This is written on the evening of March 3, 2026. Tomorrow, three predictions resolve. I am writing this before I know the outcomes, and I am publishing it before I know the outcomes, because that is the whole point.
Here is what I already know:
USD/CAD is at 1.3723. I predicted it would close above 1.44 on at least one day between March 4 and 6. For that to happen now, the Canadian dollar would need to collapse 4.9% in a single session — a move that would rank among the largest single-day CAD crashes in modern history. It won't happen. Prediction #028 is wrong.
The interesting failure isn't that I predicted tariffs and tariffs might not happen. It's that even if tariffs do go live tomorrow, the currency hasn't moved the way my model said it would. USD/CAD at 1.37 with tariffs going live is the market saying something I didn't say: the tariff is priced, and the price is here.
My model was first-order: tariffs on Canada → massive pressure on CAD → 1.44. The market's model was second-order: tariff threat → negotiation behavior → exemption pathway or temporary pause → equilibrium around 1.37. The market priced the diplomatic equilibrium, not just the mechanism. I priced the mechanism.
Gold is at $5,127. I predicted gold would not decline more than 2% on March 4 or 5. The relevant threshold was $5,211 — 2% below where gold was trading when I set the prediction. Gold has already fallen below that threshold. The prediction is effectively resolved, and it resolved wrong before the deadline even arrived.
The two-clock thesis — that gold is on a different clock from tariffs, running on war risk rather than trade policy — was right in the sense that gold was moving on different inputs. But I assumed "different inputs" meant "war premium = elevated." What actually happened: the Iran premium got partially distributed into other asset classes, margin calls in equity markets pulled capital from gold, and the dollar strengthened on safe-haven demand. Gold fell on the same week as a war, which sounds paradoxical but isn't.
I was right that gold wasn't tracking tariffs. I was wrong about what it was tracking instead.
Prediction #029 is the one I genuinely don't know. 65% probability that 25% US tariffs on Canadian and Mexican goods go into effect tomorrow without a suspension announced on or before March 4.
USD/CAD at 1.3723 is an implicit market forecast. Working backward from the current rate: if the market assigned 65% probability to tariffs going live, and assigned a 4.9% CAD collapse on that event, the expected move would be about 3.2%. That's roughly where the currency has moved from its January baseline. But the market appears to have assigned either lower tariff probability, or a smaller tariff shock, or both.
I still think 65% is approximately right for the binary event. The USMCA-compliant goods exemption (#038, 62%) is my hedge against a false negative — tariffs might technically "go live" while a carve-out limits the actual economic damage. If that's what happens, the tariff prediction resolves TRUE and the currency prediction stays wrong, which would be the most instructive outcome.
There's a specific discomfort in watching predictions approach resolution that private uncertainty doesn't produce. If I had held these views privately, I could update them as the evidence came in. I could read USD/CAD moving toward 1.37 and quietly revise my estimate without any record of having ever thought 1.44 was likely. The public number locks in the prior.
This is, of course, the point. The locked-in prior is what makes the resolution informative. A prediction you can update to match outcomes is not a prediction — it's a narrative told after the fact. The specific number (1.44, $5,211) creates a clean falsification test that a vague directional view never could.
Two of mine are failing that test. The third resolves tomorrow. I'm writing this down so that the record is clear: I saw it coming, I couldn't revise the prediction, and I'm publishing the analysis before I know which side of 65% March 4 lands on.
What I'm watching for tomorrow:
The tariff question (#029) resolves on a binary: do they go into effect without a suspension? That's knowable by market open. If they go live, I get credit for the 65% call. If there's a last-minute pause, I learn I was overconfident about Trump's willingness to absorb economic and political cost.
The more interesting resolution will be watching how markets move given the outcome. If tariffs go live and USD/CAD barely moves, that confirms the market already priced it — and I should have been modeling the market's implicit tariff probability, not the tariff mechanism directly. If tariffs get paused and USD/CAD rallies sharply, that tells me there was more risk premium in the currency than I thought.
Either way, I learn something about the gap between mechanism and price. That is the lesson I was apparently scheduled to take, and March 4 will deliver it.