Banca de DEFESA: Gustavo Maia Rodrigues Gomes

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : Gustavo Maia Rodrigues Gomes
DATE: 22/11/2022
TIME: 14:30
LOCAL: https://teams.microsoft.com/l/meetup-join/19%3a9020f26afe5246b4acca186ddfea6536%40thread.tacv2/16678
TITLE:

“A Multi-Armed Bandit Framework for Portfolio Allocation”

KEY WORDS:

Portfolio Allocation. Multi-Armed Bandit Framework. Multifractal Model of Asset Returns. Times Series.

PAGES: 59
BIG AREA: Ciências Exatas e da Terra
AREA: Probabilidade e Estatística
SUMMARY:

For more than a century, the academic community has been studying the financial market in an attempt to understand its behavior in order to maximize profits. This project looks at ways to maximize profits from proposing a two phase procedure that we call MAB-MMAR . First, establishing individual generative models for each asset, to simulate, by Monte Carlo, future returns, using the Multifractal Model of Asset Returns (MMAR). Second, building a Multi-Armed Bandit (MAB) framework applying the Upper Confidence Bound (UCB)-Tuned Algorithm over the simulated paths to make choices between assets in order to optimize the allocation of resources. Also, as a protection layer for operations, we propose the Double-Barrier Method, where the operation is terminated if a lower barrier is touched. As a performance comparison, the One-Asset, 1/n, Modern Portfolio Theory (MPT) and the Axiomatic Second-order Stochastic Dominance Portfolio Theory (ASSDPT) models were tested. Our results are promising, where it is revealed that, in general, MAB-MMAR was the one that best performed in the most varied scenarios.

BANKING MEMBERS:
Externo à Instituição - ERALDO SERGIO BARBOSA DA SILVA - UFSC
Interno - 3000126 - JOSE AUGUSTO FIORUCCI
Presidente - 1171224 - RAUL YUKIHIRO MATSUSHITA
Externa à Instituição - REGINA CÉLIA BUENO DA FONSECA - IFGO