Discriminatory versus uniform auctions under non-competitive auction: Evidence from Japan

Abstract

As Ausubel et al. (2014) describe, the general revenue ranking of uniform and discriminatory auctions is an empirical question. However, although many empirical studies investigate whether discriminatory auctions outperform uniform auctions in government bond markets, the empirical results are mixed. Pycia and Woodward (2021) suggest that the uncertainty of aggregate supply could reconcile the mixed result by focusing on non-competitive auctions. To address this research question, we take advantage of the Japanese government bond (JGB) auction because (1) we can obtain the variations from a uniform and discriminatory auction, (2) it issues 10–30% of a JGBs via a non-competitive auction, creating the potential uncertainty of aggregate supply, and (3) the data of When Issue market is available. Our empirical results show that discriminatory auctions tend to outperform uniform auctions and that discriminatory performance tends to be better when the share of a non-competitive auction is higher. Our results empirically support the predictions of Pycia and Woodward (2021).

Keywords: Discriminatory Auction; Uniform Auction; Markup; When-Issued Markets; Japanese Government Bond

JEL Classification: C57; D82; G18; G28