The Case for a Decentralized Open Banking Infrastructure

The challenges of distributing relief dollars have exposed structural issues that run deeper and need fixing; they’re more pressing now, but they’ve existed for a long time

Published in

The Startup

12 min readApr 13, 2020

Open Banking and blockchain could speed up the process of distributing relief dollars from weeks or months to minutes.

Fintechs have stepped up and are leading the way through the crisis, but we must think harder about driving structural change to create a resilient financial ecosystem.

Technology has made us more resilient than merely a decade ago. This is in part because technology itself is more flexible and adaptable; It’s less purpose-built and more centered around use cases, making it a more powerful countercyclical tool.

Such flexibility has greatly contributed to how we respond to the COVID-19 crisis. It’s behind tech (and particularly fintech) companies’ ability to repurpose or reengineer an existing piece of software to create a solution in record times. Thankfully, the number of fintech companies that are stepping up and the list of measures they have announced is already long and keeps growing by the day.

Nevertheless, there is also increasing frustration from businesses and individuals, as the wheels of the government and legacy financial institutions, ─of all sizes─ charged with processing applications and distributing relief funds do not move nearly as fast.

Millions of jobs have already been lost, while millions more remain on the line. Yet, powerful technology remains largely sidelined while we painfully wait for stimulus dollars to slowly trickle-down the long and winding apparatus. In 2020 we’re still bounded by systems built decades ago, despite the banks touting annual investments in the billions every year to upgrade their tech stacks.

Government Dollars: The Long And Winding Road

In a best-case scenario, individuals will receive a payment from the IRS roughly three weeks after it was announced. For people without direct deposit, this could take as long as five months. For those who felt like providing/updating this information, the IRS took weeks ─literally─ just to put together a webpage. One small detail, though; the first webpage is only for non-filers. People who simply wish to avoid waiting for a paper check must wait a little longer, for there is a second webpage that isn’t live yet.

The outlook is just as bleak for SMBs, who can expect average waiting periods of two to three months, in addition to having to collect a ton of paperwork for an application, a process with challenges of its own. Unclear guidelines, concerns over legal liability, and manual/legacy processes only add to the problem.

Is this the best we can do? While our priority remains to solve for our current challenges, we can also use this time to think about how to build a resilient financial ecosystem, the one that knits together our economic activity. We must capitalize on the behavioral shifts driving increased adoption of digital financial products, as well as on the government’s shifting attitudes to drive profound, impactful changes.

From Roadblocks to Roadmap

A fundamental proposition is that the difficulties associated with distributing relief payments to individuals and businesses merely bring attention to more fundamental inefficiencies in the system. Thus, some of these issues run much deeper but are only highlighted by our current situation.

Fintechs are diligently helping to speed up the distribution of relief funds by addressing one or more pain points along the road, namely data aggregation, processing/validation, and disbursements.

The process differs for individuals and companies (e.g., the IRS requires no action from most individuals, while SMBs need to apply with an eligible lender for the PPP). Therefore, the specifics of the issues to solve for also change, but they mainly fall along these lines.

To illustrate this better, let’s just recover a few notable examples. The links to these are

Data Aggregation / Pre-processing: retrieve data from multiple sources, assemble machine-readable information packets.

2. Data processing / Validation: deploy software to automate largely-manual review and approval processes.

3. Disbursements: provide the earliest possible access to funds

While by no means exhaustive, this list allows us to observe some common elements. First, people and SMBs are not in the driver’s seat when it comes to their information. Second, there’s no single source of truth; in an ideal world, data aggregation, processing, and validation would take place simultaneously, but we’re far from there yet.

A third and more nuanced point relates to our inability to underwrite individuals. It’s almost ironic that the only tool we have to accelerate the process for people are cash advances, which stem from a massive inefficiency: that it takes days up to five days for the money to clear.

Can we think of a system that tackles all three issues simultaneously, thus allowing us to be expeditious in how we price risk, make decisions, and move money? The short answer: Yes, and we don’t need to invent anything new; we mostly have the pieces that we need, and it’s about putting them together thoughtfully.

There are many ways to mix-and-match the existing tech toolkit to innovate on this topic. But here’s my take on it, in which I think that Open Banking and blockchain (or a similar open-source, decentralized architecture) show particular promise.

