The S&P/TSX Composite closed Wednesday at 35,235.87, which is 2.85% below its August 31 close of 36,270.48 and 1.09% above its June 30 close of 34,856.99. Both statements describe the same index on the same day. Behavioural finance has a name for why only one of them tends to be felt: myopic loss aversion.
The September decline was real and had identifiable causes. The Federal Reserve raised its policy range by 25 basis points to 3.75% to 4.00% on September 16, its first increase since 2023, in a unanimous 12 to 0 vote, according to CNBC. Reuters reported that rising global bond yields and escalating Canada-U.S. trade tensions also weighed on the benchmark, which ended a five-month winning streak. The same Reuters coverage, carried by Baystreet.ca, put the index on course for a ninth straight quarterly gain, the longest such run on record.
One Index, Three Reference Points
Daniel Kahneman and Amos Tversky, publishing in 1979, showed that people evaluate outcomes as gains or losses relative to a reference point rather than as levels of wealth. Their 1992 follow-up estimated that a loss is weighted about 2.25 times as heavily as a gain of the same size. The reference point is therefore the whole story, and September offered at least three of them.
An investor anchored on the August 31 statement sees a 2.85% loss. One anchored on the June 30 statement sees a 1.09% gain. One anchored on the August peak of 37,069.11, the record level listed by Trading Economics, sees a 4.95% drawdown. Three baselines produce three emotional verdicts from a single set of closing prices.
How Frequent Checking Manufactures Losses
Shlomo Benartzi and Richard Thaler argued in 1995 that the historical equity premium is best explained by investors who combine loss aversion with frequent evaluation of their portfolios. The more often results are checked, the more often a loss appears, and a loss-averse mind registers each one at roughly double weight. The evaluation period consistent with the historical premium, they estimated, was about one year.
Thaler, Tversky, Kahneman and Schwartz tested the mechanism directly in 1997. Participants who saw their results after every round allocated less to the risky asset than participants who saw results only after several rounds, even though both groups faced the same return distributions.
The 12 sessions from September 15 to September 30 show the mechanism at work. The index closed lower on 8 of them and higher on 4, with the average down day at 0.51% and the average up day at 0.70%, according to TMX Money closing data. An investor who checked daily met a loss two times out of three. An investor who checked once a quarter met one gain.
Measured from June 30, the index sits above its starting point on all 12 of those sessions, while measured from August 31 it sits below on 11 of them.
The June 30 close of 34,856.99 sets the quarter baseline and the August 31 close of 36,270.48 sets the month baseline, with both lines computed from the same 12 published closes. The index closed above its August 31 level once in the period, on September 22.
The Anxiety Has Rational Roots
Reframing the window does not make the month disappear, and the behavioural literature does not treat the discomfort as a defect. The inputs did change. The Fed moved from holding to hiking, the Bank of Canada held at 2.25% on September 2 for a seventh consecutive decision, and Statistics Canada reported August inflation of 3.0%, with gasoline up 22.8% from a year earlier.
Availability compounds the effect. Tversky and Kahneman showed in 1973 that people judge how likely an event is by how easily examples come to mind. Wednesday supplied vivid examples: Royal Bank of Canada fell 1.1%, Toronto-Dominion Bank fell 0.9%, and First Quantum Minerals closed 15.3% lower after a Panamanian commission recommended the orderly closure of the Cobre Panama mine, according to Trading Economics. A headline about a snapped winning streak is a more available memory than a statistic about nine consecutive quarterly gains.
What Resets the Reference Point Next
Thursday is the first session of the fourth quarter, so the quarterly baseline resets to the September 30 close of 35,235.87. A tenth consecutive quarterly gain would extend the record and a loss would end it, and either outcome will arrive as a headline.
Two scheduled events will move the reference points before the new quarter is a month old. Statistics Canada publishes September CPI on October 19, and the Bank of Canada announces its next rate decision, with a Monetary Policy Report, on October 28. Each date offers investors a fresh, shorter window in which to judge themselves, and the research suggests that the shorter the window, the less comfortable the reading.