Hong Kong · HKUST Quantitative Finance ’30

Mika Han

Hello! I’m Mika, a first-year Quantitative Finance student at HKUST. Mathematics is what I enjoy most, and finance is where I want to use it.

FocusIB Mathematics

01 · IB Mathematics AA HL

Where it started.

IB Mathematics AA HL is where I got curious. Vectors, calculus and proof stopped being exercises and started explaining things: shapes in space, populations, money.

Every question on the right is from my final IB exams, in syllabus order.

1 / 12

02 · Research

Then, real problems.

Then I took the mathematics outside the classroom: a university research programme, a forecast for a game I speedrun, and the finances of our own stall.

Each animation is rebuilt from the numbers in my paper. The full papers are in Research below.

03 · HKUST Quantitative Finance

Now, markets.

At HKUST I’m applying mathematics to far more complex systems: stochastic processes, derivative pricing and risk.

dSₜ = μSₜ dt + σSₜ dWₜ

Geometric Brownian motion

01/ 03Scroll
Education

From Seoul to Hong Kong.

  1. 2019 – 2026

    Nord Anglia International School Hong Kong

    IB Diploma

    • Moved from Seoul to Hong Kong at eleven.
    • Learned English and Spanish.
    • Became curious about mathematics.
    Mathematics AA HL 7Chemistry HL 7Economics HL 7SAT Math 800TOEIC 970UKMT Senior GoldHK Chemistry Olympiad finalist
  2. 2026 – 2030

    The Hong Kong University of Science and Technology

    BSc Quantitative Finance

    • Scholarship recipient, HKUST (2026).
    • Member, Mathematics Students’ Club (MATHSC).
    • Bloomberg Global Trading Challenge 2026.
    • CME Group University Trading Challenge.
    • UK University Integration Bee.
Research · 2025

Research.

University of Hong Kong · Mathematics behind AI · 1st of 60+

Dimensionality optimisation and spectral thresholding

PCA derived from first principles with Lagrange multipliers, then random matrix theory (the Marchenko–Pastur law and the BBP phase transition) to decide which eigenvalues are signal and which are noise. Written with Bona Han.

Lagrange multipliersSpectral decompositionRandom matrix theory
Full paper · 9 pages1 / 9Open PDF ↗
Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9

Independent research

Forecasting the next speedrun world record

32 world records in Getting Over It: an exponential progression curve for record times, then exponential and Weibull models for the waits between records, fitted by maximum likelihood. The Weibull fit showed that a record gets harder to break the longer it stands, and the paper checks both forecasts against what actually happened.

Maximum likelihoodWeibullLikelihood ratio test
Full paper · 29 pages1 / 29Open PDF ↗
Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29

Economics research

Fiscal Policy & Externalities Analysis

How South Korea’s HP2030 taxes and advertising rules changed e-cigarette consumption, spending and second-hand exposure, measured with interrupted time series.

Interrupted time seriesSegmented regressionSurvey, n = 97
1 / 4
Male e-cigarette use before and after HP2030, with segmented regression
Demand curve estimated from survey responses
Welfare-loss diagram for a Pigouvian tax
E-cigarette use by age group, 2013 to 2020

EcoSphere · IB Mathematics AA HL

Profit maximisation modelling

Revenue and cost functions fitted to eleven weeks of sales, then a Lagrangian to find the most profitable mix of small and large pots under a resource constraint, confirmed with the bordered Hessian: about 61 small and 22 large pots, HK$1,125 a week.

RegressionLagrangianBordered Hessian
Full paper · 15 pages1 / 15Open PDF ↗
Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15
Positions

Positions.

Off the clock · Clash Royale

Six-time Ultimate Champion.

First reached at fourteen. I maintain it is serious training in real-time resource allocation under imperfect information.

×6Ultimate Champion
14Age at first finish
3.0Average elixir
Log Bait 3.0 deck: evolved Goblin Barrel, evolved Knight, Rocket, Princess, Goblin Gang, Ice Spirit, The Log and Cannon
Deck · Log Bait 3.0 cycle4 / 10 elixir
Contact

Get in touch.

hhan.mika@gmail.com

mhanal@connect.ust.hk