Start with the question
Every model serves a concrete investment decision.
Toronto · Quantitative Finance · U of T
Charting markets.
Testing assumptions.
I’m Mile Xue, a Mathematics and Statistics student who turns noisy data into clear, decision-ready investment research.
- 7
- research builds
- 20K
- Monte Carlo paths
- SAS
- certified specialist
01 · The navigator
A clear read in uncertain waters.
I study Mathematics and Statistics at the University of Toronto and build practical research across public markets, private-equity replication, derivatives, portfolio risk, and early-stage investing.
My process is simple: define the decision, test the assumptions, quantify the risk, and write the honest conclusion—even when the answer is “pass.”
Stress the story
Scenario analysis exposes what smooth averages can hide.
Make the call
Research ends with a concise recommendation and its limits.
02 · Captain’s log
Education and experience
2025—2029
Education
University of Toronto
Honours Bachelor of Science · Mathematics and Statistics
Coursework includes economics, Python, linear algebra, calculus, and statistics.
Jun—Aug 2026
Experience
Financial Data Analyst Intern
Quantum Finance · Remote
Researched funds, stocks, and ETFs; translated factor models, scenario analyses, and risk-return trade-offs into concise investment memos.
Sep 2026
Certification
SAS Certified Specialist
Base Programming Using SAS 9.4
03 · The voyage map
Selected research
Two research deep dives, followed by five further explorations in markets, models, and investment decisions.
01 / Internship · Project 2Quantitative research
What hides behind
a smooth return?
Private Equity Replication & Desmoothing
Can listed equities replicate private-equity exposure—and how much risk does appraisal smoothing conceal?
1.18Market-factor beta
96.7%Five-factor model R²
1.64×Volatility after desmoothing
Same series. A different risk picture.
The downside becomes harder to ignore.
Line: 10th–90th percentile · Dot: median
60% world equity / 40% PE proxy · Simulated, not realized returns
Explore the factor exposures & methodology
Market+1.177
Size · SMB+0.214
Value · HML−0.382
Profitability · RMW+0.275
Investment · CMA−0.081
Fit excess returns to five equity factors (36 observations), apply a first-order desmoothing correction, then compare risk and simulated one-year portfolio outcomes. The fitted exposures suggest strong public-equity sensitivity and a growth tilt. R² describes the full five-factor model, not the market factor alone.
Research takeaway
Smooth reported returns can obscure risk. A liquid replication strategy still needs to be understood through its public-equity exposures.
Source: Project 2 exported factor_model.csv, risk_comparison.csv and scenarios.csv. Research illustration: the project documentation identifies the private-equity benchmark as placeholder data; these outputs are not verified live fund performance.
02 / Internship · Project 3Investment due diligence
A compelling product.
A disciplined pass.
Seed-Stage Deal Review · Anonymized case
Strong early traction is only the start. Does the evidence support the growth plan, entry price, and ownership required?
01 · Traction02 · Valuation03 · OwnershipPass
Evidence lens / Traction
Separate live revenue from ambition.
Live ARR→Committed ARR→Forecast ARR
The review distinguishes revenue already live from customer commitments and the future growth target. A large shop pipeline supports the opportunity, but conversion and higher revenue per shop still have to materialize.
What would change the view?
Evidence that the pipeline converts into paying shops and that expansion revenue supports the planned ramp.
Investment judgment / Pass
Early traction and product fit were promising. The ambitious revenue ramp, demanding entry valuation, and difficulty securing the required ownership and board position outweighed the opportunity.
Based on the completed Project 3 deal review. Company and investor identities, transaction figures, and source documents are withheld; this view presents the analytical framework and the review's conclusion.
03
Convertible Arbitrage & Delta Hedging
Combined bond-floor valuation, equity-option analysis, and a delta-hedged simulation to screen convertible securities for volatility mispricing.
- 20K
- paths
- 126
- trading days
04
Defined Outcome Pricing
Recreated buffered ETF payoff structures with listed options and measured the trade-off between downside protection, financing, and upside caps.
- 4
- payoff structures
- 0.25–0.90
- cap gap
05
Antifragile Barbell Analysis
Compared growth, drawdown, convexity, and probability of ruin across four portfolio structures, including sensitivity to return smoothing and block length.
- 180
- monthly observations
- 0.55%
- modeled ruin risk
06
AI Power Pair Trading
Screened power-sector pairs for mean reversion, controlled for multiple testing, and rejected the strategy after costs when the evidence did not hold.
- 24
- pairs screened
- Pass
- final decision
07
U.S. Housing Price Prediction
Cleaned macroeconomic data, selected features, and iteratively refined a linear model in Google Sheets using MAE and MSE for evaluation.
- <10%
- prediction error
- MAE/MSE
- evaluation
04 · Ship’s toolkit
Methods for the journey
Code & data
- Python · pandas · NumPy
- SAS 9.4
- Excel · Google Sheets
- Data cleaning & EDA
Quantitative methods
- Linear regression
- Fama-French factor models
- Monte Carlo simulation
- Desmoothing & scenarios
Investment analysis
- Public & private equity
- ETF & options analysis
- Seed-stage valuation
- ARR multiples · MOIC
Beyond the models
English · IELTS 7.0
Oil painting and sketching sharpen the same habits I use in research: observation, patience, and composition.
05 · Next destination
Have a question worth navigating?
I’m always interested in thoughtful conversations about markets, models, and investment decisions.
monkeydmile@gmail.comToronto, Ontario, Canada