Lesson A5.5 · 5 of 5 · Factors and investment strategies
Factor investing strategy: rules, costs and how long bad spells last
A factor strategy only helps if you can hold on through the years when it lags, and there can be a lot of them. In August 2026 the value factor was still below its December 2006 peak. Rules written in advance, a spread of factors and low costs help you avoid quitting at the wrong moment.
Terms in this lesson
- Maximum drawdown
- The largest fall in price from a peak to the lowest point after it.
- Diversification
- You spread your money across several investments, so one mistake does not hurt too much.
- TER
- What an ETF or fund costs you each year, as a percentage of the money invested.
- Rebalancing
- You bring your portfolio back to its target percentages, selling what rose and buying what lagged.
How long the bad spells lasted: value, momentum and the market
Time below peak measures how long it takes from a high until the investment gets back to it. We worked it out from the monthly data in Kenneth French’s library, July 1963 to August 2026, for the US market with dividends and for the factors, as paper portfolios with no costs.
| July 1963 - August 2026 | Longest time below peak | Largest fall |
|---|---|---|
| Market, with dividends | 73 months (Aug 2000 - Sep 2006) | -50.3% (Oct 2007 - Feb 2009) |
| Value (HML) | 236 months and counting (since Dec 2006) | -57.8% (Dec 2006 - Sep 2020) |
| Momentum (Mom) | 213 months and counting (since Nov 2008) | -57.8% (Nov 2008 - Sep 2009) |
| Equal mix of HML, Mom, RMW | 168 months (Nov 2008 - Nov 2022) | -25.8% (Nov 2008 - Dec 2020) |
Our calculations from Kenneth French’s data library, version built from the August 2026 CRSP database. The factors are gaps between groups of shares, compounded monthly, before costs. The mix puts a third each in value, momentum and profitability, rebalanced monthly.
Three lines on a logarithmic scale, showing the year-end value of a notional 100 invested at the end of June 1963, from Kenneth French’s data library, before costs. Value (HML) climbs to 977 at the end of 2006, falls to 432 at the end of 2020 and stands at 664 in August 2026. Momentum climbs to 6,688 at the end of 2008, drops to 3,155 in 2009 and stands at 4,524 in August 2026. The equal mix of the two with profitability rises more smoothly, to 1,588 in 2008, falls to 1,199 in 2020, reaches 1,656 in 2022 and stands at 1,573 in August 2026.
The rest of this lesson is in the subscription
The subscription unlocks every chapter of Intermediate, Trading school, Investor psychology and Advanced, with their exercises and unlimited reviews.
Create a free accountWritten rules and spreading across factors
A strategy you can stick with starts on paper, before the first pound goes in. Write down your answers to a few questions and keep them in your decision journal. The guide to stock market risk helps you set your limits.
- Which factor and why: what evidence you read, over what period, before and after publication.
- How much of the portfolio: a cap you never exceed, so that 20 years below peak cannot upend your life.
- When you rebalance: say once a year on a fixed date, not according to how prices have moved.
- When you would quit: reasons unrelated to recent years, such as higher costs or a fund that has changed its rules, not “it lagged for three years”.
Spreading across factors helps because factors do not always fall together. In the French data the monthly correlation between value and momentum was -0.19 from 1963 to 2026, and MSCI notes that the two typically move in opposite directions. An equal mix of value, momentum and profitability averaged 4.57% a year with volatility of 6.25%, against 10.28% for value alone.
Costs and factor ETFs
Returns in research belong to paper portfolios: they buy the strong group, short the weak one and pay nothing. A factor ETF (often called “smart beta”) tracks an index built on a rule, such as shares with a low P/B or strong momentum. MSCI points out that a portfolio that only buys can overweight the top scorers but can barely underweight the weakest. So you get only part of the factor, plus nearly all of the market’s swings.
It comes off the return every year. Compare it with the charge on a whole-market fund.
Momentum changes holdings often; every change costs money, even if it does not show up in the TER.
The index definition differs from the research one, and the fund does not short-sell.
The psychology: why people quit a strategy at the wrong moment
Three bad years are normal for a factor. In the French data, 33.3% of all 36-month stretches from 1963 to 2026 were negative for value, 14.5% for momentum and 11.6% for the three-factor mix. The trouble is that after three weak years the strategy looks broken, and whatever did well looks like the obvious switch.
Three weak years seem to say everything about the next three. The recency lesson shows how money chases recent returns.
Lagging with a strategy you picked hurts more than lagging with an index. It pushes you to sell right at the low.
When others are making money elsewhere, the pull to follow them grows. Your decision journal shows what you believed beforehand.
Lessons are educational. They are not a recommendation to buy or sell. Investing involves risk and you can lose money. Sources: Kenneth R. French Data Library (Dartmouth), the “Fama/French 5 Factors (2x3)” and “Momentum Factor (Mom)” files, version of August 2026, built from the August 2026 CRSP database (monthly data July 1963 - August 2026), taken from an automated copy of the library files made on 3 October 2026; the time below peak, largest falls, equal mix rebalanced monthly, correlation and 36-month stretches are our own calculations, before costs; MSCI, “MSCI Factor Indexes” page (value and momentum typically move in opposite directions) and “The Long and Short of Factor Indexing”, 11 September 2026 (long-only portfolios), accessed 29 October 2026. The cost example is illustrative.