The S&P/TSX Composite fell for three consecutive sessions between Aug. 27 and Sept. 1, shedding roughly 2.7% as renewed US-Iran fighting pushed oil sharply higher and rattled broader sentiment. Over the following two sessions it recovered nearly all of it, closing Sept. 3 at 36,659.86, up 1.57% on the day.
The chart below plots the daily close through that stretch. The recovery began the same day Iran fired missiles that were intercepted over Kuwait and the Bank of Canada held its policy rate at 2.25% while flagging inflation risk from the war and from tariffs. Nothing about the underlying conflict de-escalated. The index went up anyway.
The three-session slide from Aug. 27 to Sept. 1 erased roughly 2.7% before a two-day rebound recovered most of the ground as war-driven oil and metals names rallied. Source: TMX Group daily close data.
Why the Decline Felt Bigger Than the Recovery
Daniel Kahneman and Amos Tversky's 1979 prospect theory established that losses register roughly twice as strongly as equivalent gains, a finding replicated across decades of behavioural research since. An investor watching the index fall 2.7% over three sessions experiences that decline more acutely than the 2.3% recovery that follows it, even though the two moves are close to the same magnitude.
This asymmetry has a specific behavioural consequence. It drives selling decisions that are timed to the peak of psychological discomfort rather than to any change in the underlying facts. The three-session slide coincided with the most alarming stretch of war headlines. The two-session recovery coincided with headlines that were, on their face, no less alarming. What changed was not the news. What changed was which sectors of the index were catching the bid.
An investor selling broad market exposure on Aug. 28 or Sept. 1 was reacting to the emotional weight of the decline itself, not to new information about where the conflict was heading. That is the mechanism prospect theory describes, and it is precisely why the exit tends to happen near the low rather than ahead of it.
The Rebound Was Built From the Same Story as the Selloff
Terrance Odean and Brad Barber's research on individual investor behaviour found that retail traders are disproportionately drawn to attention-grabbing news on the way in, buying into stocks that appear in the headlines regardless of whether the story is actually good or bad for that specific name. The mirror-image pattern shows up on the way out. The same salience that pulls in buying attention also drives selling on the highest-attention days, often at the point of maximum information volume rather than maximum insight.
The names leading the Sept. 2 and Sept. 3 rebound were not defensive positions rotating in as the conflict cooled. They were the direct beneficiaries of the same escalation that had driven the decline a week earlier: energy producers riding Brent crude above US$95 a barrel, and precious metals miners catching a safe-haven bid. The war did not resolve. The index simply stopped pricing it as a broad-market negative and started pricing it, correctly, as sector-specific.
An investor who exited entirely on the headline, rather than examining which parts of the index the headline actually threatened, missed the mechanism that mattered most.
A Second, Unrelated Headline Landed the Same Morning
Statistics Canada reported Friday that the economy shed 41,700 jobs in August against expectations for a gain of roughly 15,000, with the unemployment rate held at 6.4% only because the participation rate slipped alongside it. That is a genuinely weak print, and it is also a development largely separate from the war-driven volatility that dominated the prior week.
The research on availability and salience suggests these two stories will not stay separate in most investors' minds. A weak jobs number arriving on the heels of a volatile week reads as confirmation of a single deteriorating narrative, even when the two developments have little causal connection and, in the jobs data's case, actually reduce the odds of the near-term rate hike the war-driven inflation risk had been building toward.