The Power Of Information: Fulfilling The Promises of Open Banking

Open Banking primarily seeks to return people the de-facto ownership of their financial information. They can choose to share this information with financial institutions, effectively leveraging it to get access to better products and services. Financial players also benefit, as they can now better understand their customer base and improve their value proposition to drive the growth of the business.

Nevertheless, in most of the Western hemisphere ─the US included─, while accessible, our information remains scattered. This has to do with how technology evolves and how we use it.

The process starts when challengers set out to solve for one problem big-enough to achieve product-market fit and create a business around it. But because they naturally have bigger ambitions, primarily to become the center of their customer’s financial lives, and because customers look for simplicity, the rebundling and disruption cycle starts over (plus, add to that new products).

Therefore, paradoxically, the tools that should allow us to recover the power over our financial information might make it harder for us to keep track of it. This is because the end-users of the tools to move around our information are developers, not people. We need to scramble over an increasing number of data sources while this information is seamlessly shared between applications.

This is not an argument against Open Banking. On the contrary, these tools have created a fertile environment for innovation that benefits us all, but we all know that sometimes there are unintended consequences or negative externalities. While we should continue to create powerful software tools for developers, it’s also time to turn to the consumers.

For instance, despite (or precisely because of) all the budgeting and personal finance features offered in all of our apps, there’s still a tremendous use case for Microsoft Money (a collaboration between Plaid and Microsoft), as yet another effort to make sense of our scattered financial lives. Remember Plaid’s product to streamline the process of preparing payroll data for SAB applications, a process that remains mostly manual despite many businesses already using SaaS services.

Another potential consequence is that some information asymmetries remain. These hurt both consumers and financial providers, as these represent mutually beneficial business opportunities that are lost ─a sub-par equilibrium. Neither customers nor their new digital partners have the full picture of their financial lives.

In summary, Plaid ─and similar “information plumbing” companies, have done a titanic job of building the infrastructure to integrate with thousands of FIs and move information around. That, by all means, is a net positive.

Yet, we’re not on the driver’s seat when it comes to our financial information; this doesn’t relate exclusively to privacy, but also as to how we can readily access it and make use of it. We either painfully scrape our accounts and files or sign a blank check to someone to do it for us, sometimes helping and other times compounding the issue[1]. This is especially problematic when time is of the essence.

A Single Source of Truth

I bet the word blockchain popped in some people’s heads (and surely caused some eyes to roll). Full disclosure, I am a believer in blockchain, and ─obviously─ think this could be a great use case, but please feel free to substitute blockchain for the cloud (or any other network architecture for that matter) that feels better. I do believe that a blockchain might offer some additional benefits, but we’ll get to those.

Now, imagine that all of a person’s or business’s financial information (including say, tax returns) was stored and updated in real-time on a blockchain infrastructure in a standardized machine-readable form. Every credit card purchase, mortgage information, loan obligation, and bill payment is safely recorded. We could think of it as a massive spreadsheet on the cloud or over a distributed network that can be simultaneously used by humans over some graphic interface (similar to the Microsoft tool) or by computers in a suitable format.

Using a private key, people/SMBs could grant access to a financial partner on the network, say, to apply for a mortgage, or for a business to file an entire SBA application in minutes. Also, at every point in time, people/businesses would be able to see every financial partner that has access to their information, including type and frequency. These permissions could be revisited every time just as easily.

In turn, fintechs (or banks) could use that same source of consensus information to adjust their offerings and price risk accordingly. In our current world, a bank partner could place a download button to access and process the relevant information and authorize advances for SMBs quickly. For consumer-facing apps, this would help to make the transition towards an underwriting model, something that a few are doing, but that is not common yet. Chime could pre-qualify more of their customers and underwrite the $1,200 government check (or future payroll, hopefully).

Let’s think about this for a moment in the context of our initial discussion. We haven’t skipped any steps, but by continuously updating the blockchain, we have effectively automated and merged the processes of aggregating all relevant information and have it processed/validated. Thus, we moved or shifted both steps forward, allowing us to save a great deal of time. Now, it could literally take us minutes what today takes weeks. So far, so good.

Do We Need A Decentralized Network?

There’s the question, again, of whether we need a blockchain for this. The answer, still, is that we might not need one. Moreover, we mentioned how much of this infrastructure already exists. In fact, I think that the next step for Plaid could be to put a human-friendly, intuitive interface on top of their software to help people understand and manage their flow of information (similar, but broader in scope to the Excel case). More people/SMBs using a product as such would make the platform incredibly more valuable.

