Ivan Hung

Education

Aug 2025 —
Aug 2026

MSc Finance — Investment and Wealth Management

Imperial College Business School, London

Financial Econometrics · Machine Learning · Derivatives · Asset Allocation & Investment Strategies · Applied Quantitative Macro Strategies

Sep 2022 —
Jun 2025

BSc Economics and Finance

University of East Anglia, Norwich

Ranked 1st of 250 students in first and final year · Brightspark Scholarship (£3,000) · Distinguished Performance Award, both years · Alternative Investments (86%) · Corporate Finance (83%)

Professional Experience

03 roles

Jun 2026 —
Aug 2026

Wealth Management Intern — International Team, APAC

UBS AG

London, UK

Return offer · 2027 Graduate Program

Produced a presentation on Sukuks and built three risk-profiled model portfolios using Solver, closing client knowledge gaps to support portfolio discussions
Analyzed internal advisory portfolios, identified cash allocation gaps, and drove a 7% performance improvement through reallocation
Built an automated prospecting tool using LLM prompt engineering to identify and screen potential clients, bringing £10M in Net New Money
Created quarterly performance decks for client advisors, highlighting forward-looking risk exposures and key KPIs

Jun 2025 —
Aug 2025

Data Science Intern

NatWest Group plc (RoosterMoney)

London, UK

Return offer · 2026 Graduate Program

Applied machine learning models (logistic regression, XGBoost) via scikit-learn to mitigate customer churn risk, reducing exit rates by 17%
Used prompt engineering and an agentic workflow to deploy a generative AI-powered chatbot, cutting response time by 20%
Led an end-to-end churn analysis using SQL (Snowflake/AWS) and Tableau, identifying a 46% higher churn rate after a first declined transaction

Jun 2024 —
Sep 2024

Wealth Management Intern

Chow Tai Fook (CTF) Life

Kowloon Bay, HK

Collaborated with a team of four on due diligence for Allianz, addressing regulatory and operational risks to pitch an investment deck with 32% ROI
Applied financial planning and pension product knowledge to help close a HK$450,000+ premium deal
Spearheaded a team of 10 to curate a portfolio website showcasing the firm's financial offerings, increasing traffic by 12%

Research Projects

05 studies

PYTHONQuant Finance

Statistical Arbitrage in HK Equities: A Cointegration & Kalman Filter Strategy

Developed a cointegration-based statistical arbitrage strategy using Engle-Granger and ECM frameworks on HK equities across tech and commodities
Implemented dynamic hedge ratios via Kalman filtering to capture time-varying relationships between assets
Constructed systematic trading signals using spread z-scores with entry/exit thresholds and position management logic to reduce false-signal exposure

PYTHONDerivs Trading

Rolling Call Strategy on Copper Futures: Options Pricing & Regime Analysis

Backtested a rolling 12-month ATM call strategy on copper futures using Black (1976) pricing with realistic roll mechanics and transaction costs
Sub-sample analysis revealed strong supply-shock hedging properties — 0.86 Sharpe and -10% max drawdown during the 2016 China supply cycle, vs. 0.24 full-sample
Derived annualized volatility inputs from 36-month rolling log-return standard deviations, isolating regime-dependent convexity as the key performance driver

PYTHONQuant Finance

An Ensemble Approach to the Overnight Equity Anomaly

Constructed a survivorship-free, point-in-time backtest over the S&P 1500 (2010–2024) with 50+ cross-sectional features, eliminating look-ahead bias
Developed a walk-forward ensemble (Huber, LightGBM, IC-weighted composite) yielding significant predictability — Spearman IC 2.1%, t-stat 6.7
Built a costed dollar-neutral long/short book with realistic costs and liquidity caps, isolating turnover as the binding constraint on net returns

PYTHONQuant Macro

Cross-Asset Carry: Currencies, Commodities & Government Bonds

Developed a systematic cross-asset carry strategy across FX, commodities and bonds using dollar-neutral long-short ranking with inverse-volatility weighting
Validated signal quality via information coefficients, quintile sorts and panel regressions, blending sleeves to deliver a Sharpe of 0.82 at 0.06 equity beta
Stress-tested robustness through parameter sensitivity, signal-lag and transaction costs, finding a conservative out-of-sample Sharpe of 0.45
Presented findings to lecturers and co-authored the final report as part of a 6-person team, translating results for a non-technical audience

PYTHONQuant Finance/Econometrics

Oil Futures Spread Trading Strategy: Forecasting and Economic Performance

Top grade in cohort

Developed and evaluated forecasting models for WTI crude oil futures spreads (CL2–CL1), exploiting term structure dynamics to generate trading signals
Implemented ARIMA and ARIMAX models with rolling out-of-sample backtesting, demonstrating the robustness of parsimonious models in volatile regimes
Translated forecasts into a systematic spread strategy, achieving higher risk-adjusted returns (Sharpe > 1), ~63% hit rate, and lower drawdowns

Hackathons

02 entries

PYTHONAlgo Trading / AI Agents

Eventus — Options Alpha Agents (Alpaca AI Trading Agents Hackathon)

Built solo, in one week (competed against 400+ teams): an autonomous options-trading agent that reads Alpaca through its MCP server and writes orders through the Alpaca CLI, with a deterministic 15-check risk engine as the only component permitted to approve a trade
Ran three uncorrelated strategy books on one paper account — a short-volatility iron condor carry book, a VWAP-triggered intraday momentum book, and an earnings-driven directional book gated on confirmed report dates rather than indicator thresholds
Backtested over a Feb–Aug 2026 replay generating ~8,600 trade ideas and 72,753 logged rejections, and audited the book's own results to find a credit-to-width accounting flaw inflating headline P&L — then reported the gated, corrected figure instead
Live dashboard: eventus-algo.streamlit.app

WEB APPFinTech / EdTech

Stackt — Financial Literacy for Gen Z (Work in Fintech AI Summit Hackathon)

Selected as 1 of 130 participants from an applicant pool of over 1,300 for this full-day, in-person hackathon running 8:30am–6pm — my first in-person hackathon
Developed the app with my team over the two weeks leading up to the event, while working full-time at UBS
Built a Duolingo-style learning app tackling low financial literacy among 17–21 year olds, turning macro, ISAs/savings, investing and scams into short interactive lessons
Designed a judgment-based mechanic: users read a real financial scenario alongside competing accounts of it and must identify the credible one, with every fabrication corrected on screen immediately
Live deployment: stackt-pi.vercel.app

Builds

Actively building and developing software… if you need another pair of hands contact me :)

Coming Soon…

Skills

Finance & Risk

Asset Allocation Stress Testing Scenario Analysis Model Risk Management Macro Finance Stochastic Calculus Financial Modelling

Programming & Data

Python R SQL Excel VBA Microsoft Office Stata Tableau AWS SageMaker

Analytics & Modeling

Data Visualization Dashboard Development XGBoost scikit-learn PyTorch

Tools & Platforms

Bloomberg Terminal GitHub Confluence Jira VS Code

Certifications & Languages

In Progress

CFA Level I FIMC — Wall Street Prep Data Scientist Associate — DataCamp

Languages

English — Native Cantonese — Fluent Mandarin — Intermediate

Activities Outside of Work

Padel Golf Bouldering Calisthenics Gym Taekwondo Piano — ABRSM Grade 8 (Merit) Drums