Made Card's credit edge.
Most issuers segment by spend category. Travelers get travel cards. Foodies get dining cards. We built ours around something more fundamental — homeownership — and the data underneath is the most under-appreciated edge in consumer credit today.
Most credit card issuers segment their customers by spend category. Travelers get travel cards. Foodies get dining cards. Cash back maximizers get flat-rate cards. The segmentation logic is sound, but it misses something more fundamental: homeownership.
At Made Card, we built our entire product around homeownership. Here's why we think it's the most important and under-appreciated edge in consumer credit today.
01 — The opportunityThe total addressable market is large.
The U.S. homeownership rate is 65%, representing approximately 86 million owner-occupied households. Homeowners earn nearly double the median income of renters ($86,000 vs. $42,500) and hold 40 times the median net worth ($225,000 vs. $6,300). That is a large and financially stable population that has been almost entirely ignored as a distinct credit card segment.
Owners are 65% of households — but 80% of spend capacity.
Owners outearn renters roughly 2× on median income, so the spend-capacity skew is sharper than the household share.
Share of households
65% of US households own their home — 86M out of ~133M total.
Total addressable market
Owners earn 2× more ($86K vs $42.5K median) — blending households × income, owners represent 80% of spend capacity.
Homeowners carry an average FICO score of 754, solidly in the "excellent" range. The average renter scores 638 — "fair." That's a 100+ point gap. This gap in the credit score directly translates to loss rates.
In a business where net interest margin is measured in hundreds of basis points, a 200–300 bps spread is the difference between a mediocre portfolio and a great one.
The Federal Reserve Bank of New York's Consumer Credit Panel — a nationally representative sample of Equifax credit bureau data covering tens of millions of consumers — shows a persistent and meaningful gap of 200 to 300 basis points in credit card delinquency rates between homeowners and renters across multiple economic cycles.
The gap is persistent — through every cycle since 2001.
Credit card delinquency rates by mortgage status (proxy for homeownership), quarterly, seasonally adjusted. Homeowners track 200–300 bps below renters across boom, GFC, and post-COVID periods.
Moreover, even conditioned on credit score stratification, the gap remains tangible. A data study of personal loans issued between July '21 and June '23 indicates that within each Vantage Score band, loans issued to consumers with a mortgage tradeline performed significantly better than ones issued to consumers without. For example, for the Prime bucket (661–699), loans without a mortgage had a 110 bps higher 60-day delinquency rate versus ones with a mortgage — 7.4% vs. 6.3%. The trend persists across credit segmentation.
The mortgage tradeline lowers target rates within every Vantage band.
12-month delinquency study of ~2.4 Million personal loans issued July '21 to June '23.
| Vantage bucket | With mortgage | Without mortgage | ||
|---|---|---|---|---|
| % of total | 12-MONTH 60-DAY DELINQUENCY RATE | % of total | 12-MONTH 60-DAY DELINQUENCY RATE | |
| 1 · Super Prime (781+) | 8.8% | 0.6% | 7.5% | 1.0% |
| 2 · PrimePlus (721–780) | 8.5% | 1.8% | 11.5% | 2.8% |
| 3.1 · Prime-2 (700–720) | 3.7% | 3.9% | 8.5% | 4.8% |
| 3.2 · Prime-1 (661–699) | 6.4% | 6.3% | 18.3% | 7.4% |
| 4 · Near Prime (601–660) | 5.3% | 9.8% | 21.6% | 11.1% |
| Grand total | 32.5% | 3.9% | 67.5% | 6.8% |
02 — The mechanismWhy does the gap exist?
The gap is a direct byproduct of mortgage underwriting. To get a home loan, a borrower must demonstrate stable, verifiable income, meet debt-to-income requirements, sustain a creditworthy history, and survive one of the most rigorous financial reviews a consumer will ever face. The mortgage process is, in effect, a pre-screen — and it's one that every homeowner has already passed before they ever apply for a Made Card.
When we underwrite a Made Card applicant, we are starting from a population that has already been filtered by one of the toughest underwriting processes in consumer finance, and one that has to conform to rigorous guidelines established by Fannie, Freddie, et al. We then apply our own credit model on top of that foundation. The result is a meaningfully lower risk profile than a mass-market card issuer targeting the general population.
03 — DistributionThe distribution moat.
If homeowners are such an obvious target, why hasn't anyone built this before?
The data infrastructure to identify homeowners exists. Property records, mortgage databases, and credit bureau attributes can flag homeownership status. The problem is cost. Purchasing homeowner data from third-party marketplaces is expensive, and that expense flows directly into customer acquisition cost. A mass-market card issuer running a national direct mail campaign targeting homeowners will spend considerably more per application than one with organic access to that population at the moment of purchase.
We have organic access. Made Card has built a network of mortgage company partnerships that gives us direct reach to homebuyers at the exact moment they are closing on a home — not weeks later through a mailer, but in the transaction itself, facilitated by the loan officer who just helped them close.
