JD Case Study · Concora Credit
VP, Risk — Fraud Strategy & Analytics
A VP owning fraud strategy and the analytics behind it for a subprime credit-card book — the exact risk/decisioning world I lived in at Acima: income-and-bank-data underwriting, first-payment default, charge-offs, the approval/loss frontier. This is the closest fit in my portfolio to a hiring manager’s actual day.
What they’re actually buying
Protect a subprime portfolio’s economics without strangling its growth. Own the fraud and risk-decisioning strategy, the models underneath it, and the test-and-learn that tunes the approval/loss frontier. Stripped down, the job is one trade-off, quantified: every basis point of fraud and credit loss against every approved account you turn away.
How I’d own it — first 90 days
Days 1–30
Map the loss curve
Decompose fraud and credit losses by segment, channel, and vintage. Baseline the applications → approved → funded → first-payment-default → charge-off funnel, and meet Risk, Ops, and Analytics.
Days 31–60
Stand up the decisioning frontier
Ship a risk-frontier dashboard — approval rate vs loss rate by segment — to show where tightening protects margin and where it’s quietly throttling good accounts. Identify the two or three highest-value moves.
Days 61–90
Run the first risk experiment
A/B a fraud-strategy or decisioning change with an explicit loss guardrail. Deliver the business case — loss reduction vs approval impact, in dollars — and set the model-governance and experimentation cadence.
Signature analysis: the subprime decisioning funnel
Where strategy pays off: first-payment default is the earliest fraud/credit gate, and the cheapest to fix. A tuned fraud model cuts FPD with minimal approval drag — the prize is loss reduction without turning away good accounts. That’s the frontier I’d manage, not a blanket tightening.
Illustrative figures — the same decisioning funnel I built for Acima (delivery-verified funding, first-payment gate, charge-offs), here on a subprime card book.
Requirement → proof
| What the JD asks for | What I’ve already shipped |
|---|---|
| Subprime risk / decisioning strategy | I modeled Acima’s funnel end-to-end: approval → delivery-verified funding → first-payment gate → charge-offs. |
| Experimentation with loss guardrails | An A/B / incrementality readout with a charge-off guardrail and a sample-size calculator. |
| Segmentation & loss / unit economics | An LTV / unit-economics model where the charge-off rate is the central lever. |
| SQL / SAS / Python | Show-the-SQL blocks, channel-aware, with the charge-off guardrail built in. |
| Subprime consumer-credit domain | A career in lease-to-own (deep-subprime), plus published spotlights across the subprime-credit field. |
The same engine
The engine behind this whole portfolio — approval/funding funnel, first-payment-default gate, charge-off economics, experimentation with a loss guardrail — is the same machine a subprime card book needs. I built it for lease-to-own; the merchandise changes, the risk/growth frontier doesn’t. I’d walk in fluent in the trade-off this role exists to manage.
See the decisioning funnel I built →Subprime risk isn’t a domain I’d ramp into — it’s where I’ve worked and what I’ve already modeled in public. Let’s talk about your loss curve. — Paul Brown