Why Issuing Credit Cards is Hard

3 FINTECH NEWS STORIES

#1: The Difference Between a Bank and a Network

What happened?

Capital One is teaming up with Stripe and Adyen:

While Capital One has built models to protect its customers from fraud, it was getting stuck because it just “didn’t have enough data,” said head of fraud strategy Jon Borman.

So it built an open source project called Direct Data Share, which Borman said is an API that allows merchants or anyone in the payment stack to send real-time transaction data. This is particularly useful for e-commerce transactions, which he described as “pretty unsafe” compared to in-person purchases where a user is verifying the chip on a credit card, for example.

With Direct Data Share, every time an online purchase was made, “a bunch of data” flowed through Capital One’s API to the bank, which would be used to help prevent fraud for more customers and merchants, he said.

But by partnering with Stripe and Adyen with Direct Data Share, Capital One can act like a data clearinghouse, identifying fraud across all of their rails.

So what?     

Anytime a new fraud data consortium pops up, you can be assured I will write about it.

This one is obviously notable because of the names involved.

Capital One isn’t exactly known for playing nicely with others in the ecosystem, but as skee ball aficionado Matt Janiga pointed out on Twitter, the company may be trying to turn over a new leaf in the wake of its planned acquisition of Discover. 

Operating a payment network requires finding as many ways as possible for all network participants to win at the same time. Stopping fraud is an easy and obvious place to start, and this positions Capital One to grow acceptance of the Discover network (assuming the acquisition goes through) in the high-growth merchant segments that Stripe and Adyen specialize in.

It’s also not great news for the established vendors in the transactional fraud detection space (Actimize, Sift, Kount, FICO, etc.), though Capital One is likely not interested in competing with them directly (at least not yet).    

#2: Consumer Report

What happened?

Plaid launched a new credit solution, built on its consumer reporting agency (CRA):

With Consumer Report, companies can now embed cash flow data and credit risk insights directly into their underwriting and verification workflows with a single API.

Our customers typically have two main goals in today’s credit environment: they want to expand access to qualified applicants based on net positive cash flow, regular rent payments, and stable employment. And they want to ensure their applicants have the ability to repay the loan using real-time signals of cash flow health, such as the stability of their bank balances or income. In many cases, these signals are not captured by other credit bureaus, but through Consumer Report, our customers get these critical insights instantly.

So what?     

Consumer Report (which is a strangely traditional name for the product) currently includes three things:

1. Categorized Cash Flow Insights – These are the basic inflow and outflow attributes summarizing a consumer’s bank account(s) for the prior 24 months.   
2. Income Insights – These are a set of more sophisticated attributes on the inflow side, summarizing everything from debt-to-income ratios and historical and forecasted net and gross income to employer name and projected next paycheck date. 
3. A cash Flow Underwriting Score – for the score (and associated attributes), Plaid partnered with Prism Data.

This is the logical next step for Plaid, following its move to spin up a CRA (which Fintech Takes reported last fall). Nothing here is shocking to me, although I do find it interesting that the company chose to partner with Prism on scores rather than develop one of its own. It’s not unheard of, certainly. Equifax, TransUnion, and Experian spent decades reselling the FICO Score before they decided to team up on VantageScore.

Plaid’s challenge, similar to that faced by traditional credit bureaus, is to avoid becoming commoditized. Plaid’s strategy for overcoming this challenge appears to be bundling the broadest set of solutions (credit, fraud management, payments) around its core product (data aggregation). 

#3: Why Issuing Credit Cards is Hard

What happened?

LoanPro built a new integration with Visa:

LoanPro … today announced it will utilize their direct integration with Visa DPS, one of the largest issuer processors of Visa debit transactions globally, to provide card issuers, brands, financial institutions, and fintechs with a unified platform to seamlessly launch, manage, and service next-generation credit card programs, while unlocking competitive differentiation.

This integration of LoanPro’s modern credit platform with Visa DPS’ digital-first issuer processing paves the way for the deployment of first-to-market credit card programs that ensure scalability while driving growth through unique features such as transaction level credit. Transaction level credit offers unprecedented precision in managing transactions, allowing card programs to customize interest rates, credit limits, and grace periods based on the specifics of each transaction, including amount, location, merchant type, and more.

So what?

I’m oversimplifying a bit, but to launch a debit card, you need three things: a bank partner, a ledger, and an issuer processor.

Today, there are lots of different infrastructure providers out there that can help you get these things. Most modern card issuers offer both the issuer processing and the ledger. And if you want the whole kit and caboodle, that’s where BaaS middleware platforms come in (though Synapse may end up ruining that model for everyone … we’ll see).

If you want to launch a credit card – a true credit card, not a charge card – it’s trickier. In addition to the bank partner and issuer processor, you also need debt capital (this is the money that you’ll be borrowing and lending back out) and a loan servicing system. The loan servicing system is essentially a ledger, but 1000x more sophisticated (if you need a primer on why loan servicing is much more complex than you might imagine, here you go).

It’s tough to find an infrastructure provider that brings all of these capabilities to the table. There are co-brand credit card platforms – your Deserves and Cardlesses – that will get you close. But you’ll give up a lot of control (and a decent chunk of the unit economics) to use them.

This is (likely) a more appealing solution for most fintech companies interested in launching a credit card right now (and given the focus on profitability in fintech, that’s most fintech companies): an easy-to-integrate-with issuer processor (Visa DPS) combined with a modern loan servicing system (LoanPro).
There are a few other companies looking to assemble a similar bundle (Marqeta, after its acquisition of Power, comes to mind), and someone still needs to solve for the debt capital part of the puzzle (Sivo, perhaps?), but this news is definitely a step forward.

---

2 FINTECH CONTENT RECOMMENDATIONS

#1: Funding Fears Transform Into Profitability Woes (By Kiah Haslett, Bank Director) 📚

There are ample sources of wholesale funding for community banks, and community banks have gotten better (and more aggressive) at tapping them over the last 18 months.

However, reliable funding doesn’t equal cheap funding, and recent data suggests that the funding costs incurred by community banks are having a deleterious effect on those banks’ profitability.

As always, I feel smarter after going to Kiah Haslett school.     

#2: Real-Time Payments in 1,000 Words (by Matt Brown) 📚 

Another excellent primer from Matt, this time on real-time payments, which is a subject that most can explain in 2 minutes, but struggle to more deeply explain in 20. 

---

1 QUESTION TO PONDER

Why is it that B2C fintech companies are so reluctant to go with a classic secured card as their credit builder product? What are the primary objections to this product type?