A population that is credit-starved, financially pre-qualified, and sitting on a large, immediate spending need.
The net effect of this strong origination channel is a steady-state LTV/CAC ratio that is 2–4× higher than comparable direct-to-consumer fintech companies. More importantly, the moment we target — the period immediately following a home purchase — is a perfect product-market fit. During the mortgage process, buyers are instructed not to apply for any new credit from pre-qualification through closing, a period of three to six months. The day they close, that restriction lifts — and they immediately face large discretionary spending: moving costs, furniture, appliances, repairs, paint, landscaping, contractors. The average new homeowner spends thousands of dollars in the first 90 days after closing, often across the exact categories where a rewards card delivers maximum value.
04 — UnderwritingRisk management is in our DNA.
Made Card's Chief Risk Officer was our second full-time hire. That sequencing was intentional. We believed in embedding risk discipline in the company's culture from day one — not retrofitted after the first loss spike.
We use a modular underwriting engine that allows us to iterate on underwriting policy without constant engineering involvement. In our first year of operation, we ran 65 versions of our underwriting workflow. Most of those iterations were executed entirely outside the engineering team, by our Chief Risk Officer working directly in the decisioning layer. Our integration with an industry-leading fraud risk management provider took all of 30 minutes.
Fraud risk system
Our underwriting flow first filters for eligibility. Eligible applicants are then evaluated for fraud risk in a waterfall approach that separates the highest- and lowest-risk applicants for efficient decisioning. Applicants at the margin are stepped up for additional evaluation and manual review. Key features:
- Partnerships with industry-leading identity verification providers
- Scoring models for every piece of PII
- Use of consortium data
- Digital intelligence covering device, network and behavior signals
- Evaluation of first-party fraud risks (which has seen a rising trend)
Credit model
Only applicants that pass our fraud checks go through credit checks. To get an informed view of unit economics, we invested in a bureau data study that looked at the 12-month delinquency performance of 1 million credit card accounts opened by customers who had a mortgage (as a proxy for homeowners). We chose Vantage Score 4.0 given its comparable performance to FICO at better economics.
Vantage Score provided good credit discrimination. However, we knew from experience that an internal model built on our specific customer base and product would outperform a generic credit score. We partnered with HudsonData to build a machine learning underwriting model that uses XGBoost (Extreme Gradient Boosting), a state-of-the-art ensemble technique known for high predictive power and efficiency. The model delivers meaningful lift over single-score decisioning, as shown in the cumulative gains chart below.
The V1 model captures 95%+ of defaulters in the top 4 deciles vs. 85% for Vantage.
Cumulative gain plot — share of defaulters captured at each rank decile. Higher curves at low deciles = better discrimination.
This risk score is used for approve-deny, pricing, and line assignment decisions. For every model-driven decline, clear reason codes are provided using SHAP (SHapley Additive exPlanations) values to quantify how each feature influenced the score. This promotes transparency, aligns with fair lending principles, and fulfills regulatory expectations for explainability in automated credit decisions. Our risk-based pricing means our best customers receive APRs that mass-market issuers cannot match, because those issuers are pricing for a riskier average.
We continue to invest heavily in data. Whether it be device signals or credit report data, we capture every signal obtained throughout the application process and customer lifecycle. Using AI tools (and with limited engineering help), we have been able to build dashboards that provide real-time insights into underwriting and portfolio performance, allowing us to fine-tune our models against real portfolio data faster than incumbents can move...
The excess spread from our homeowner credit advantage goes back to the customer in the form of richer rewards or lower rates. That is our value proposition: a card that is materially better for the people it is designed to serve, funded by the credit quality of the population itself.
Written by
|
Alex Song
Alex is a Co-Founder of Made Card, a consumer fintech company shaking up the mortgage and credit card ecosystem. Prior to Made Card, Alex was an early executive and Head of Finance at Ramp. He brings 13+ years of experience as a hedge fund investor across asset-based financing, private credit, structured credit, and ABS/RMBS. In addition to being a seasoned fintech operator, he is also an avid angel and early-stage investor, and has advised companies including Gemini and Navan on their card programs.
|
|
Sahas Ranganathan
Sahas is the Chief Risk Officer at Made Card, where he leads credit, fraud and decision science. He brings over two decades of risk management experience spanning fintech, big banking, and enterprise financial services. Before Made Card, he served as CRO at Balance Homes, a home equity investment startup, where he built risk pricing, underwriting, and collections capabilities from the ground up. Prior to that, he held senior risk leadership roles at Wells Fargo and American Express, where he led Enterprise Risk Management and Model Risk Management functions across regulatory, capital, and new AI/ML models. He began his career in consumer credit risk at American Express, building authorization strategies for the cards portfolio.
|
It’s time your home started paying its share.
