Projects and competitions

From research to practice

Projects that start from papers and industry ideas, modeled in Python and documented by the league itself, plus our teams' results in university challenges.

Competitions

League teams in university quantitative finance challenges.

Top 10
2026
Desafio Quant AI 2026
Itaú Asset Management
Alpha HelixEl Dorado
Itaú Asset Management · 2026
Desafio Quant AI 2026
The 7th edition of Itaú Asset's university challenge, where students build investment robots with generative AI. Out of more than 1,300 groups, two league teams made the top 10. Mentored by Murillo Toda.
Alpha HelixTop 10
Members: Vitor Guardiano, Natalie Kaori, Enrico Poggi
Strategy: Builds alpha from the “DNA” of Brazilian funds' portfolios: it creates embeddings of assets and investors, measures fund–asset similarity and derives each asset's popularity among similar funds. It invests in assets that stayed persistently popular in the recent past, on the hypothesis that popularity reflects peer consensus and demand.
El DoradoTop 10
Members: Antônio Freitas, Luisa Obara, Pedro Berta
Strategy: Hunts the risk premium in commodities from physical inventories: every week it compares each of ten markets' inventory with its last 156 weeks and ranks them by “tightness”. The tighter the inventory, the higher the risk of a shortage and, since commodities are hard to substitute, the higher the premium for carrying that risk.
Former league members
Augur · 2nd place · Hugo Villanova, Eduardo Zorzi
Organizer's announcement(opens in a new tab)
3rd place
2025
Desafio Safra Top Gestor 2025
Banco Safra
FGV Quant
Banco Safra · 2025
Desafio Safra Top Gestor 2025
An asset management challenge with over 280 registered groups. The team ran a model-based approach: macro scenario, systematic selection of stocks and BDRs, and risk-focused portfolio construction.
FGV Quant3rd place
Members: Antonio Paiva, Bruno Fogaça, Hugo Villanova, Thiago Lebrão
1st place
2024
Datathon de Moedas Digitais (5ª edição)
FGV EESP
Crypto X Quant
FGV EESP · 2024
Datathon de Moedas Digitais (5ª edição)
A data challenge on real problems in the crypto asset and digital currency ecosystem, with 20 teams. The team joined members of FGV Quant and FGV Crypto.
Crypto X Quant1st place
Members: André Togni, Bernardo Rubin, Francisco Zipman, Leonardo Ono
5th place
2024
Desafio Quant Itaú Asset Management 2024
Itaú Asset Management
FGV Quant
Itaú Asset Management · 2024
Desafio Quant Itaú Asset Management 2024
Over 500 registered teams. As a finalist, the team presented Itaú's portfolio managers a strategy combining deep reinforcement learning (SAC, PPO and A2C) and NLP to build sentiment-adjusted portfolios. Mentored by André Maranhão.
FGV Quant5th place
Members: Juan Puzzo, Matheus Moreira Espindola, Henrique Motta, Ricardo Paschoa

Projects

By semester: research with partner asset managers, the league's simulated fund and each trainee cycle's standout projects.

Simulated fundOngoing since 2025.2

FGV Quant Capital

The league's simulated fund, running since July 2025: it simulates a quant asset manager's routine with real data, internally validated strategies (Pairs Trading and Momentum) and a full Python stack.

Current team

Current team: Antônio Freitas (lead), João Henrique Cunha, Carolina Yumi, Valentina Fantozzi, Maria Lucia Lucatelli

See FGV Quant Capital

2026.2In progress
- Project with Avantgarde Asset Management
  An ongoing project this semester; its topic and results will be published at the end of the cycle.
  Members: Enrique Ernandes (lead), Rebeca Lavalle, Luiz Henrique Lopez, Maria Júlia Bongiovanni
- Project with Vault Capital
  An ongoing project this semester; its topic and results will be published at the end of the cycle.
  Members: Pedro Berta (lead), Bianca Campos, Helena Minns, Bernardo Suplicy
- Project with B3
  An ongoing project this semester; its topic and results will be published at the end of the cycle.
  Members: Marina Silva (lead), Fernando Suguimoto, Maria Eduarda Boone, João Vitor Henrique
- Independent project
  An ongoing project this semester; its topic and results will be published at the end of the cycle.
  Members: Rodrigo Ciofi (lead), Enrico Poggi, João Menezes, Kirghean Pacheco
2026.1
- With Avantgarde Asset Management
  How many dimensions does the zoo have? Information redundancy among return characteristics
  For IBRA stocks from 2010 to 2024, PCA and hierarchical clustering over 45 candidate factor-zoo characteristics show that about 11 to 12 genuinely distinct dimensions suffice: 12 components explain 82% of the variance and 11 clusters emerge from the dendrogram.
  Members: Calebe Howang, Pedro Berta, Vitor Guardiano
  Download paper (PDF)
- With Avantgarde Asset Management
  Fundamental volatility in Brazilian equities
  Building on Huang (2009), it tests whether the historical volatility of operating cash flow predicts returns in the IBrA. The long-short between extreme deciles returned 90.5% cumulatively from 2012 to 2022 after costs and borrow fees: above the Ibovespa, below the Selic.
  Members: Carolina Yumi, João Menezes
  Download paper (PDF)
2025.2
- With Avantgarde Asset Management
  Allocation-focused regimes and dynamic factor investing in Brazilian equities
  A replication and adaptation, for Brazilian equities (2000–2024), of a method that defines regimes as periods in which one factor persistently beats another: a model with a jump penalty labels the sample and an XGBoost on financial and macro features forecasts the regime out of sample to drive allocation.
  Members: Arthur Formiga, Kevin Lima, Luisa Obara, Luis Poliseli, Rodrigo Kandir, Théo Ogier
  Download paper (PDF)
- The Momentum factor on B3
  A study of price-trend persistence tested on real B3 data: the factor's logic, portfolio formation and results.
2024.2
- With Avantgarde Asset Management
  Pairs trading in Brazilian utilities
  Pairs trading on utility stocks, with a five-year backtest and pair selection by Engle-Granger cointegration, augmented Dickey-Fuller and Euclidean distance. Presented to the firm in March 2025.
  Members: Hugo Villanova, Marcelo Bettarello, Antonio Paiva, João Schmitt, Rosa Zhu
  Download paper (PDF)
2024.1
- With Constância Investimentos
  A predictive model for interest rates
  A semester project in partnership with Constância Investimentos.
  Members: Gabriel Daud, Diego Natel, Paula Pécora, Ellen Kim, Célika Matsui
- With Constância Investimentos
  AI and event studies for trading
  A semester project in partnership with Constância Investimentos.
  Members: Henrique Motta, Juan Puzzo, Paula Maluf, Henri Boteon