Model-free time-aggregated predictions for econometric datasets

TL;DR

This paper improves the NoVaS method for predicting squared log-returns in financial data, demonstrating enhanced long-term and short-term prediction accuracy over existing methods like GARCH(1,1).

Contribution

A new, more parsimonious NoVaS variant is developed, offering better performance for time-aggregated financial predictions.

Findings

Outperforms existing NoVaS in long-term predictions

Surpasses GARCH(1,1) in accuracy

Effective for both short- and long-term forecasts

Abstract

This article explores the existing normalizing and variance-stabilizing (NoVaS) method on predicting squared log-returns of financial data. First, we explore the robustness of the existing NoVaS method for long-term time-aggregated predictions. Then we develop a more parsimonious variant of the existing method. With systematic justification and extensive data analysis, our new method shows better performance than current NoVaS and standard GARCH(1,1) methods on both short- and long-term time-aggregated predictions.