Identifying the Effects of Chinese Treasury Purchases Using High-Frequency Data

Abstract

I use a new measure of surprise foreign official Treasury purchases, high-frequency data, and identification by heteroskedasticity to estimate the effect of Chinese official purchases of US Treasury securities on Treasury yields. Over the past decade, foreign official institutions have purchased Treasuries at an unprecedented rate. While existing studies suggest these purchases should depress yields, much of the empirical literature suffers from endogeneity bias. Applying a new identification technique, I overcome this bias and accordingly estimate larger impacts of purchases on yields. My results indicate that Chinese official interventions lowered Treasury yields in the mid-2000s by perhaps 100 basis points.

Keywords: foreign Treasury purchases, capital flows, event study, identification by heteroskedasticity

JEL Classification: E43, E58, F31, F42