SPEAKER： 任玥璇（新加坡管理大学）

TIME/DATE：2024.12.24   12:00-13:00

CLASSROOM：A505室

TENCENT：952710081

PW：676496

LINK：https://meeting.tencent.com/dm/MNA27j8OfnvD

ABSTRACT

We propose using a conformal predictive analysis for high-frequency event studies. Unlike existing literature, we recast the inference problem of cumulative abnormal return (CAR) as a counterfactual prediction problem for cumulative return. The general continuous-time model for spot regression can be approximated by a linear regression model with independent and stable-distributed random variables under the fixed-$k$ asymptotic setting, thereby establishing the asymptotic validity of the conformal prediction interval. Extending the theory to incorporate a counterfactual model with many control units, the proposed prediction interval remains valid when using the synthetic control estimator. An intraday event study of AMD’s conference session illustrates the empirical application.

GUEST BIO

Yuexuan Ren is a job market candidate from Singapore Management University. Her research interest is econometrics theory with a specialization in financial econometrics. She holds a master’s degree in Statistics from Renmin University of China and a bachelor’s degree in Economics from Shandong University.