This paper shows that consumption-based asset pricing puzzles arise from using globally concave-shaped consumption utility. We empirically find that asset returns correlate negatively with many individuals' low-quantile consumption growth. This finding challenges most mainstream models and...

This paper proposes an estimable asset pricing model that builds upon micro consumption andreference-dependent preference. Central to the model is an S-shaped consumption utility function that is convex below the reference point. The model quantitatively accounts for both low risk-free rates and...

Inspired with the research, we present an enhanced Deep Neural Network (DNN) framework that combines Symbolic Genetic Algorithm (SGA) with Long-Short Term Memory (LSTM) in predicting the cross-sectional price returns using 245 fundamental indicators in China. The study addresses the challenges...