Purpose--This paper aims to examine the relationship between the conditional variance of the factors from the Fama-French three-factor model and macroeconomic risk, where macroeconomic risk is proxied by the conditional variance for a default risk premium and real gross domestic product (GDP) growth. Design/methodology/approach--A generalised autoregressive conditional heteroscedastic model is used to generate the conditional volatilities and bivariate Granger causality tests are used to examine the empirical relationship between the risk measures. Findings--Past values of the conditional variance for a default risk premium have information that is precedent to the conditional volatility for value premium and the small stock risk premium, and the conditional variance for the market risk premium has information about the future volatility of macroeconomic risk, as proxied by the conditional variance for GDP growth. Research limitations/implications--The implications are that conditional volatility associated with default is related to current and future volatility in value premium; however, volatility associated with the market risk premium appears to be a predictor of future macroeconomic risk. A caveat is that the results are dependent on the proxies used for macroeconomic risk and more refined measures of macroeconomic risk may yield different results. Practical implications--This paper suggests that examination of the relationship between the volatility of macroeconomic factors and the explanatory factors in asset-pricing models will help to further understanding of the relationship between risk and expected return. Originality/value--This paper focuses directly on the links between risk associated with the Fama-French factors and macroeconomic risk. This added knowledge is beneficial to practitioners and academics whose interest lies in asset price modelling. Keywords Autoregressive processes, Value analysis, Premium pricing, Stock returns, Risk assessment Paper type Research paper