Trend Strategy That Beat Markets for 100 Years and Cut Risk

High-level summary

The video reviews an academic/systematic trend‑following strategy applied to industry groups. According to the referenced paper, the approach outperformed buy‑and‑hold over ~100 years (1926–present) while materially reducing drawdowns. The core idea is to rotate exposure across US industry portfolios (or sector ETFs) by following trend breakouts/breakdowns and moving to cash/T‑bills when industries aren’t trending.

Core concept: hold only industries/ETFs that are in an uptrend (trend breakout), size positions by recent volatility, and move to cash (1‑month T‑bills) when no trend exists.

---

Assets, instruments and sectors mentioned

• 48 US industry portfolios (historical backtest universe)
• 31 sector ETFs from State Street Global Advisers (modern tradable implementation)
• S&P 500 (benchmark)
• One‑month Treasury bills (cash / safe asset)
• General categories referenced: stocks, sector ETFs, managed futures / hedge‑fund‑style trend followers
• No specific tickers were named in the subtitles

---

Methodology — step‑by‑step framework

1. Universe
   • Use either the 48 historical US industry portfolios (backtest) or map the model to 31 State Street sector ETFs for a tradable implementation.
2. Trend indicators
   • Breakout‑based channels were used (subtitles said “Donian and Kelner channels” — likely Donchian and Keltner channels).
   • Entry rule: go long when price breaks above the 20‑day upper channel.
   • Exit rule: close the position when price breaks below a lower channel (40‑day lookback used for the lower channel to bias toward staying long).
3. Trailing stop
   • Use the maximum of the lower bands from the Donchian and Keltner channels as the trailing stop.
   • The trailing stop is never moved downward (locks in profits, limits losses).
4. Position sizing
   • Weights are inversely proportional to each industry’s recent 14‑day volatility (higher volatility → smaller weight).
5. Cash allocation
   • When an industry is not in an uptrend, the system moves to cash / 1‑month T‑bills rather than holding the whole market.
6. Practical implementation notes
   • The century‑long industry model was mapped to 31 State Street sector ETFs and the same rules were applied, with trading costs accounted for.

---

Risk management specifics

• Volatility‑adjusted position sizing (inverse of 14‑day volatility).
• Trailing stop combining Donchian & Keltner lower bands; stop never trailed downward.
• Strategy reduces overall market exposure by holding only the strongest trending industries/ETFs (reported lower beta ≈ 0.40 in ETF test).
• Defensive posture: moves to cash / T‑bills when no clear trend is present.

---

Key numbers, timelines and performance metrics

• Backtest timeline: 1926 through the present (~100 years).
• Long‑run industry‑portfolio results (historical backtest)
  • Average annual return: ~18.2%
  • Benchmark (market) average return referenced: ~9.7% annually
  • Sharpe ratio: ~1.3–1.4
  • Drawdowns: approximately 60% smaller than buy‑and‑hold market drawdowns
• Modern tradable ETF implementation (31 State Street sector ETFs; includes trading costs)
  • Trading cost assumed: 0.0035 per share (subtitle value)
  • Sharpe ratio: 0.61 (strategy) vs 0.59 (market)
  • Annualized alpha: +2.7% above the market
  • Max drawdown: 24% (strategy) vs 55% (S&P 500 during the period shown)
  • Beta: 0.40 (lower market exposure)

---

Implementation / practical notes called out

• The paper’s authors mapped the century‑long theoretical backtest to a tradable ETF implementation to show feasibility in real markets.
• Trading costs were included in the ETF test.
• The ETF implementation frequently held only a subset of the strongest ETFs, which lowers overall market exposure and beta.
• The strategy is fully rules‑based and reactive (it does not attempt to forecast tops or bottoms; it uses momentum / channel signals).

---

Explicit recommendations or cautions from the video

• The video presents the approach as a simpler alternative to stock‑picking or forecasting and prompts viewers to consider whether they would trade such a system.
• The supplied subtitles did not include an explicit legal disclaimer or “not financial advice” statement.

---

Potential subtitle / naming errors to note

• Subtitles used “Donian and Kelner channels” and authors “Carlos Xeratini and Gary Anton” — these are likely misspellings:
  • “Donian and Kelner” → likely Donchian and Keltner channels
  • Author names likely intended as Carlos Zeratini and Gary Antinacci (spelling in subtitles may be incorrect)

---

Sources / presenters (as named in the subtitles)

• QuantifiedStrategies.com (video/channel)
• Paper referenced (subtitle): “A Century of Profitable Industry Trends” by Carlos Xeratini and Gary Anton (as shown in subtitles)