Why fintech innovation is broken at its core

By James Garvey, CEO of Self Financial

Fintech innovation is broken at its core.

Our financial institutions employ some of the sharpest minds in the world — yet, the majority of Americans are still doing the equivalent of “driving to Blockbuster to rent a VHS tape” when it comes to banking.

Have you ever wondered why banks are so slow to innovate?

At Self, we needed to integrate into a bank partner’s backend software. We patiently waited 8 weeks for a phone call that would dictate the integration cost and timeline.

To our dismay, we found out that we needed to spend $X**X,XXX** and wait Y months — just to get API documentation and sandbox API access — which would allow us to perform the integration. On top of that, we would need to pay a large annual fee.


The amount of money and time delay was unacceptable — yet we needed the functionality. We aren’t the only fintech company with this problem.

What is the core?

Core banking software or simply “core” is the backend software that powers a financial institution.

The core is the center of all bank operations. As a centralized record keeping system, the core processes deposits, payments, loans, and also manages customer information.

Ostensibly, the core is just a spreadsheet layered on top of a big pile of money. Your checking account has the look and feel of a structured tranche of money — but, it’s just software that provides the illusion of separation.

If you have a background in computer science, and an understanding of accounting, I guarantee that you can build your own core. However, for most U.S. bank CEOs, this concept is absolutely inconceivable.

4 companies own 96% of U.S. cores

According to American Banker, just 4 companies control 96% of core banking software. These four companies are:

  1. DH — 7,000 financial institutions (FIs); $1.13B annual rev → $162K/bank

  2. Jack Henry — 10,900 FIs; $1.25B annual rev → $115K/bank

  3. FIS — 14,000 FIs; $1.58B annual rev → $113K/bank

  4. Fiserv — 14,500 FIs; $5.06B annual rev →$349K/bank

Servicing small-dollar loans is expensive

Self helps consumers establish credit history through a small, secured installment loan. For years, these “credit builder loans” have been offered through a retail setting by credit unions and small banks. The problem with small-dollar loans (e.g. $1,100) is that the servicing costs can be more expensive than the generated interest income.

Here’s why:

Let’s estimate the number of checking accounts in the US:

  • 88.4% of U.S. households have a checking account1

  • total of 242mm U.S. adults2

  • → estimated ~213.92 million checking accounts

Given that the 4 core providers control 96% of the core processing market:

  • DH + Jack Henry + FIS + Fiserv → $9.02 billion in revenue

  • Assuming equal distribution, then 100% of the core software market value → $9.39 billion in revenue

  • $9.39 billion / 213.92mm checking accounts → $43.89 per account per year

  • Just for good measure, let’s cut the number in half → $21.94/account

Now, suppose you’re the CFO of an average American bank that uses one of the big core providers. What kind of loans would you offer?

  • Core software costs to maintain 1 member account: $21.94/year

10 loans → $219 core fees

  • (10) x $110,000 loans → $1.1 million in loans

  • $219.40 in core servicing costs

1,000 loans → $21,940 core fees

  • (1000) x $1,100 loans → $1.1 million in loans

  • $21,940 in core servicing costs

The math is obvious: banks who rely on these core providers cannot be profitable on high volume, small-dollar transactions.

Yet, banks are stuck using these core systems. Industry experts often claim that “switching cores is like changing jet engines during flight.”

This is why 96 of the top 100 banks are still using mainframes3 and legacy code written in COBOL from the 1980s.

Don’t believe me? Here’s a screenshot from one of the big core provider’s jobs pages:

For reasons of accounting, compliance, security, and credit reporting — banks want data integration with their core. With so many financial institutions stuck on legacy architecture, innovation is hard to embrace.

The unfortunate truth is that:

  • the core providers have an oligopoly;

  • core software is the lifeblood of banks;

  • and therefore, any third-party integration with a core provider is going to be extremely expensive and very time-consuming.

We couldn’t afford the additional burden of core fees — yet we needed the functionality.

This is why we’ve built our own core.

We’re about to go through an SSAE 16 audit which will give our partners the confidence and assurance that they need. Our core will allow us to do the same underlying backend bank operations without having to pay third-party core maintenance fees in perpetuity.

The beauty of it all? API-driven consumer banking.

We’re so excited about the technology that powers Self that we’re letting the world know!


  1. https://fdic.gov/householdsurvey/2013execsumm.pdf

  2. http://quickfacts.census.gov/qfd/states/00000.html

  3. http://www.americanbanker.com/news/bank-technology/why-citi-is-buying-ibms-new-mainframe-for-mobile-transactions-1072161-1.html

About the Author

James Garvey is the CEO and co-founder of Self Financial, Inc.

self logo
Written on October 2, 2015
Self is a venture-backed startup that helps people build credit and savings. Comments? Questions? Send us a note at hello@self.inc.

Disclaimer: Self is not providing financial advice. The content presented does not reflect the view of the Issuing Banks and is presented for general education and informational purposes only. Please consult with a qualified professional for financial advice.

Take control of your credit today.

Please enter a valid email.