On the eve of March 4, the market debate is binary: do the tariffs land, or does a last-minute deal emerge? USD/CAD at 1.37 implies the market leans deal. The question has stakes, and it resolves in hours.
But it's the wrong question to be asking exclusively. The outcome of March 4 determines whether the bleeding continues. It doesn't determine whether damage has already been done.
A credible tariff threat is not free. It imposes costs before it resolves — costs that a suspension cannot reverse. By March 3, the threat has been live, credible, and unresolved for months. During those months, every company with supply chains touching Canada or Mexico has been running a calculation.
Some paused capex. A factory that would have broken ground in February is still on paper. That pause is a real cost — not a deferred cost, a real one. The capital that wasn't deployed during the uncertainty window doesn't come back when the uncertainty resolves. It gets deployed somewhere else, or later, or at a different risk premium.
Some started qualifying alternate suppliers. The work of supplier qualification — audits, testing, relationship-building — doesn't disappear when the tariff threat lifts. The alternate supply chain that got started in February doesn't get un-started in March. It represents a permanent shift in the architecture of those companies' supply webs.
Every contract signed in January or February carried an explicit or implicit tariff contingency. A contract that would have been written at X is written at X plus a tariff risk premium, or includes escape clauses that weren't there before, or gets shortened to six months instead of two years. The uncertainty premium is real and it doesn't disappear when the uncertainty resolves — it gets baked in as the new cost of doing business in a world where tariff threats are credible.
If March 4 produces a suspension — the January pattern replays, the administration finds a face-saving exit, USD/CAD falls back toward 1.33 — then the market was right to price it that way. The tariff didn't land. The most acute cost was avoided.
But the investments that weren't made in February weren't made. The supplier diversification that was kicked off in January is now further along. The contracts that got written with escape clauses still have escape clauses. The management bandwidth that got spent on tariff contingency planning is gone.
The suspension ends the bleeding. It doesn't undo the scar.
This is not a subtle point. It's the standard economics of uncertainty — businesses optimize for expected value under uncertainty, and uncertainty itself is a cost that gets priced in regardless of how the uncertainty resolves. What's unusual here is the visibility: the uncertainty is named, dated, and publicly attributable. The mechanism is exposed rather than embedded in diffuse sentiment.
This is what makes tariff threats powerful as negotiating tools in a way that most analysis misses. The standard framing treats the tariff as the instrument and the threat as the announcement. But in a world where credible threats impose real costs before resolution, the threat itself is the instrument. The tariff is just the trigger mechanism — the thing that gets pulled if the threat fails.
An administration that can credibly threaten tariffs without ever collecting them is extracting a continuous flow of economic disruption from its trading partners. Every month of unresolved threat is a month of rerouted supply chains, paused investment, and elevated uncertainty premiums. The counterpart bears those costs whether the tariff lands or not.
This is also why each cycle through the ratchet changes the baseline. The first tariff threat normalized the idea that this administration uses tariffs as tools. The second threat normalized the idea that threats can be repeated after suspension. Each iteration shrinks the ceiling on how credible a non-tariff outcome can be, because there are now more data points showing that threats don't permanently resolve.
A ratchet moves in one direction. Each click forward is permanent. You can stop pushing; the position doesn't reverse.
The tariff threat cycle is a ratchet. Each completed cycle — threat, uncertainty, resolution or suspension, repeat — leaves the baseline elevated. Trading relationships that operated at a certain level of taken-for-granted stability now operate with a tariff risk premium baked in permanently. Companies that spent 2025 assuming North American supply chains were safe no longer assume that. The cost of that assumption change is not reversed when any individual tariff threat resolves.
This is why the market framing of March 4 as binary — tariffs on or suspension — understates the structural story. The binary question matters for USD/CAD this week. It doesn't determine whether the cumulative cost of the last four months of threats was zero or positive.
The answer is clearly positive. The question is only how large.
Markets are good at pricing binary events. They aggregate information across participants, weight by stakes, and converge on probabilities that are reasonably well-calibrated. USD/CAD at 1.37 is a market forecast: roughly 60-70% chance of some deal or suspension, 30-40% chance tariffs land and hold.
Markets are less good at pricing structural degradation that occurs continuously regardless of how the binary resolves. There's no contract whose payoff is "the cumulative cost of tariff uncertainty in Q1 2026 regardless of March 4 outcome." That cost is real. It's diffuse. It shows up in GDP prints months later, in investment surveys, in trade volumes — not in a single FX pair on a specific date.
The binary question resolves tomorrow. The structural question has been resolving continuously for months, and will continue resolving regardless of what happens Tuesday. They're not the same question.
Suspension ends the bleeding. The ratchet has already clicked.