digest
Short Summary
Main recommendation: Stay invested through market volatility and plan using long-run expected returns rather than expecting "normal" returns in any given year, since annual real stock returns are rarely near the average. Ben Felix argues that the realistic approach is to build reasonable expectations about the future, choose an asset allocation you can tolerate in a worst case, and avoid trying to dodge downturns.
Key Points
- Long-run data (Credit Suisse Global Investment Returns Yearbook, back to 1900, 23 countries in US dollars) shows global real equity returns averaging 5.2% per year, even including total losses for Russia (1917) and China (1949).
- Averages are misleading for single periods: US stocks underperformed Treasury bills about 15% of the time over rolling 10-year periods since 1926, and Canadian stocks did so over rolling 10-year periods since 1970.
- From 1974 to 2021 (a 47-year period), real compound average returns in Canadian dollars after Canadian inflation were about 5.2% (S&P/TSX Composite), 5.3% (MSCI EAFE), and 7.0% (S&P 500), yet the average return was earned in exactly 0 of the 47 years.
- Extreme years are the norm: the band of -8% to +15% real return captured only about half of annual returns in each market, meaning the other half were very positive or very negative, so extremes should be expected.
- Peter Lynch's point is cited: far more money has been lost by investors preparing for or anticipating corrections than in the corrections themselves.
- Expected returns are estimated by blending a 50-year historical equity risk premium with a valuation-based estimate using the Shiller CAPE, which reflects current conditions while staying more stable.
Details Worth Noting
- Concrete instructions:
- Estimate inflation using the yield difference between a 30-year government bond and a 30-year real-return government bond (about 1.7% at the time of recording).
- Start from the 50-year historical real return of each asset class as the equilibrium cost of capital, but vary the equity risk premium year to year (or by entry year) as conditions warrant, per Damodaran's 2017 observation that premiums change quickly.
- Compute a second estimate from current valuations: the Shiller CAPE, i.e., current price divided by 10-year trailing inflation-adjusted earnings. High CAPE implies lower expected future returns; low CAPE implies higher.
- Average the historical-premium estimate and the CAPE-based estimate to get a figure that reflects current conditions but is more stable.
- Pick an asset allocation you can live with in a worst-case scenario, recognizing the worst case may be worse than anything in your lifetime.
- Stay invested regardless of how bad markets look, since avoiding downside has historically made investors worse off.
- The speaker notes the valuation-based method is more volatile and changes more rapidly than the long-term risk premium, which is why blending the two is preferred.
- The full method is laid out in a 2016 PWL Capital white paper, "Great Expectations," by Raymond Kerr and Dan Vallatti, linked in the video notes.
- Canadian and US market figures cover 1974 onward; the global 5.2% figure covers 1900 onward and includes two markets that went to zero and were later reintroduced.
- These are investment-planning observations from the speaker, not personal financial advice.
Bottom Line
Long-run real stock returns are a useful planning anchor, but single-year results are almost always far from average, so a sound plan blends a reasoned forward estimate with a risk level you can hold through extremes.
