Factor Investing: Value, Quality, Momentum, and Beyond

Factor investing is an approach that groups stocks by shared, measurable characteristics such as valuation, quality, or momentum, based on academic and practitioner research showing these characteristics have historically explained differences in returns.

The value factor captures the tendency of stocks trading cheaply relative to fundamentals like earnings or book value to outperform more expensively priced stocks over long periods, though the premium is cyclical rather than constant.

Factor crowding occurs when a large share of investors position around the same factor signal, which can compress the factor's future premium and create sharp, correlated drawdowns when the crowded trade unwinds.

A factor is a characteristic associated with a persistent difference in returns across a broad universe, documented over long periods and across markets, with an explanation for why the difference should persist. A characteristic that correlates with returns in one dataset and has no rationale is a pattern rather than a factor. The distinction matters because patterns without explanations tend not to survive out of sample.

Most explanations for why factors work involve either compensation for a risk that occasionally materialises or a behavioural pattern that other participants can also exploit. Either explanation implies extended periods of underperformance: risk premia are earned by holding through the bad periods, and behavioural effects weaken as capital pursues them. A factor that never underperformed would attract enough capital to eliminate itself.

Each factor has several competing definitions, and value in particular has been measured with book value, earnings, cash flow, and sales, producing meaningfully different portfolios. Research findings are conditional on the definition used. Comparing a study's conclusion against a product's implementation requires checking whether they measure the same thing, which they frequently do not.

A concentrated portfolio built from a factor screen provides exposure but with far more single-name risk than the diversified portfolios the research describes, so the outcome is dominated by stock selection rather than by the factor. Achieving the exposure the research documents typically requires many positions and periodic rebalancing. Funds built for the purpose exist, and their construction rules deserve the same scrutiny as any active strategy.

Crowding describes capital concentrating in the same factor exposure, which can compress the expected premium and increase the risk of a sharp unwind when positions are exited together. Observable indications include valuation spreads within the factor narrowing, rising correlation among stocks sharing the exposure, and growth in assets tracking the factor. None of these is a timing signal, and each is visible only in aggregate.

They interact, sometimes substantially. Value and momentum have historically shown negative correlation, so holding both has smoothed results relative to either alone, while quality and low volatility overlap enough that combining them adds little diversification. A portfolio built by stacking screens can also end up with unintended concentrations, since several factors may select the same sector at the same time.