Bettors now see a tariff hike on Canada as more likely than a deal

At 00:59 UTC on 21 July 2026, bettors on Polymarket priced a 63% probability that a new US tariff increase on Canadian goods would take effect before the end of the calendar year. Twenty-four hours earlier, on the same platform, the implied odds of a negotiated US–Canada trade deal by 31 December 2026 stood at 12%. The gap between those two contracts, more than five to one, captures the betting market's working assumption about where the bilateral relationship is actually headed: toward escalation, not accommodation.

The two contracts, read together, sketch a market view that the official statements from Ottawa and Washington have so far refused to draw. One side sees a tariff hike as the modal outcome; the other sees a deal as a tail risk. The middle of the distribution, where most diplomatic processes are supposed to live, has thinned out.

What the contracts actually price

The Polymarket instrument titled "Will the US increase tariffs on Canada by the end of 2026?" traded at 63% on 21 July, with the underlying resolution tied to whether a tariff increase on Canadian goods enters effect before the end of the calendar year. The companion contract, "US makes a trade deal with Canada by end of 2026?," settled at 12% the prior evening at 21:49 UTC on 20 July.

Neither contract specifies which goods, which tariff line, or which statutory authority would carry the increase. That detail matters: a Section 232 national-security adjustment on steel and aluminum would look very different from a broad Section 301 reciprocal-tariff package of the kind Washington has applied to other partners in 2025 and 2026. The contracts price the binary event, not the mechanism. A trader reading 63% is taking a position that something labelled "a tariff increase" will land, by any of the available routes, in the next five months.

The 12% on a deal is the more striking print. Prediction markets typically assign a non-trivial floor to negotiated outcomes in active disputes, on the theory that both sides have an incentive to keep talking. A single-digit-to-low-teens reading suggests the betting public believes the diplomatic channel has been deprioritised, or that whatever is being negotiated does not yet rise to the level of the contract's resolution criteria.

Why the political coverage is lagging the market

Headline reporting through the first half of 2026 has tracked the familiar choreography: a senior official floats a tariff threat, Canada announces a retaliatory list, business groups file comments, both governments signal that talks are "ongoing." The market is now pricing through that choreography. It is treating the theatre of negotiation as cover for an underlying decision that has, in effect, already been made.

There is a structural reason for the gap. News cycles reward movement, and movement is what ministers and trade representatives generate when they meet, call, or threaten. Prediction markets, by contrast, are anchored on a calendar date and a binary outcome. They compress the noise of weekly diplomacy into a single price that moves only when new information arrives. The fact that 63% has held through a series of "productive" and "frank" readouts implies that the market reads those readouts as performance rather than progress.

Counter-read: the bear case for 12%

The case against the dominant 63% is not frivolous. A 12% probability is not zero. Canadian federal and provincial governments retain leverage in several sectors where US supply chains are genuinely exposed, including critical minerals, energy, and softwood lumber. US business constituencies with Canadian exposure have, in prior rounds, succeeded in slowing or carving out tariff actions. And the political calendar in Washington imposes its own discipline: an administration that wants tariff revenue on the books by year-end also wants the political optics of a "deal" if the alternative is a sustained consumer-price reaction in border states.

The reason that case has so far failed to move the contract is timing. Five months is short. New tariff authorities typically require notice-and-comment periods, even when invoked under national-security provisions. A deal of sufficient scope to satisfy Polymarket's resolution criteria would, at this stage, almost certainly require a political decision in Washington to prioritise the bilateral relationship over the broader tariff regime. The market is betting that decision will not be made.

What remains uncertain

The contracts do not specify the mechanism of the tariff increase, the tariff line, or the magnitude. They do not capture the possibility of a partial deal that resolves some sectors while leaving others under threat. And they do not price the political contingency that a domestic US shock, a court ruling, or a shift in Congressional posture could upend the schedule before the end of the year. A 63% probability is not certainty; a 12% probability is not impossibility. Both numbers are best read as the market's current best guess about which way the next five months will break, given the information available on 20 and 21 July.

The Polymarket prints also say nothing about what a tariff hike, if it comes, would actually contain. The political cost of a Section 232 steel and aluminum adjustment on Canadian producers is qualitatively different from the political cost of a broad reciprocal tariff. The market is signalling direction. It is not, by itself, signalling scale.

The structural read

What the two contracts together suggest is a market that has stopped expecting the US–Canada bilateral relationship to be governed by the framework that held from the early 1990s through roughly 2024. That framework, whatever its strains, rested on the assumption that integrated North American supply chains produced a political floor under the commercial relationship. The Polymarket pricing implies the floor is no longer trusted. Whether that distrust is justified is the open question of the next five months. The market's current answer is: probably, and the deal that would prove otherwise is unlikely.

Desk note: Monexus treats prediction-market pricing as one input among several, not as a substitute for primary-source reporting. The Polymarket contracts are useful here precisely because they compress a complex diplomatic standoff into two clean numbers, but neither the platform nor the contracts disclose the composition of the betting pool. Treat the 63/12 split as a signal of trader conviction, not as a poll of Canadian or US public opinion.