math.STNov 3, 2026

Estimating "Realized" Skewness using Convolutional Neural Network

Authors: Haozhe Jiang, Ostap Okhrin, Michael Rockinger

Organizations: Technische Universität Dresden. “Friedrich List” Faculty of Transportation. Chair of Applied Statistics. · Technische Universität Dresden. “Friedrich List” Faculty of Transportation. Chair of Applied Statistics. Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI) Dresden/Leipzig, Germany. · University of Lausanne. Faculty of Business and Economics, CH 1015 Lausanne, Switzerland.

Abstract

We propose a new estimator of low-frequency skewness (monthly) that exploits high-frequency data (5 min) through a direct functional mapping consisting of layers of Convolutional Neural Networks followed by Multi Layer Perceptrons. Monte Carlo experiments involving stochastic volatility models, with and without jumps, show that the estimator reduces finite-sample bias relative to existing realized-skewness estimators and remains accurate under model misspecification. Empirically, our estimator exhibits temporal stability and delivers cross-sectional performance in actual skewness-sorted portfolios. We characterize the finite-sample bias of conventional plug-in skewness estimators and establish risk consistency properties of the proposed NN estimator under standard regularity conditions.

Figures & tables

Explore similar work

CardsList
  1. Bridging Structured Knowledge and Data: A Unified Framework with Finance Applications

    Apr 1, 2026Yi Cao, Zexun Chen, Lin William Cong +1Parameter EstimationQuantitative Finance

  2. Nonparametric GARCH: An Iterated Neural Estimator

    Nov 3, 2026Ruizhi Deng, Guohao Shen, Ngai Hang ChanTime Series ModelingParameter Estimation

  3. End-to-End Neural Shrinkage of Indefinite Pairwise Correlation Matrices for Small-Cap-Inclusive Portfolios

    Aug 31, 2026Christian Bongiorno, Lorenzo VillasseroQuantitative FinancePortfolio Optimization