The full PDF of this report is available for download below.
Anatomy of A Canadian Bear is one of several market-specific bear market studies we are compiling. We wrote it to prepare for a bear market, not to forecast one. Decisions about how to act in a drawdown are best made before it begins, against the historical record, and these studies are part of our firm's effort to prepare.
The S&P/TSX Composite has fallen 15% or more from a closing high sixteen times since 1979. The average drawdown during those episodes was 27.6% over eight months; the median was 23.4% over six. The gap is almost entirely the two structural bears, 2000-02 and 2008-09, which fell roughly 50% and took an average of 51 months to regain their peaks. The ten event-driven bears averaged a 22.7% decline over three and a half months. The type of bear matters more than the average.
The findings most relevant to positioning:
The recognition date, the first close 15% below the peak, has been a better point to add than to cut. From that date, the median further decline was 8%, and the twelve-month return was positive in eleven of sixteen episodes. The four exceptions were the two structural bears, 1980-82 and 2020.
Consumer staples and utilities outperformed the index by an average of 27 and 19 percentage points through the declines. The rotation out of defensives and into cyclicals at the trough has a strong track record of outperformance; energy has been a reliable source of downside and an unreliable source of recovery.
Small caps fall further and take longer to heal, with a median of 23.8 months to regain their peak against 13.2 for large caps. The small-cap rebound premium appears in the 2008-09 and 2020 recoveries and is absent otherwise.
In local currency, the TSX fell less than the S&P 500 in 15 of 24 US corrections since 1979. In US dollars, the cushion disappears: -21% on average against -20%, because the Canadian dollar weakens when risk is sold.
Dispersion is wide. In 2000-02 the index halved while the median constituent fell 6.9%. That dispersion breadth is where stock pickers earn or lose in a bear market.
Our Canadian Bull/Bear composite reads 60% through September 18th (the data cutoff for this study), below the 66% to 82% range that preceded past rate-driven and structural bear markets. Valuation is holding the composite up: the CAPE stands at 29.2, its 98th percentile since 2001. We expect a bear market starting from here to be more likely event-driven or cyclical than structural, and we would treat its recognition date as a point to add, at least given current data. That view would change if growth momentum and real commodity prices move into their upper quintiles with valuation unchanged, the configuration that preceded 2008.
