Moment set selection for the SMM using simple machine learning

Kód výsledku v IS VaVaI

RIV/67985556:_____/23:00574253 - isvavai.cz

Nalezeny alternativní kódy

RIV/00216208:11230/23:10466898

Výsledek na webu

https://www.sciencedirect.com/science/article/pii/S0167268123001944?via%3Dihub

DOI - Digital Object Identifier

10.1016/j.jebo.2023.05.040

Jazyk výsledku

angličtina

Název v původním jazyce

Moment set selection for the SMM using simple machine learning

Popis výsledku v původním jazyce

This paper addresses the moment selection issue of the simulated method of moments, an estimation technique commonly applied to intractable agent-based models. We develop a simple machine learning extension reducing arbitrariness and automating the moment choice. Two algorithms are proposed: backward stepwise moment elimination and forward stepwise moment selection. The methodology is tested using simulations on a Markov-switching multifractal framework and two popular financial agent-based models with increasing complexity. We find that both algorithms can identify multiple moment sets that outperform all benchmark sets. Moreover, we achieve considerable in-sample estimation precision gains of up to 66 percent for agent-based models. Finally, an out-of-sample empirical exercise with S&P 500 data strongly supports the practical applicability of our methodology as the estimated models pass the validity test of overidentifying restrictions.

Název v anglickém jazyce

Moment set selection for the SMM using simple machine learning

Popis výsledku anglicky

This paper addresses the moment selection issue of the simulated method of moments, an estimation technique commonly applied to intractable agent-based models. We develop a simple machine learning extension reducing arbitrariness and automating the moment choice. Two algorithms are proposed: backward stepwise moment elimination and forward stepwise moment selection. The methodology is tested using simulations on a Markov-switching multifractal framework and two popular financial agent-based models with increasing complexity. We find that both algorithms can identify multiple moment sets that outperform all benchmark sets. Moreover, we achieve considerable in-sample estimation precision gains of up to 66 percent for agent-based models. Finally, an out-of-sample empirical exercise with S&P 500 data strongly supports the practical applicability of our methodology as the estimated models pass the validity test of overidentifying restrictions.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

GA20-14817S: Propojení finančních a ekonomických multiagentních modelů: ekonometrický přístup

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Journal of Economic Behavior & Organization

ISSN

0167-2681

e-ISSN

1879-1751

Svazek periodika

212

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

26

Strana od-do

366-391

Kód UT WoS článku

001021137800001

EID výsledku v databázi Scopus

2-s2.0-85161338880