What is the q-factor model and how does it differ from Fama-French five-factor?

Question

AcadiFi · Accepted Answer

The q-factor model of Hou, Xue, and Zhang (2015) is derived from the neoclassical q-theory of investment, not from empirical data mining. It has four factors: market (MKT), size (ME), investment (I/A), and profitability (ROE).

q-theory says firms invest until the marginal benefit equals the marginal cost, so cross-sectional differences in expected returns (discount rates) are linked to current investment and expected profitability. Firms with low expected ROE and high current investment (cheap capital) have low discount rates; firms with high ROE and low investment have high discount rates.

Differences from FF5:
- q-factor uses annual I/A based on total assets growth; FF5 CMA uses asset growth but constructed slightly differently.
- q-factor uses quarterly ROE (current profitability); FF5 RMW uses annual operating profitability divided by book equity.
- No value (HML) factor — the model claims HML is subsumed by investment and profitability once both are included.

In empirical horse races, the q-factor model explains anomalies roughly as well as FF5, with comparable redundancy of the value factor. Practitioners picking between them often choose based on software availability and academic reviewer preferences.

Example: Chesterfield Quant Fund regressed on both models yields β_I/A = -0.22, β_ROE = 0.31 in q-factor versus β_CMA = -0.18, β_RMW = 0.26 in FF5 — almost identical economic message with slightly different loadings.

What is the q-factor model and how does it differ from Fama-French five-factor?

Master Level III with our CFA Course

Related Questions

Practice Questions