Misspecification and Weak Identification Zhaoguo Zhan
OVERVIEW
A common practice across disciplines is to infer conclusions from a theoretical model using empirical data. However, two inherent problems jeopardize its usefulness. First, models are only approximations of reality, and their analysis typically suffers from the so-called misspecification problem, or biased coefficients, error terms, and/or parameter estimates. Second, even if a model is theoretically correct, researchers may not have sufficient empirical data to infer its structural parameters. This problem is termed weak identification. Our project focuses on the asset-pricing literature to develop novel econometric methods to jointly address both theoretical and empirical questions related to misspecification and weak identification.
46 | Summer Research Fellowship
Made with FlippingBook - Online catalogs