Yet, I feel there are still arguments to create a decentralized architecture around something like this. There are the benefits of transparency, immutability, and security that a distributed infrastructure can offer, but that’s all I’ll say about those. Then, there’s the more libertarian issue of thinking about whether we want to give a single entity so much control over our lives, regardless of how good its products are.

But more importantly, such open-source networks can become incredibly powerful avenues for innovation. Surely, it takes time to attract enough interest from developers worldwide, but over the long-term, more people building on it makes it more valuable. Moreover, the fintech community is already humming and thriving, powerful, and large enough to give it the critical mass to achieve network effects.

Furthermore, dissimilar to other open-source or decentralized network projects, we must underscore how much of the infrastructure has already been built, with multi-billion dollar businesses created around it. Thus, we would be potentially much closer to new killer use cases. Remember our introduction on a flexible tech stack? Well, this is not about building from scratch but rather repurposing it, sort of putting it on steroids.

The fact that it is decentralized means that it wouldn’t be bounded or controlled by any individual company’s agenda. It could propel new companies, products, and business models while allowing for far-ranging collaborations that will benefit customers and companies individually and as a society.

I think this is a compelling argument and one that goes beyond advancing any libertarian or blockchain-evangelist argument. If you believe all of this can be done better or more efficiently without the need of a blockchain, by all means, I take it and more power to you.

Smart Contracts As An Underwriting Tool

And there’s this, of course. Before people remind me about the DAO hack (which I don’t mean to underplay), I ask you to please remember that pretty much every other fraud in the modern history of humanity has had a paper contract backing it, written and revised by ─sometimes incredibly expensive─ attorneys. This is a well-documented behavioral bias called algorithm-aversion. It essentially says that we tend to be much more forgiving with human error, which also helps to explain why a crash involving an autonomous car grabs headlines, while the tens of thousands of human-made accidents rarely do. But back to our discussion.

As a general premise, cash advances as we know them shouldn’t exist, so it feels weird that they are a feature of a fintech offering. This feature is made possible by an inefficiency in the system, typically the timing window for processing deposits. Thus, we are not genuinely underwriting customers.

Again, my point here is not to underplay what Revolut, Chime, Current or similar personal finance companies do. Particularly in times like this, that several-day window provides much-needed oxygen. Moreover, I bet that if they had a certainty of repayment, they would happily underwrite more of their clients’ future cash flows, be these in the form of relief payments of paychecks.

Well, what if I could willingly enter into a smart contract with my fintech for future streams of cash? This obligation would, on principle, be self-executable and would be added to the decentralized ledger. Regardless of where I elect to receive my check, there’s an automated order (maybe in the form of a push payment) to transfer that money to my fintech provider the moment it hits that account.

I believe this could have tremendous implications in the future, as it would open the door for widespread underwriting model approach to consumer finances. We could start offering consumers similar products as the ones PayPal or Square offer by extending credit lines that are repaid from future card purchases.

In creating this middle ground between incredibly expensive short-term revolving lines and asset-secured finance (e.g., mortgages), we would substantially lower the risk (mainly that a person loses their job) and the cost for millions. Moreover, both in good times, but especially in bad times involving adverse scenarios, this could greatly reduce our response times. Just imagine that people who pledged a portion of their relief check could have received an advance the moment it passed into law.

Transparency is also enhanced in this marketplace, as other would-be lenders could see that this particular future money stream is earmarked, avoiding the said double-spending problem. In the case of a Government guarantee of the type offered through SBA loans could be documented as part of the credit agreement on the smart contract code. Thus, worthy individuals and business customers could access funds immediately, while transparency could limit credit overextension.

Lastly, another interesting topic that has gathered attention relates to the Fed creating a digital currency. That topic is rich enough in itself, so we’ll leave it for the next time, but keep in mind that implementing that would also change the paradigm of monetary policy and the flow of money.

Not A Lost Opportunity: Building A Resilient Financial Ecosystem

This is just but one way of imagining what the future could look like. A future that is grounded on technology and that is responsive enough to spare people and businesses from much of today’s sufferings.

A financial system can be leverage in both ways: either enhancing people’s lives or acting as a regressive tax on the most vulnerable and on SMBs, which are the backbone of the economy. No one knows when or how, but we can be certain that our system will be tested again in the future and shall be prepared.

[1] It would be an interesting exercise to quiz people on the type and frequency of information that each financial app on their phone retrieves, from purely transactional to meta-data.