**-** **** **, #**, Long Island City, NY ***** Abigail Loh adk68u@r.postjobfree.com 646-***-****

EDUCATION

Columbia University New York, NY

M.A. in Mathematics of Finance (STEM) GPA: 3.6/4.0 Sep 2019 – Dec 2020

• Courses: Statistic Methods in Finance, Stochastic Process, Model & Trade Derivatives, Financial Engineering, Time Series, etc.

• Cross Registration: Machine Learning, Applied Data Science, Natural Language Processing (NLP), Algorithms

• Activity: Quant Analyst at Columbia Quant Finance Group Shanghai University of International Business and Economics Shanghai, China B.A. in Finance GPA: 3.9/4.0 Summa Cum Laude Sep 2015 – Jun 2019 SKILLS

Computing: Python, R, SQL, TensorFlow, Keras, Scikit-learn, NLTK, RapidMiner, MATLAB, LaTeX, Tableau, Power BI Project Github: https://github.com/AbigailLoh/Quant-Finance WORK EXPERIENCE

UnionPay International Singapore

Quant Strategy Analyst Oct 2018 – Aug 2019

• Developed a new risk analytic product (UPDIndex) by applying Principal Component Analysis (PCA) on credit card transaction data and selecting six statistic components; UPDIndex was later used as an in-house benchmark for credit risk management

• Scraped and analyzed alternative datasets with Python to track and predict default rate and customer traffic around ATMs

• Built regression models to estimate the budget for each marketing campaign and designed A/B tests to discover key factors that influence consumer behaviors, increasing marketing efficiency by 20%

• Created business intelligence reports based on model outputs and presented strategic investment feedback to senior strategists SinoLink Asset Management Shanghai, China

Model Research Intern Aug 2018 – Oct 2018

• Implemented equity vanilla options pricing through the Black Scholes model; conducted the benchmark test to compare the Black Scholes model validation with SinoLink in-house pricing system

• Compared the pricing results from different numerical methods in Black Scholes model; gave the internship final presentation on Black Scholes model validation results

• Built automation in VBA to construct policy portfolios with equities, fixed-income ETFs and gold as alpha benchmarks PwC Shanghai, China

Audit Data Intern Jul 2018 – Aug 2018

• Assisted in debugging and developing Aura (PwC in-house software), integrated and accelerated enterprise data flow

• Audited the 2018-3 Financial Bond Prospectus of $1 billion for Ford Automotive Finance (China) Limited PROJECTS

Portfolio Construction with Dynamic Hedging (Python) New York, NY EDHEC Research Lab, Mentors: Vijay Vaidyanathan, PhD and Lionel Martellini, PhD Dec 2020 – Feb 2021

• Designed and calibrated CPPI strategy with drawdown constraints to protect against systemic risk

• Generated Asset Returns with Geometric Brownian Motion (GBM), used CIR to model Drift (µ) and Volatility (σ) in Rates

• Back-tested and visualized CPPI algorithm by creating interactive chart with six parameters: n_scenarios, µ, σ, m, floor level and risk-free rate, and designed three methods to balance µ, σ, m in practice

• Applied Quadratic Programming on Cov Matrix to locate Efficient Frontier, Tangency Portfolio, plotted EW and GMV portfolios

• Asset Liability Management – analyzed funding ratio and matched durations with dynamic risk budgeting between PSP and LHP Design an Event-driven Back-testing System to Test a Mean-reverting Strategy (Python) New York, NY Capstone Project, Mentors: Siyuan Liu, PhD and James Ma Weiming Nov 2020 – Jan 2021

• Designed a back-testing engine that interacts with various modules components which handle tick data, fetching historical prices from Quandl, handling order and position updates, and simulating a streaming price feed that triggers z-score calculations

• Ran the engine by using GS historical data from 2015 to 2017, with buy/sell z-score thresholds of -1.5 and 1.5, gained annualized return of 15.9%, Sharpe ratio of 1.2 and maximum drawdown of 7.2%

• Performed multiple runs with varying parameters and determined the optimal buy/sell threshold tuple to be (-1.5, 1.5) Predict Time Series COVID-19 Deaths with Machine Learning (Python) New York, NY Columbia CS Department, Mentor: Sherif A. Tawfik Abbas Oct 2020 – Nov 2020

• Created and trained SARIMAX, Prophet, Neural Networks and XGBOOST models to predict future COVID-19 deaths

• Checked stationary with Dickey-Fuller test, iterated hyperparameters and optimized SARIMAX model with tuple p=2, q=1, d=3

• Calculated Mean Average Error (MAE) of all four models, XGBOOST has the least MAE of 392.15 and fits historical chart best INTERESTS

Interests: Performance-level Pianist, Diary, Ballerina, Handicraft, Diving, Meditation, Fashion Show, OOTD, Foodie

Contact this candidate