How long do Trump's tariff threats last?

A typical presidential tariff threat lasts about seventy-two hours before the retreat or the exemptions arrive. That is the finding of a detailed event study of trade announcements made between April 2025 and January 2026, which found the pattern predictable enough to model. The sample runs from the April tariffs on almost every trading partner to this month's threat against European allies over Greenland, and the shape of each episode is strikingly similar.

Threat, panic, climbdown

The study of fifty-two executive trade announcements finds the same cycle again and again. First comes a surprise threat at the maximum, usually over a weekend on social media or in an executive memo, and stocks and currencies drop. Next, markets absorb the shock for 48 to 72 hours while corporate lobbyists fill Washington with requests for exemptions. Finally, the administration announces a 'constructive dialogue', pushing back deadlines or exempting large industries.

Three retreats that set the pattern

The template was cast in April 2025. The sweeping tariffs announced on 2 April gave the S&P 500 its worst two-day fall since the pandemic crash of 2020, and a week later, with Treasury yields jumping as well, the White House paused most of the higher rates for 90 days. That pause produced the index's biggest one-day gain since 2008. In May came two more reversals: tariffs on China were cut from 145 per cent to 30 per cent after talks in Geneva, and a threatened 50 per cent levy on European Union goods was postponed within days of being announced. By late spring a Financial Times columnist had given the trade a name, TACO, and the acronym spread across trading desks.

The autumn supplied another example. In October, after Beijing tightened its controls on exports of rare earths, the president threatened an additional 100 per cent tariff on Chinese goods. At the end of the month he met Xi Jinping in South Korea, and the threat gave way to a truce in which Washington trimmed some existing duties and China paused the new export controls. The sequence of maximum threat, market fall and negotiated retreat ran to form, only over weeks rather than days.

Why the cycle repeats

The mechanism has three parts. A maximal threat gives the administration a headline and an opening position, and costs nothing until customs officers start collecting. Falling equity prices and, more tellingly, rising bond yields then raise the political cost of carrying it out, because the Treasury has a large deficit to finance and mortgage rates follow long yields. Finally, the exemption process gives the White House a way to retreat that looks like a negotiation, with each carve-out presented as a concession won rather than a threat withdrawn.

The credibility problem

Game theorists have studied this dynamic for decades. Thomas Schelling's work on bargaining in the 1960s showed that a threat is only as strong as the other side's belief that it will be carried out, and that a negotiator who repeatedly backs down teaches opponents to wait. Trading partners have drawn that lesson. After the first rounds, several governments stopped offering quick concessions and instead let markets do the arguing, calculating that falling share prices and rising yields in America would bring the deadline forward faster than any counter-threat.

Markets stop flinching

Traders have watched the same retreat play out with Mexico, Canada, Europe and Asia, and markets react less each time. Spikes in implied volatility fade faster, and trading algorithms are now set up to buy the first dip. The VIX index of expected equity volatility went above 50 in early April 2025, a level associated with the crises of 2008 and 2020, and later threats pushed it only a fraction as high. The pattern feeds on itself: each quick recovery rewards the traders who bought the dip, and their buying shortens the next sell-off. The same logic applies to options: when every sell-off reverses within days, those who sell insurance against a crash are repeatedly paid for taking the risk, which draws in more sellers and cheapens protection. That leaves markets less prepared for the one threat that is carried through.

Retreats are not refunds

The case against complacency is that a retreat from the maximum is rarely a return to zero. A baseline tariff of 10 per cent on most imports stayed in place after the April pause, tariffs on steel and aluminium were doubled to 50 per cent in June, and the average effective American tariff rate is the highest since the 1930s. Companies that treated every threat as noise have still paid more at the border. The event study measures how fast the headline threat fades, and it says less about the level at which each episode settles.

The sample also has a selection problem. The retreats are memorable because they move markets, but a number of threats were carried out in full or close to it. Tariffs of 50 per cent on goods from Brazil and India took effect in August, the first justified by a political dispute and the second partly by India's purchases of Russian oil, and they stayed in force for months. An investor who assumed every threat would dissolve within three days would have been wrong on those, and the cost of being wrong fell on the exporters and importers who had to pay.

There is also a risk in the pattern becoming too well known. A threat that markets price as a bluff gives the administration less leverage, and a president who dislikes being called predictable may one day carry a threat through simply to restore its force. The indicators to watch are whether a threatened tariff is published as a formal order with an effective date, whether exemption requests are refused rather than deferred, and whether Treasury yields, rather than equities, start to fall on the news. The threats are wearing out. After a year of retreats, markets now treat a new tariff threat as a buying opportunity, and the White House will probably find each one moves prices less than the last.