Factor investing, explained honestly

Factor investing sits between index investing and active management, with a substantial academic literature and an implementation record that is more mixed.

What a factor is

The concept.

A characteristic that has historically been associated with different returns across a large group of securities.

The most studied include: value, meaning cheaper securities relative to fundamentals; size, meaning smaller companies; momentum, meaning recent relative performance; quality, meaning profitability and stable earnings; and low volatility.

Each has decades of academic research behind it in various forms.

Why they might exist

Two families of explanation.

Risk-based: the factor represents a genuine additional risk for which investors are compensated.

Under this explanation the premium should persist, since the risk does not disappear.

Behavioural: the factor arises from systematic investor errors.

Under this explanation the premium may shrink as it becomes known and exploited.

Both explanations have support for different factors, and the distinction matters for whether the premium should be expected to continue.

The honest caveats

Which are substantial.

Data mining: with enough searching, patterns emerge in historical data that have no future validity, and the volume of published factors — hundreds have been proposed — suggests this is a real problem.

Out-of-sample performance has generally been weaker than in-sample, which is what data mining would predict.

Several well-known factors have underperformed for extended periods after publication.

Implementation costs — turnover, trading, capacity — reduce returns and are absent from academic studies.

And definitions vary between providers, meaning two funds targeting the same factor can hold quite different portfolios.

The duration problem

The practical obstacle.

Factor premiums are not consistent: they operate over long periods with substantial stretches of underperformance.

The value factor in particular underperformed for well over a decade in major markets, which is longer than most investors will hold anything.

Which means capturing a factor premium requires holding through periods long enough that most people conclude the strategy has stopped working.

And the temptation to abandon it near the point of maximum underperformance — which is exactly when a mean-reverting premium would be most attractive — is substantial.

Smart beta products

The retail implementation.

Funds tracking indices constructed on factor rules rather than market capitalisation.

Costs sit between index funds and active funds.

Quality varies enormously: some are thoughtful implementations of well-researched factors, and others are marketing applied to arbitrary screens.

Checking the methodology, the turnover, the cost and the actual holdings is necessary.

And multi-factor funds combining several factors reduce the risk of any one underperforming, at the cost of diluting each.

How to assess one

Practical questions.

Which factor, and how is it defined?

What is the economic rationale — risk-based or behavioural?

What is the cost premium over a plain index fund?

What is the turnover, which drives transaction costs?

How concentrated is the resulting portfolio?

What is the worst relative underperformance in the backtest, and could you hold through that?

And how much of the record is backtested rather than live, since backtests are constructed with knowledge of the outcome?

Where a small allocation is defensible

For balance.

An investor who understands the evidence, accepts that premiums may not persist, and commits to holding through a decade of underperformance may reasonably tilt a portion of a portfolio.

The conditions: a modest allocation; a well-implemented low-cost fund; a broad market core alongside; and a written commitment to the holding period.

Without the last condition, the likely outcome is abandoning it at the worst point, which produces a worse result than never having tilted.

What most investors should do

Stated plainly.

A broad market-cap-weighted global index fund captures the market return at minimal cost and requires no view on factors.

Which is a defensible position and is what a substantial share of the evidence supports for people who will not tolerate long relative underperformance.

Factor investing is not a mistake and it is not necessary, and treating it as necessary is where retail investors get into difficulty.

The general lesson

Beyond factors.

Any strategy with a historical premium requires holding through the periods when it does not work, and those periods are longer than backtests make them feel.

The gap between a strategy's returns and its investors' returns is widest for strategies requiring patience.

Which means the question is not whether a factor works but whether you would still be holding it after ten disappointing years.

General information only, not investment advice. Investments can fall in value and past performance does not indicate future returns. Consult a regulated financial adviser.