Title: Testing for spurious factor analysis on high dimensional nonstationary time series Authors:  Yi He - University of Amsterdam (Netherlands) [presenting]
Bo Zhang - University of Science and Technology of China (China)
Abstract: Spurious factor behaviors arise in large random matrices with high-rank random signal components and heavy-tailed spectral distributions. The aim is to establish analytical probabilistic limits and a distribution theory for these spurious behaviors in high-dimensional non-stationary time series. We transform scree plots into Hill plots to detect spectral patterns in these spurious factor models and develop max-t tests to distinguish between spurious and genuine factor models. Simulations confirm the excellent size and power performance of our test in finite samples. Applying the tests to three real-life datasets, we detected spurious factors in both economic and climate data, and genuine factors in finance data.