Cram\'er type moderate deviations for intermediate trimmed means

Abstract

In this article we establish Cram\'er type moderate deviation results for (intermediate) trimmed means $T_n=n^{-1} \sum_{i=k_n+1}^{n-m_n}X_{i:n}$, where $X_{i:n}$ -- the order statistics corresponding to the first $n$ observations of a~sequence $X_1,X_2,\dots $ of i.i.d random variables with $df$ $F$. We consider two cases of intermediate and heavy trimming. In the former case, when $\max(\alpha_n,\beta_n)\to 0$ ($\alpha_n=k_n/n$, $\beta_n=m_n/n$) and $\min(k_n,m_n)\to\infty$ as $n\to\infty$, we obtain our results under a~natural moment condition and a~mild condition on the rate at which $\alpha_n$ and $\beta_n$ tend to zero. In the latter case we do not impose any moment conditions on $F$, instead, we require some smoothness of $F^{-1}$ in an~open set containing the limit points of the trimming sequences $\alpha_n$, $1-\beta_n$.