Statistical inferences for price staleness
提出非参数统计方法，检验高频资产价格零回报的独立同分布性及日内模式，实证发现NYSE十只股票零回报并非独立同分布，且日内变化不能仅由确定性模式解释。
This paper proposes a nonparametric theory for statistical inferences on zero returns of high-frequency asset prices. Using an infill asymptotic design, we derive limit theorems for the percentage of zero returns observed on a finite time interval and for other related quantities. Within this framework, we develop two nonparametric tests. First, we test whether intra-day zero returns are independent and identically distributed. Second, we test whether intra-day variation of the likelihood of occurrence of zero returns can be solely explained by a deterministic diurnal pattern. In an empirical application to ten representative stocks of the NYSE, we provide evidence that the null of independent and identically distributed intra-day zero returns can be conclusively rejected. We further find that a deterministic diurnal pattern is not sufficient to explain the intra-day variability of the distribution of zero returns.