Abstract
This paper examines tick size reduction in the U.S. 2-year Treasury Note futures market around January 14, 2019. We employ a difference-in-differences regression with 5-, 10-, and ultra 10-year Treasury futures markets as control groups, and we generally find significantly improved market quality. Trading costs reduce for liquidity takers and the smaller tick size has limited ability to relax its binding constraint. Our mechanism analyses indicate that traders significantly withdraw less liquidity after undercutting in response to non-execution risks and when they react to snipping risks at the opposite side of undercutting orders. Moreover, we find traders do not shift to market orders to acquire immediacy. Conversely, significantly fewer trades after undercutting. The smaller tick size significantly changes the quoting behavior, where depths are dispersed through the new pricing grid, suggesting that traders can post sub-optimal quotes and limit orders are layered in a more granular pricing grid, with limit order size significantly becoming smaller. Last, trading volume does not significantly change, while more trades occur with smaller trade size. Pricing efficiency does not significantly change neither does price discovery. Our findings suggest that reducing the tick size in tick-constrained liquid futures market could be promising.
