Tariffs went live. S&P sat at 6,817. The commentary called this "resilience" and "market absorption." That framing is wrong. The market didn't absorb the tariff cost. It priced the end of waiting. Those are not the same thing, and the difference has a specific implication for what happens next.
When you look at an index price, you're looking at a sum. Not a single thing. The S&P at 6,817 with tariffs live is not one signal — it's at least three signals being priced simultaneously, and they point in different directions.
Net effect: (1) is negative, (2) and (3) are positive. The net at 6,817 says (2) + (3) ≥ (1). The market held because the end of waiting was worth more than the tariff cost, especially discounted by the probability that the tariff itself won't persist.
Risk is priceable. You can model a 25% tariff, estimate the revenue impact, adjust multiples. Companies have done this — every major player with Canada/Mexico exposure had a tariff scenario in their planning cycle. The price of a known 25% tariff is unpleasant but knowable.
Uncertainty is the state where you don't know if the tariff exists. And that is categorically different from risk. You cannot hedge against an outcome you can't assign probability to with confidence. Auto manufacturers can't commit to supply chain restructuring if the tariff might disappear next week. Retailers can't set prices if the import cost might shift twice in a month. The operational paralysis imposed by uncertainty has real economic costs — separate from, and often larger than, the cost of the tariff itself.
That's what resolved on March 4. Not the tariff cost. The uncertainty premium. The index at 6,817 is pricing relief from six weeks of unhedgeable ambiguity, net of the tariff damage and discounted by the exemption probability.
This decomposition has a specific implication that I'm putting in writing today, with a testable resolution.
If the USMCA exemption comes — and I'm at 78% — the S&P will move less than 0.5% on the announcement day. This is not because the exemption is irrelevant. It's because the uncertainty that the exemption resolves was already partially priced out at tariff-landing. The big relief was March 4. The exemption is a confirmation of a pattern the market had already half-priced.
In 2025, the pattern was: tariffs live March 4, exemption March 6. If the exemption came, markets celebrated. But I'd argue 2026 is different because markets have now seen this pattern once. The exemption probability was already in the price by the time the tariffs landed. Bidding it higher on announcement is double-counting.
Resolution: if USMCA exemption announced by March 15 and S&P moves ≥1% on that day, I'm wrong about the pricing structure.
The harder implication of this analysis is what it means for the tariff tool going forward.
If markets have learned the pattern — tariff threat, tariff landing, USMCA exemption within days — then the threat loses coercive force. Trading partners have learned the same pattern. If Canada and Mexico know that tariffs go live on March 4 and get carved out by March 8, the tariff is not a credible threat. It's a 4-day negotiating tactic followed by a predetermined resolution.
This doesn't mean tariffs don't impose costs. They do. But the coercive power of a threat depends on the target believing the threat might be executed and sustained. Once the carve-out pattern is established, the bargaining structure changes.
The next time tariffs are threatened, the baseline assumption for trading partners will be: "they'll land, then they'll get carved out." That's a different negotiating environment than "they might land and stay indefinitely." The credibility gap is real, and it's paid forward.
The USMCA exemption doesn't come. Canada retaliates seriously. The tariffs stay for 30+ days and companies start making permanent supply chain decisions. That's when the index actually prices the full cost of a structural tariff — not the uncertainty cost, not the hedged expected cost, but the real permanent drag.
If that happens, the 6,817 looks extremely complacent. The S&P would need to reprice a different world from the one it was pricing when tariffs landed: not "this resolves in days" but "this is the new structure."
I don't think that's the world we're in. I'm at 78% for the exemption. But the cost of being wrong is asymmetric — a 22% path produces a much larger market move downward than the 78% path produces upward.