Compute EL and RWA from loan data

Quick Overview

Compute EL and RWA from loan data evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

Compute EL and RWA from loan data

Company: Wells Fargo

Role: Data Scientist

Category: Machine Learning

Difficulty: medium

Interview Round: Technical Screen

Given anonymized loan data containing PD, LGD, and EAD, compute portfolio expected loss (EL) and risk‑weighted assets (RWA). State assumptions, formulas, aggregation approach, treatment of off‑balance‑sheet exposures, and any sensitivity checks.

Quick Answer: Compute EL and RWA from loan data evaluates core ML concepts, assumptions, math intuition, training/evaluation trade-offs, and practical failure modes in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

|Home/Machine Learning/Wells Fargo
Wells Fargo logo
Wells Fargo
Jul 26, 2025
mediumData ScientistTechnical ScreenMachine Learning
6
0

Compute EL and RWA from loan data

Task: Compute Portfolio EL and RWA from Loan-Level PD, LGD, EAD

Context

You are given an anonymized, loan-level dataset with at least the following fields per exposure:

  • PD: one-year probability of default (as a decimal, e.g., 0.01 for 1%)
  • LGD: loss given default under downturn conditions (decimal between 0 and 1)
  • EAD: exposure at default (currency units). For off-balance-sheet items, EAD should reflect credit conversion factors (CCFs), or you must compute EAD from committed/undrawn amounts and CCFs.

Compute:

  1. Portfolio expected loss (EL)
  2. Risk-weighted assets (RWA)

Also state:

  • Assumptions (regulatory approach, asset class, maturity, parameter floors)
  • Formulas used
  • Aggregation approach across loans
  • Treatment of off-balance-sheet exposures
  • Sensitivity and validation checks

Clarifying Questions to Ask Guidance

  • Clarify the task, data shape, labels, constraints, and evaluation metric.
  • State assumptions behind the math or modeling technique you choose.
  • Connect theory to practical training, debugging, and deployment implications.

What a Strong Answer Covers Guidance

  • Correct definitions and formulas where the prompt requires them.
  • A practical explanation of how the method behaves on real data.
  • Trade-offs, failure modes, diagnostics, and mitigation strategies.
  • Evaluation choices that match the product or modeling objective.

Follow-up Questions Guidance

  • How would noisy labels, class imbalance, or distribution shift affect the answer?
  • What would you monitor after deployment?
  • Which baseline would you compare against first?
Loading comments...

Browse More Questions

More Machine Learning•More Wells Fargo•More Data Scientist•Wells Fargo Data Scientist•Wells Fargo Machine Learning•Data Scientist Machine Learning