Cash flow underwriting is having a moment. And for good reason. Research from FinRegLab shows that machine learning models combining cash flow and credit bureau data can increase credit approvals by about 4% at mainstream risk cutoffs, without increasing defaults. To put that in perspective: applied to 2023 origination volumes, that translates to roughly two million additional credit card accounts and 152,000 additional mortgages. For an estimated 32 million Americans who can't be scored under widely used credit models, that's not an incremental improvement. It's a fundamentally different outcome.
But here's what tends to get lost in the excitement: underwriting models are only as good as the infrastructure beneath them. The conversation about smarter credit rightly focuses on what cash flow data can do. It's critical to understand what it takes to make it work at scale, reliably, and for everyone.
Most of the early traction in cash flow underwriting has come from lenders using their own customer data (what the industry calls "on-us" or first-party data) to enhance credit decisions. That's a solid starting point, but the real scale opportunity is in "off-us" data: financial information held at other institutions that a consumer chooses to share with a lender.
Think about it from a lender's perspective. You're making a credit decision, and you can see transaction activity at one institution. But the borrower's full financial picture (income deposits, rent payments, recurring obligations) might live across two or three accounts at different banks and credit unions. Cash flow underwriting reaches its potential when lenders can access a complete, consumer-permissioned view of financial behavior across institutions.
That's where the infrastructure question becomes critical. In a recent survey conducted by Akoya and American Banker, we found that the ambition is real but the execution is still catching up. 83% of financial institutions call open finance a priority. Nearly a third of those already consuming open finance data cite alternative lending models (including cash flow underwriting) as a top use case driving investment. The demand signal is clear.
The progress the industry has made is worth acknowledging. The FDX ecosystem now connects approximately 114 million accounts through standardized APIs. More institutions are moving toward open finance programs, building developer portals, and standing up the connectivity infrastructure that makes consumer-permissioned data sharing possible.
But the industry still have a long way to go. In that same American Banker survey, only 3 in 10 financial institutions exclusively share data via APIs. About half operate in a hybrid model, mixing API connectivity with screen scraping depending on the third party requesting data. The rest haven't made the shift at all.
This matters for cash flow underwriting specifically because data quality and reliability are not negotiable inputs for credit models. Screen scraping is fragile. It breaks when an institution updates its online banking experience. It can't enforce what data gets accessed. It introduces noise and inconsistency into the data that lenders are building models on. When a lender is making a credit decision, you want the data to be exact (the precise account balance, the verified deposit history, the complete transaction record). API-based connectivity, built on standards like FDX, delivers that reliability. Screen scraping does not.
From Akoya's perspective, this is our core focus. We operate as the infrastructure layer (the pipe, not the pump) that connects financial institutions to the broader ecosystem through secure, standardized APIs. We don't build credit scores or analytics. We make sure that the data feeding those models is accurate, complete, and delivered in real time directly from the source. Because models are only as good as the underlying data.
The institutions that treat open finance as truly critical are seeing results. Our research found they are almost twice as likely to view open finance as a path to new revenue and markets compared to those treating it as a moderate priority (46% vs. 25%). That's a meaningful signal that the business case is separating from the compliance case.
Here's where the infrastructure conversation turns into a fairness conversation.
If a consumer's bank isn't set up to share data via secure APIs, that consumer can't participate in cash flow underwriting. Full stop. It doesn't matter how sophisticated the model is or how much predictive power cash flow variables add. If the data can't flow reliably from the institution to the lender, the consumer is left out.
With only 3 in 10 institutions fully on APIs and roughly half still relying on a hybrid mix, there's a meaningful coverage gap. Consumers at institutions with mature API connectivity get access to better credit outcomes through reliable, high-quality data. Consumers at institutions that haven't made the transition are stuck with either fragmented, scraping-dependent data (or no data sharing at all). That's a two-tiered system, and it disproportionately affects consumers at smaller banks and credit unions who may already be underserved by traditional credit infrastructure.
Over 96% of American households have a bank account. The financial behavior data exists. The question is whether the infrastructure exists to make it available in a way that credit models can actually use.
This is why expanding FI participation isn't just a growth strategy for Akoya, it's a coverage imperative. The more institutions that move to API-based connectivity, the broader the population that can benefit from cash flow underwriting. Partners who build analytics on top of our network (like Nova Credit, whose cash flow and income analytics are powered by Akoya's secure data access) can only reach consumers whose institutions are connected. Every bank or credit union we bring onto the network expands who gets a fair shot at credit.
FinRegLab's earlier research reinforces this point: cash flow data provides independent predictive value across demographic subgroups without acting as a proxy for race, ethnicity, or gender. The data works. But it only works if it's accessible. Infrastructure readiness determines whether inclusion is a reality or a talking point.
Consumer trust is the currency of open finance. Without it, none of this scales.
Trust is built through a combination of factors. Transparency, so consumers understand who they are sharing data with and for what purpose. Active, permissioned consent, so consumers opt in to data sharing rather than having it happen in the background. Revocation, so consumers can change their mind and cut off access at any time. And data minimization, so that a third party only retrieves and accesses the data needed to fulfill the product or service the consumer wants.
All of these components build confidence in the ecosystem. They build confidence for consumers that their financial data is handled responsibly. And they build confidence for lenders that the data they're receiving is reliable, consented, and compliant. When lenders can trust the integrity of the underlying data, their models carry less risk and produce better outcomes.
Akoya is purpose-built for this. We are API-only (no credentials, ever). Consumer permission is mandatory for every data request. Our architecture enforces data minimization by design, not as an afterthought. These are not policies we bolted on. They're day one architectural choices that reflect the fact that Akoya was built by financial institutions, for the ecosystem, with the governance and accountability standards that come with that.
The American Banker survey underscores why this matters at the institutional level. Security, regulatory uncertainty, and data liability are the top three concerns blocking broader open finance adoption. 77% of FIs currently integrating open finance cite third-party risk management as a significant or moderate challenge. Institutions want a permission-based system that's built to scale and easy to govern. That's the infrastructure Akoya provides.
Trust also has a compounding effect. When consumers see that they control what's shared, when they can revoke access simply, when they understand the purpose (and that purpose is getting better access to credit), adoption grows. And when adoption grows, the ecosystem becomes more valuable for everyone: consumers, lenders, and financial institutions.
If you're a lender or financial institution thinking about cash flow underwriting, here's what I'd leave you with.
First, don't wait for regulatory certainty. The direction of policy (whether through the CFPB's Personal Financial Data Rights Rule or broader industry alignment around FDX standards) is clear: secure, permissioned, API-based data exchange is where the industry is headed. Institutions that are already operating with this infrastructure aren't scrambling to comply. They're already there.
Second, the business case is already here. 83% of FIs call open finance a priority. Research shows that cash flow data meaningfully improves credit outcomes. Institutions that lean in are seeing it as a path to new revenue and deeper customer engagement, not just a compliance exercise.
Third, the infrastructure layer matters more than most people think. The analytics and scoring innovations being built across the industry are impressive. But they depend on reliable, high-quality, consumer-permissioned data flowing from financial institutions to the companies building those products. Getting that foundation right is what Akoya focuses on every day.
If you're looking to stand up the data sharing infrastructure that makes cash flow underwriting (and the broader open finance opportunity) possible, set up time to connect or dig into Akoya's Open Finance Solution.