Brisbane, AU · BMath ’27 · The University of Queensland

Maxwell Hanks

Hi! I'm a second-year Bachelor of Mathematics (Mathematical AI + Computer Science) student at UQ. I enjoy building projects across machine learning, statistics, and finance, and read the wonderful literature and research behind them for fun.

Reading: Geiss - Measure, Probability & Functional Analysis

See the projects Download resume

3 / 124

IMC Trading AlgoJam ’26 - adaptive multi-model time-series trading

6.25 / 7.00

cumulative GPA (83 WAM)

250+ hrs

teaching mathematics to high-school and university students

Projects

Selected work

Projects across machine learning, statistics, and finance. Code on GitHub.

2026 · Pythonregime-adaptive forecasting

IMC x UQ FinTech AlgoJam 3 - regime-adaptive multi-strategy forecasting

Designed instrument-specific time-series models spanning peg mean-reversion, Holt's ETS trend-following, volatility-gated regime-switching, and seasonal decomposition against a fitted annual cycle. All positions routed through a budget-constrained sizer with priority-ranked trimming.

3 / 124

final placing

instruments traded

ETS

framework

Holt's ETSregime-switchingseasonal decompositionminority game

2026 · Pythonstatistical arbitrage

Susquehanna Algothon ’26 - adaptive multi-model time-series trading

Selected cointegrated pairs offline via Engle–Granger and ADF stationarity testing, modelled mean-reverting spreads as Ornstein–Uhlenbeck processes, and hedged residual exposure against the index. Volatility-scaled sizing; validated by walk-forward backtesting on unseen data.

161 / 394

final placing

instruments traded

spreads

cointegrationADF / Engle–Grangerwalk-forwardvol targeting

2026 · Python, MATLAB

Optimal execution via HJB & neural networks

Formulated liquidation as continuous-time stochastic control, solved the linear-quadratic benchmark analytically, and extended to nonlinear market impact with neural-network approximation. (Still a work in progress :D)

ε-opt

validated vs dynamic programming

stochastic controlHJBneural nets

2026 · Python, NumPy, Pandas

Barra-style multi-factor equity risk model

Daily factor returns via cross-sectional WLS; EWMA/Newey–West factor covariance with PSD enforcement and shrunk specific risk. Euler risk decomposition and walk-forward volatility-forecast validation.

stock universe

style factors

factor modelscovariance estimationrisk decomposition

About

About myself

I’m studying a Bachelor of Mathematics at UQ, majoring in Mathematical Artificial Intelligence with a minor in Computer Science. My coursework centres on statistical machine learning, probability & stochastic processes, software engineering, and linear algebra & optimisation.

Outside lectures I design and build projects that bridge data science, machine learning, and mathematical modelling, and I keep a running stack of maths journals and books on the go.

Aside from my studies in mathematics, I have a passion for music performance and composition, including playing piano and percussion in pit bands, symphony orchestras, and other ensembles!

Education

Bachelor of Mathematics

The University of Queensland — Mathematical Artificial Intelligence (Major), Computer Science (Minor)

2025 – 2027

Algorithms & Data StructuresDeep LearningObject-Oriented ProgrammingStatistical Inference

Experience

Teaching the maths I love

Connect Presenter

Connect Education - Brisbane, QLD

2026 – now

Deliver targeted exam-preparation workshops, using IA/EA data to tailor content and close knowledge gaps.
Guide students through past exam questions to build exam technique aligned with QCAA criteria.