Moment Component Analysis: An Illustration With International Stock Markets

By Eric Jondeau, Emmanuel Jurczenko, Michael Rockinger. 2018

52 citations recorded by the source corpus.

We describe a statistical technique, which we call Moment Component Analysis (MCA), that extends principal component analysis (PCA) to higher co-moments such as co-skewness and co-kurtosis. This method allows us to identify the factors that drive co-skewness and co-kurtosis structures across a large set of series. We illustrate MCA using 44 international stock markets sampled at weekly frequency from 1994 to 2014. We find that both the co-skewness and the co-kurtosis structures can be summarized with a small number of factors. Using a rolling window approach, we show that these co-moments convey useful information about market returns, for systemic risk measurement and portfolio allocation, complementary to the information extracted from a standard PCA or from an independent component analysis.

Subfields: Economics and Econometrics, Finance, Management Science and Operations Research

Citations in the corpus

• A literature survey of low‐rank tensor approximation techniques (2013, neutral): “…• computational tasks in micromagnetics [80,81]; • rational approximation problems [242]; • computational homogenization [42,96,176]; • computational finance [134,137]; • the approximation of stationary states of stochastic automata networks [38]; • multivariate regression and machine learning [25].”