ƽÌØÎå²»ÖÐ

Event

PhD Thesis Defense Presentation: Qing Xu

Monday, December 9, 2024 10:00to12:00

Ms. Qing Xu, a doctoral student at ƽÌØÎå²»ÖÐ in the Finance area will be presenting her thesis defense entitled:

Three Essays on Recovering Subjective Beliefs and Risk Preferences in a Non-Parametrical Approach

Monday, December 9, 2024, at 10:00 a.m.

(The defense will be conducted in Hybrid mode)

Ìý

Qing Xu

Student Committee Chair: Professor Anisha Ghosh and Professor Laurent Barras.

Please note that the Defence will be conducted in Hybrid mode. Only the student and their committee members will participate in the presentation.


Abstract:

This thesis comprises three essays that utilize a non-parametric approach to recover subjective beliefs and risk preferences from asset prices, shedding light on the behavior of heterogeneous investors across varying market conditions and forecast horizons.

The first essay separately identifies various investor types’ risk preferences and beliefs based on their portfolio holdings. It documents differing levels of risk aversion and belief patterns among investors who primarily invest in different types of portfolios. This study further explores the identity of these investors, finding that large and value investors, who share similar risk preferences and beliefs, are predominantly institutional investors, whereas small-growth investors consist of a higher proportion of professional individuals.

The second essay investigates the impact of downturn experiences on investors’ beliefs and risk preferences. The estimation results show that investors who have weathered market downturns and remained in the stock market exhibit greater pessimism and risk tolerance compared to those who have never faced downturns. The hypothesis is that investors with lower risk tolerance tend to exit the market following downturns, thus concealing their preferences and beliefs from market prices. An analysis of survey data supports this argument.

The third essay introduces a novel approach to empirically recover the conditional pricing kernels using the survey data. The survey-implied pricing kernels exhibit a tilde shape in the short term (1-month), gradually transition to a U-shape for longer horizons, and form a checkmark shape for long-term (10-year) forecasts. These results suggest that investors consistently place a high value on cash flows from future high-return states, a behavior that is more pronounced during adverse market conditions.

Back to top