Hello everyone,

I have applied Fama-Macbeth cross-sectional regression on Fama and French five-factor model (2014). On the left-hand side are the portfolio returns for sixteen size - B/M portfolios. On the right-hand-side are five the factors i.e. market, size, investment, and profitability factor.

The issue I am facing is that in 16-time series regressions, my intercepts were all positive and significant. Factor loadings for each factor were also strong and significant. However, when I applied t cross-section regressions (3439 days), the intercept of cross-section regression turned out to be negative and insignificant. Also, the risk premiums are very small (down to 4 decimal points).

Kindly help me make sense out of this.