Open Banking: Is Availability the Key to Adoption?

by | May 20, 2020

When talking about IT availability, the goal is to get as close to 100% available as possible. Just as nothing in life is certain, nothing in IT is always available. Hardware and software eventually break down.

Nevertheless, companies can get very close to meeting the established expectation that tech should be always on. The holy grail of reliability standards is dubbed “five-nines reliability.” That’s 99.999% availability, which is the percentage of time during which a system is ready to be used when it’s needed. The calculation of availability accounts for any downtime whether something is broken, out of service, or unavailable due to routine maintenance.

As a percentage of the time, five-nines reliability equates to less than five minutes and 15 seconds of downtime per year. That’s not an easy mark to hit. Frankly speaking, five-nines are very expensive and very difficult to achieve as well as maintain. However, it remains the gold standard in availability and it’s exactly what you want in industries like digital banking. The closer we get to achieving real-time payments, the more likely that downtime will become universally unacceptable.

Availability in open banking

One of the biggest challenges of any new digital initiative, including open banking adoption, involves the question of availability. When a company implements a new process, there’s a natural learning curve and considerable room for error.

On top of that, it’s also incredibly difficult for traditional banks to change at all. Decades of ad hoc regulation and process requirements make up the foundation of these banks. They are not set up to adapt quickly—and so, they typically don’t.

Whether banks are ready or not, change is coming. In the UK and EU, mandates such as the CMA order and PSD2 legislate the open banking process. As such, EBA Guidelines 2.2 and 2.4 require two KPIs that ASPSPs, or account servicing payment service providers, should document and how to calculate them. Not surprisingly, those quarterly KPIs are for availability uptime and downtime.

To set a standard, the OBIE recommends a quarterly uptime of at least 99.5% and a quarterly downtime of less than 0.5%. That may not sound bad, but that is about 52 minutes of downtime a year.

At first, the UK CMA9 banks struggled with open banking availability as to be expected. They reported an average of 96.97% in Q1 2019, according to OBIE. However, that number is rapidly improving. As of March 2020, the average API availability is 99.14% and climbing.

As banks get closer to five-nines reliability by smoothing out their processes and upgrading their systems, we expect to see more banks in the U.S. get comfortable with the idea of open banking adoption.

Availability and the mainframe

When it comes to availability, the mainframe is king. One of the primary benefits of the mainframe is its knack for being reliable. Typical z/OS features, like the tight coupling of data structures, make the mainframe both fast and reliable. Two characteristics that naturally lead to high availability.

In fact, many mission-critical enterprise applications still run on z/OS today. An estimated 30 billion business transactions are processed on mainframes daily, including most major credit card transactions, stock trades, money transfers, manufacturing processes, and ERP systems. Today, 96 of the world’s top 100 banks still depend on mainframes.

However, if you asked business leaders what’s stopping them from fully embracing digital transformation efforts and open banking, they would likely say that legacy systems inhibit modernization. And that’s not completely wrong. While mainframes are extremely reliable, they do struggle to connect to modern applications and systems.

In a recent webinar, Alex Heublein, CRO of GT Software, points out that the biggest challenge in open banking is the vast amount of variability in terms of integrating the mainframe with open banking APIs. Both the mainframe and API standards, like OBIE and FDX, are incredibly reliable. The problem lies in the integration layer between the two.

Fortunately, there are tools to help securely unlock the mainframe, so that banks and financial services companies can have the best of both worlds. It’s absolutely possible to keep the reliability of your mainframe investments while interfacing with modern standards.