Savvy Nickel
by Michael W. Covel
Michael Covel's comprehensive case for trend following — the trading strategy that buys assets trending up and sells assets trending down, regardless of fundamentals. Backed by decades of live performance data from the world's most consistently profitable traders.
*Disclosure: This article contains affiliate links. If you purchase through these links, we may earn a commission at no additional cost to you. We only recommend books we genuinely believe in.
Michael Covel has spent his career studying and writing about trend following — the strategy of buying assets in uptrends and shorting assets in downtrends, using systematic rules with no fundamental analysis. His book makes the most comprehensive case for trend following as an investment strategy, backed by decades of performance data from the world's best-performing CTAs (commodity trading advisors). For serious traders and investors who want to understand an alternative to both fundamental investing and buy-and-hold passive strategies, this is required reading.
| Attribute | Details |
|---|---|
| Title | Trend Following (5th Edition) |
| Author | Michael W. Covel |
| Publisher | FT Press |
| First Published | 2004 |
| Current Edition | 5th edition, 2017 |
| Pages | 480 |
| Reading Level | Intermediate |
| Amazon Rating | 4.3/5 stars |
Hardcover: Buy on Amazon
Kindle: Buy on Amazon
Michael Covel is the founder of TurtleTrader.com and has spent 20+ years researching trend following traders and their strategies. He has conducted extensive interviews with successful CTAs and written multiple books including The Complete TurtleTrader (about the Turtle experiment) and Trend Commandments. He is the most dedicated popularizer of trend following as a strategy and has done more than anyone to make CTA performance data accessible to the public.
Trend following is a trading strategy with three core rules:
What trend following is NOT:
| Not Trend Following | Trend Following |
|---|---|
| Fundamental analysis | Technical/price-based |
| Buy-and-hold | Active; both long and short |
| Market timing (predicting direction) | Trend-reactive (follow, not predict) |
| Concentrated positions | Broadly diversified |
| Stop-loss optional | Stop-loss mandatory |
The defining characteristic: trend followers never predict where markets will go. They react to where markets are going and stay with the move until evidence of reversal appears.
Covel's most important contribution is assembling decades of verified performance data from the world's top trend following CTAs.
Representative trend following CTA performance (backtested + live, through 2016):
| Firm | Annual Return (net of fees) | Sharpe Ratio | Max Drawdown |
|---|---|---|---|
| John W. Henry (JWH) | ~18-20% | 0.65-0.75 | -30 to -40% |
| Campbell & Company | ~14-16% | 0.55-0.65 | -25 to -35% |
| Man AHL | ~12-15% | 0.55-0.65 | -20 to -30% |
| Millburn Ridgefield | ~15-18% | 0.60-0.70 | -25 to -35% |
Note: These are historical figures. Trend following has become more crowded since 2000, and recent returns have been lower. The strategy still has value but expected returns have likely declined.
The crisis alpha property:
Trend following's most valuable characteristic: it tends to perform best during financial crises, when buy-and-hold portfolios suffer their worst losses.
| Crisis Period | S&P 500 Return | Representative CTA Return |
|---|---|---|
| 1987 Crash (Oct) | -21.8% | +200%+ (shorting bonds and stocks) |
| 1990 Gulf War | -14.8% | +20%+ |
| 2000-2002 Bear Market | -47.4% | +40-60% |
| 2008 Financial Crisis | -37.0% | +18-25% |
| 2020 COVID (March) | -34.0% | +5-15% |
The negative correlation with equity crises makes trend following a natural portfolio diversifier — it gains when traditional portfolios lose. This "crisis alpha" is one of the primary justifications for allocating to CTAs in institutional portfolios.
Trend followers enter positions when price breaks to new highs or lows over a specified lookback period. The most common:
Donchian Channel Breakout:
Buy signal: Price closes above the highest close of the past N days (typically 20-55 days)
Sell signal: Price closes below the lowest close of the past N daysMoving Average Crossover:
Buy signal: Short-term MA (e.g., 50-day) crosses above long-term MA (e.g., 200-day)
Sell signal: Short-term MA crosses below long-term MABoth methods catch trends after they have begun — trend followers never try to pick tops or bottoms. They accept missing the first portion of every move in exchange for confirmed trend direction.
Exits are typically trailing stops:
Long position exit: Price falls below the lowest close of the past N days (10-20 days)This allows profits to run (the exit moves up as the trend continues) while cutting losses quickly (if the trend reverses immediately after entry, the loss is small).
Trend followers size positions based on market volatility (ATR — Average True Range):
Position Size = Account Risk per Trade / (ATR × Dollar Value per Point)This means:
Example:
| Market | ATR | Account Risk | Position Size |
|---|---|---|---|
| Gold (volatile) | $20/day | $1,000 | 50 ounces |
| Treasury bonds (less volatile) | $0.50/day | $1,000 | 2,000 units |
| Corn | $0.05/bushel/day | $1,000 | 20,000 bushels |
Each position risks approximately $1,000 on a one-ATR move, creating consistent risk exposure across very different markets.
Richard Dennis and William Eckhardt debated whether great traders were born or trained. In 1983, they ran an experiment: select 23 individuals with no trading experience and train them in a specific trend-following system. If they could follow the rules, could they succeed?
The Turtles (so named because Dennis said he would "grow traders like turtles") generated approximately $175 million over the following years, with several going on to extremely successful independent trading careers.
