˙⋆ data science × economics ⟡˙

Vanessa
Feng

i build models that keep risk afloat.

data scientist & economist drifting through credit, portfolio & quantitative risk

see my worksay hi 🐚

psst, press b for bubbles ⟡

⋆˙⟡ a little about me

hello from the tide pool

Portrait of Vanessa Feng

i’m a data science & economics student at uc san diego who likes turning messy financial data into something calm and legible. most of my work lives where statistics meets markets: probability-of-default models, portfolio rebalancing, and the quiet plumbing that helps risk teams sleep at night.

i’ve interned across credit and wealth-management risk (morgan stanley, lima one capital, and rappo) and i’m headed to j.p. morgan’s asset & wealth management risk team next summer. i care about models that are accurate *and* explainable, dashboards people actually open, and a tidy threshold that catches the right loans.

fun facts

  • trilingual: en · 中文 · fr
  • SIE & bloomberg certified
  • monte-carlo enthusiast
  • my favorite animal is the shark 🦈
  • i love art & drawing 🎨
⋆˙⟡ where i'm studying

education

University of California, San Diego

San Diego, CA · Expected Jun 2027

B.S. Data Science & B.A. Economics

GPA 3.75 / 4.0

relevant coursework

  • Econometrics
  • Financial Accounting
  • Financial Statement Modeling
  • Data Modeling
  • Machine Learning & Application
  • Data Management
⋆˙⟡ where i've worked

a string of pearls

  1. J.P. Morgan

    incoming ⟡

    Asset & Wealth Management Risk Summer Analyst

    New York, NY · Jun 2026 – Aug 2026

    joining the asset & wealth management risk team next summer.

  2. Morgan Stanley

    Wealth Management Intern, Portfolio Risk Management

    New York, NY · Aug 2025 – Sep 2025

    • Tailored model portfolios for 6 new clients by rebalancing asset class, sector, and market-cap allocations to align with individual risk profiles, liquidity needs, and existing exposures; used Monte Carlo simulations to help clients hit long-term financial goals.
    • Collaborated with the financial planning team on 10+ client portfolio reviews to refresh investment objectives; designed duration- and credit-appropriate fixed-income strategies aligned with the macro-rate outlook.
    • Screened and modeled 10+ US large/mid-cap stocks against macro themes and 3-year revenue & EPS trends; built Excel forecasts through 2028 used to tactically over/underweight $100MM+ of client assets.
    • Dissected 5 private-market funds across infrastructure, real estate, and private credit to determine risk profiles and client suitability, supporting asset-allocation strategy across 50+ portfolios.

  3. Lima One Capital

    Credit Risk Intern

    Greenville, SC · May 2025 – Aug 2025

    Private lender, backed by MFA Financial.

    • Built and shipped a loan-filter system that piped the Probability-of-Default (PD) model into the watchlist, cutting false positives 30–35% and saving ~10 hrs/month of manual review for Credit & Servicing.
    • Ran sensitivity analysis across 8 PD-model thresholds (tracking recall, precision, accuracy, and ROC-AUC) to find an optimal threshold with recall above 0.90 while holding key metrics at or above firm benchmarks.
    • Built Datarails dashboards segmenting ~300 PD-flagged loans/month by product, geography, vintage, and sponsor profile to surface risk concentrations and guide underwriting updates.
    • Restructured the internal Credit Approval Memo to quantify risk assessments and reasoning, cutting loan-approval meeting time by ~50%.

  4. Rappo

    VC Analyst Intern

    San Francisco, CA · Jun 2024 – Sep 2024

    Tech networking & vendor-solutions platform.

    • Refined GTM strategy from investor feedback and built a one-pager that lifted cold-email success 10%+, helping Rappo acquire 100+ clients and generate $6,000+ in revenue within 2 months.
    • Extracted keywords and linguistic patterns from 50+ VC blogs with the OpenAI GPT API and built a content template that hit 5,000+ impressions, 12% conversion, and 34% click-through.
    • Built an internal database of 500+ target investors using Python and web scraping, plus a risk-appetite scorecard that improved targeting accuracy 50%+.
    • Designed a 10+ slide pitch deck communicating the business model, helping Rappo pitch 200+ investors and secure first-round funding.
⋆˙⟡ things i've built

treasures from the reef

Jan 2024 – May 2024 · San Diego, CA

Probability of Default Predictive Model

An end-to-end credit-risk model scoring default risk for credit-card holders and shaping mitigation strategy.

55%
KS
13%
PSI
10%
MAPE
300→10
features
  • Logistic Regression
  • Python
  • Credit Risk
  • WoE / IV
  • Vintage Analysis
  • Model Validation
  • Built a logistic-regression PD model to assess default risk across credit-card holders and design mitigation strategies.
  • Ran vintage analysis with delinquency-stabilization curves (first 30/60/90 DPD) to find the maturation period and set a 6-month performance window.
  • Performed roll-rate analysis using EVER (ever-delinquent) and EOP (end-of-period balance), computing forward/backward roll rates to track delinquency movement.
  • Reduced features from 300 → 10 via Break & Heal, Weight-of-Evidence / Information Value, stepwise regression, and cluster analysis.
  • Validated discriminatory power, stability, and accuracy (55% KS, 13% PSI, 10% MAPE) and set warning/breach monitoring thresholds.
⋆˙⟡ my little toolkit

shells in my collection

technical

  • Python
  • Java
  • SQL
  • Tableau
  • Snowflake
  • Excel
  • PowerPoint

finance & risk

  • Credit Risk
  • Portfolio Risk
  • Monte Carlo Simulation
  • Financial Modeling
  • PD Modeling
  • Sensitivity Analysis

certificates

  • Securities Industry Essentials (SIE)
  • Bloomberg Finance Fundamentals
  • Goldman Sachs Excel Skills for Business
  • Financial Reports Analysis

languages

  • English (native)
  • Chinese (native)
  • French (intermediate)
⋆˙⟡ proud little moments

highlights & honors

  • Securities Industry Essentials (SIE) certified
  • Bloomberg Finance Fundamentals certified
  • Cut PD-model false positives 30–35%, saving ~10 hrs/month of manual review
  • Tuned a PD threshold to recall > 0.90 while holding firm-benchmark metrics
  • Forecasts informed tactical positioning of $100MM+ in client assets
  • Helped Rappo acquire 100+ clients and secure first-round funding
⋆˙⟡ let's float together

say hello

drifting in from a recruiter, a classmate, or just curious? say hi! i’m happy to talk risk, models, or tide pools.

download résumé ↓
emailvanessalfeng@gmail.com
linkedinin/vanessalfeng
phone626-267-9978