax.
The market at 1.36 implies it either doesn't believe the rationale shift is real, or it believes the administration will manufacture a face-saving exit anyway. That's a possible read. It's also a borrowed prior from a structurally different event.
If tariffs go into effect and hold for even a week — the repricing from 1.36 toward 1.42 or beyond happens in hours, not days. The same dynamic played out during the January peak, when markets briefly priced full tariff implementation and USD/CAD touched 1.47 before reversing on the pause announcement. The gap between 1.36 and 1.44 isn't exotic. It's precisely the January-to-implementation repricing, compressed into a shorter window.
Markets are usually right about binary events because information aggregates across thousands of actors with different models. But they can be systematically wrong when they apply a learned pattern to a structurally different event — especially when the surface features match (tariff announcement, imminent deadline, same countries) and the structural difference is easy to miss (compliance mechanism absent).
Model transfer errors are harder to see before they resolve than after. The January reflex will look obviously wrong in retrospect if tariffs stick — but right now, it looks like reasonable caution.
The unusual thing about this moment: the silence before March 4 is a specific thesis about how administrations behave, derived from a single data point, with a public resolution date visible to everyone. Markets don't always give you that — a clear prediction, a verifiable event, a known timestamp. This one does.
If the tariffs land and hold, the market was wrong and 1.36 was expensive. If a deal emerges on March 3 or 4, the market was right and the January model still applies. Either way, we get a data point about whether the compliance-escape-hatch prior transfers to revenue tariffs.
The silence before March 4 is not neutrality. It's a bet. By Tuesday we'll know which kind of stillness it was.