The Turtle rules (simplified):
| Rule | Specification |
|---|---|
| Entry | Breakout above 20-day or 55-day high (long) or below 20-day or 55-day low (short) |
| Exit | Breakout in opposite direction over 10-day or 20-day lookback |
| Position sizing | 1 ATR = 1% account risk |
| Maximum units | 4 units per market, 6 units per correlated market group |
| Stop loss | 2 ATR from entry |
The key finding:
The system's rules were publicly available. The critical variable was whether individual Turtles could execute the system with discipline — taking every signal, not overriding the system when it felt wrong, maintaining position sizes during drawdowns.
Those who followed the rules closely achieved excellent results. Those who let discretion override the rules generally underperformed. This validated both Eckhardt's position (great traders can be trained) and the broader point that execution discipline matters more than analytical brilliance.
Trend following is intellectually straightforward but psychologically demanding:
Trend following systems regularly experience 20-40% peak-to-trough drawdowns. During these periods (which can last 1-3 years), the strategy feels broken. Most discretionary overlay destroys the system's long-run performance precisely during these drawdowns — when the trader "helps" by overriding signals or reducing size.
Historical drawdown durations:
| Drawdown | Duration Before Recovery |
|---|---|
| 1990 (JWH) | ~18 months |
| 1993-94 | ~24 months |
| 2009-2012 | ~36 months |
| 2013-2016 | ~24 months |
Three years of underperformance is psychologically devastating for most investors, who assume the strategy has stopped working and exit just before recovery.
Trend following typically wins only 35-45% of trades. Most individual trades lose small amounts. A small percentage of winning trades generate large profits. The psychological difficulty: most humans experience a steady stream of small losses (the 55-65% of losing trades) and must stay disciplined until the large winning trend arrives.
The typical trend following trade distribution:
| Outcome | Frequency | P&L Contribution |
|---|---|---|
| Small loss | 40% of trades | -$500 each |
| Tiny loss | 20% of trades | -$100 each |
| Breakeven | 5% | $0 |
| Small win | 20% | +$500 each |
| Large win | 15% | +$5,000+ each |
The 15% of large-winning trades generates all the profitability. Miss those by exiting early (because the position "has already moved a lot") and the strategy fails.
When trend followers are riding a large trend, their positions become obviously right in hindsight. But before the big move, they entered a breakout that looked like "buying at the top" to outside observers. Trend followers must accept being criticized for their entries and resist the pressure to take profits too early.
Covel's case for institutional and individual investors:
The 60/40 portfolio with trend following CTA:
| Portfolio | Annual Return | Standard Deviation | Sharpe Ratio | Max Drawdown |
|---|---|---|---|---|
| 60% S&P / 40% Bonds | 8.5% | 10.2% | 0.83 | -30.0% |
| 50% S&P / 40% Bonds / 10% CTA | 8.6% | 9.4% | 0.91 | -26.5% |
| 50% S&P / 30% Bonds / 20% CTA | 8.7% | 8.8% | 0.99 | -22.0% |
Adding a trend following allocation improves the Sharpe ratio and reduces maximum drawdown while maintaining returns. The crisis alpha property is why this works — CTAs gain when traditional portfolios lose their most.
Available vehicles for retail investors:
| Vehicle | Examples | Notes |
|---|---|---|
| Managed futures mutual funds | AQR Managed Futures (AQMIX), Equinox (MFBTX) | Accessible, daily liquidity, higher fees |
| ETFs | KMLM, DBMF, CTA | Increasingly available, lower cost |
| Direct CTA allocation | Requires accredited investor status, $100K+ minimum | Best access but high minimums |
Q: Can individual investors implement trend following?
A: A simplified version — using moving average crossovers with diversified ETFs — can be implemented individually. The crisis alpha and long-run return profile are partially achievable. Full implementation with 50+ markets requires institutional infrastructure.
Q: Has trend following stopped working?
A: Since 2010, trend following has underperformed its pre-2010 historical record. The strategy still generates positive returns with low correlation to equities, but expected annual returns appear to have declined to 5-10% net from historical 15-20%. More competition and faster markets may have eroded some of the edge.
Q: What percentage of a portfolio should be in trend following?
A: Institutional allocations typically range from 10-20%. For individual investors, a managed futures ETF representing 5-15% of the portfolio captures meaningful diversification benefit without excessive complexity.
Rating: 4.3/5
Trend Following is the most comprehensive treatment of systematic trend following available. Its performance data, Turtle experiment account, and crisis alpha analysis are each uniquely valuable. The promotional tone and repetition hold it back from a higher rating, but the core content is genuinely important for serious traders and institutional investors.
Hardcover: Buy on Amazon
Kindle: Buy on Amazon
Prices current as of publication date. Free shipping available with Prime.
by Jack D. Schwager
Jack Schwager's follow-up to Market Wizards features a new generation of trading legends — including Bill Lipschutz, Mark Ritchie, and Stanley Druckenmiller. Essential reading for serious traders who want to understand the full range of strategies and psychological disciplines that produce elite performance.
by Jack D. Schwager
Jack Schwager interviews the greatest traders of the 1980s and distills what separates consistent winners from the rest. A masterclass in trading psychology, risk management, and the diverse approaches that produce extraordinary long-term results.
by Mark Douglas
Mark Douglas's masterwork on trading psychology. The definitive guide to developing the disciplined, probabilistic mindset that separates consistent winners from traders who blow up their accounts despite having technically sound strategies.
Most people accept the first offer they get. That single decision can cost them hundreds of thousands of dollars over a career. Here's how to negotiate — and why the stakes are higher than they appear.
The ability to wait - to choose a larger reward later over a smaller one now - is the single most consistent predictor of financial outcomes. Here's the science, and how to actually build this skill.
Fractional shares let you buy a slice of any stock or ETF regardless of its price. Here is how they work, which brokerages offer them, and when they actually matter for your portfolio.