Intraday News Event Study: From Traditional to Causal Inference Methods

Explore intraday real time news events’ impact on trading risk management, understand different impact from different event types, as well as event types’ point estimate impact and cumulative impact on individual stocks’ intraday excess returns, a.k.a., idiosyncratic risk. The topic will go over traditional methods like event group vs. control group using parametric and non -parametric group comparison statistical methods, as well as causal based methodologies that create impact predictions from counterfactual time series generation. Two causal based methodologies to be discussed include Synthetic Control ( a CausalLens model) and Bayesian based Causal Impact. The purpose of the study is to evaluate the immediate and cumulative impact as well as decay of intraday news event on stock’s idiosyncratic moves and therefore, develop better intraday market making risk management.