Consistent Noisy Independent Component Analysis

Jean-Marc Robin (1) , Stéphane Bonhomme (2)

1 ECON - Département d'économie (Sciences Po)
2 CEMFI

We study linear factor models under the assumptions that factors are mutually independent and independent of errors, and errors can be correlated to some extent. Under factor non-Gaussianity, second to fourth-order moments are shown to yield full identification of the matrix of factor loadings. We develop a simple algorithm to estimate the matrix of factor loadings from these moments. We run Monte Carlo simulations and apply our methodology to British data on cognitive test scores.

Domaines: Economies et finances

Jean-Marc Robin, Stéphane Bonhomme. Consistent Noisy Independent Component Analysis. Journal of Applied Econometrics, 2009, pp.1-45. ⟨10.1016/j.jeconom.2008.12.019⟩. ⟨hal-01022621⟩