Shifts in beta and the TARP announcement

• Published in: Finance research letters 2022-06, Vol.47, p.102704, Article 102704
• Main Authors: Phin, Andrew, Prono, Todd, Reeves, Jonathan J., Saxena, Konark
• Format: Article
• Language: English
• Subjects: Event study; Intra-day returns; Realized beta; Systematic risk
• ISSN: 1544-6123; 1544-6131
• DOI: 10.1016/j.frl.2022.102704

Using high frequency return data, we propose a method for detecting level shifts in stock betas in the context of an event study. Using this method, we identify significant beta changes around the initial announcement of the Troubled Asset Relief Program (TARP). Our findings have important implications for studies of TARP that assume betas remain unchanged following the program’s announcement. Pre-TARP betas are found to be poor measures of post-TARP systematic risk exposure. Holding these betas fixed at pre-TARP levels in the estimation of cumulative abnormal returns (CARs) risks distorting the magnitudes of these CARs by multiple orders. •Using high frequency data, we propose a method for detecting level shifts in beta around an announcement, in an event study context.•Our method measures the degree to which the announcement leads to a reduction in systematic risk exposure of the risky assets (in this case, financial stocks) affected by the announcement.•We find that the TARP announcement led to a significant reduction in the betas of financial stocks (specifically, bank stocks).•Ignoring these reductions in betas when estimating cumulative abnormal returns (CARs) for these bank stocks leads, in turn, to an overstatement of CARs by multiple orders of magnitude.•Our methodology extends to event studies, more broadly, when systematic risk exposure may also be changing.