Vanessa Feng

Her work, 2024 to 2026

  1. Probability of default model Jan⁠–⁠May 2024

    300 to 10candidate variables cut to the final model

  2. Lima One Capital May⁠–⁠Aug 2025

    30⁠–⁠35%fewer false positives on the loan watchlist

  3. Morgan Stanley Aug⁠–⁠Sep 2025

    6model portfolios for new clients

  4. J.P. Morgan Jun⁠–⁠Aug 2026

    Incoming Analyst at J.P. Morgan in New York.

Works on credit and portfolio risk and builds her own models and tools. Data Science and Economics at UC San Diego, 2027.

55%

Kolmogorov–Smirnov separation in validation

Probability of default model

Project on a credit card portfolio, Jan⁠–⁠May 2024, San Diego

Logistic regression, weight of evidence and information value, vintage and roll-rate analysis

A logistic-regression model of credit card default. Vintage curves set a 6‑month performance window; roll-rate analysis tracked accounts within it.

  • Narrowed the candidate variables with break-and-heal binning, weight of evidence and information value, stepwise regression and cluster analysis.
  • Set warning and breach thresholds for ongoing monitoring.
KS
55%
PSI
13%
MAPE
10%

30⁠–⁠35%

fewer false positives on the loan watchlist

Lima One Capital

Credit Risk Intern, May⁠–⁠Aug 2025, Greenville, SC

Probability-of-default thresholds, recall and ROC‑AUC, Datarails

Piped the probability‑of‑default model into the watchlist as a loan filter, saving about 10 hours of manual review a month.

  • Tested 8 model thresholds against recall, precision, accuracy and ROC‑AUC; chose the one that kept recall above 0.90 with the other metrics at or above firm benchmarks.
  • Built Datarails dashboards that segment about 300 flagged loans a month by product, geography, vintage and sponsor.
  • Restructured the credit approval memo to quantify risk assessments; approval meetings ran about 50% faster.

50+

client portfolios her private-fund assessments supported

Morgan Stanley

Wealth Management Intern, Portfolio Risk Management, Aug⁠–⁠Sep 2025, New York

Monte Carlo simulation, Excel forecasting, fund risk and suitability

Rebalanced asset class, sector and market-cap allocations to each client’s risk profile, liquidity needs and existing exposure; tested long-term goals with Monte Carlo simulation.

  • Designed duration- and credit-appropriate fixed income strategies for 10+ client portfolio reviews with the financial planning team.
  • Screened and modelled 10+ US large- and mid-cap stocks; built the Excel forecasts through 2028 that the team used for tactical positioning of $100MM+ in client assets.
  • Assessed 5 private market funds in infrastructure, real estate and private credit for risk and client suitability, supporting asset allocation across 50+ client portfolios.
Portrait of Vanessa Feng

J.P. Morgan

Asset Management Summer Analyst, Jun⁠–⁠Aug 2026, New York

Returns full-time as an Analyst in New York.

Spent the summer in asset management risk.

Education and experience

Education

UC San Diego

B.S. Data Science and B.A. Economics, 2023⁠–⁠2027. GPA 3.75.

Econometrics, Financial Accounting, Financial Statement Modeling, Data Modeling, Machine Learning & Application, Data Management

Earlier experience

Rappo

VC Analyst Intern, Tech networking and vendor platform, Jun⁠–⁠Sep 2024, San Francisco

Designed the 10+ slide pitch deck Rappo used with 200+ investors to raise its first round.

  • Built a database of 500+ target investors with Python and web scraping, with a risk-appetite scorecard for fit that improved targeting accuracy 50%+.
  • Refined go-to-market from investor feedback; a new one-pager lifted the cold-email success rate 10%+ as Rappo won 100+ clients and $6,000+ in revenue in two months.
Certifications
  • Securities Industry Essentials (SIE)FINRA, 2026
  • Analyzing Financial ReportsWall Street Prep, 2025
  • Excel Skills for BusinessGoldman Sachs, 2024
  • Bloomberg Finance FundamentalsBloomberg
Award

The President’s Volunteer Service Award

AmeriCorps, 2021

Languages

English (native), Chinese (native), French (elementary)

Tools

Python, SQL, Java, Snowflake, Tableau, Datarails, Excel, PowerPoint