Learning about noise

This paper studies how acquisition of non-fundamental information (learning about noise) affects financial markets. We develop a rational expectations model with investors who are endowed with fundamental and non-fundamental information of heterogeneous quality and who optimally allocate learning capacity between fundamentals and noise. We demonstrate that learning about noise increases price informativeness, and the price can be the most informative when the majority of investors acquire non-fundamental information. We also find that i) investors whose prior fundamental information is relatively precise (imprecise) compared to their prior non-fundamental information learn only about fundamentals (noise) and ii) learning about fundamentals (noise) increases (decreases) the heterogeneity in the fundamental information quality across investors.