Legacy banks are great. I say that as someone who used to work for one (for the record, it was Nordic banking leader Swedbank, with net income just over €5 billion). Behemoths like Swedbank are geared up to deal with millions of consumers and include some of the biggest corporations on the planet. That’s heavy duty. 

However, typically they are not so well set up to lend to small businesses, and this is a worrying gap, as SMEs represent about 90% of businesses and more than 50% of employment worldwide. 

Many of these smaller businesses are new and can’t meet the usual collateral or trading record requirements of legacy banks. And conventional credit scoring methods and scorecards, still in use at most legacy banks, are inadequate for the sector and only serve as another barrier to growth financing faced by SMEs. 

Startups - which all begin life as SMEs - are crucial for future economic growth and vitality. So it’s troubling that access to working capital and growth capital is potentially choking off entrepreneurship.

Manual banking processes exclude smaller startups 

The basic problem is a massive mismatch between legacy banks’ processes and technology, on one hand, and the size of the SME market on the other. 

Given that they are set up to deal with consumer-scale operations, this might seem counter-intuitive. But legacy banks deal in consumer markets that are so large that they can afford to cream off the best credit risks and still have a very sizable business. And they have, mostly, invested in the digital automation needed to service these customers. 

When it comes to the SME sector, they also skim off the best risks, where they can handle a few larger, well-established SME clients using manual account management. It’s an expensive approach but the selected clients justify the outlay. Without any meaningful automation in place, smaller or newer businesses are simply not attractive enough.

It hasn’t been a massive problem for the legacy banks. They were making plenty of money as things stood. However, that is changing and the SME market is looking increasingly enticing.

The need for automation and AI

To service SMEs and make a profit, legacy banks are going to have to make huge efforts to integrate automation and AI so that they can reduce the cost of assessing and managing risk, and servicing relationships.

That looked out of reach until recently: the cost of new IT development and integrating it into banks’ complex core systems was just not realistic. 

But times change. APIs and embedded finance theoretically make it possible to add any banking capability - creating super apps, in the industry jargon - to your platform without the time and cost of developing the code.

This is fine for the so-called neobanks, which generally started out offering customers a limited set of banking options but soon found themselves in an arms race to expand those capabilities to keep up with competitors. Revolut, for example, quickly developed current accounts, FX, payments and cards, and is now spreading its reach out into decentralized cryptocurrency wallets, the mortgage sector, and, most recently, expense management, as it pushes ahead with its own strategy to deliver a super app.

And others that started out as banks have seen the opportunity to become API developers/providers and service this market opportunity. Take embedded finance platform Solaris, as an example. It started out focused on payments, then added a banking license, and now offers a widening range of BaaS APIs to help neobanks build their platforms more rapidly. 

This flexibility of new digital players is causing the old-school some pain. In Scandinavia, the incumbent banks have 80% of the market. It’s still an oligopoly but things are changing: Revolut, for example, is now number three in the Baltics. 

Legacy bank’s are mostly too complex for APIs

From a legacy bank’s perspective, however, the API revolution sounds too good to be true. Indeed, it is too good to be true.

If they want to target the SME market, it’s almost impossible for legacy banks to deploy an API strategy against their new competitors. Why? Their IT systems are just too complex. Nordea has 40 core systems whereas neobank Starling has just one. 

With just one system to think about, embedding a single API is relatively easy. With 40 plus systems it’s a different story. The interplay and potential interference between an API and any of these core systems is hard or impossible to predict. And, even if you could sort that all out, every time the API is maintained and updated by its developer, the same scope for instability creeps in again.

Then start to imagine doing all that with multiple APIs. 

And you have plenty of choices, as a quick look at the API bible, ProgammableWeb, demonstrates. But not all embedded APIs are “API first” and may lack stability as a result. Nor does integration magically vanish as an issue with APIs, the more you add, the less likely they are to work faultlessly together. 

The big banks recognise the challenge but they also want a piece of the SME pie. Knowing that there are limits to what they can achieve in their core businesses, they are investing heavily in fintechs and neobanks. To take just one deal, HSBC has invested in Monese. And they are also investing in fintechs offering embedded finance to understand how it works and have a piece of the high growth. ABN AMRO and BBVA are investors in Solaris, for instance.

3 traps for legacy banks to avoid

So, to summarise, if the bank where you work is considering the SME market and has an API-led approach in mind, what are the three ways to avoid potential traps that might derail your strategy? 

1. APIs sound exciting - they are exciting. But be realistic about whether they are suitable for your IT environment. In many cases you would be setting up for failure.

2. Investigate alternative ways of accessing the SME market and/or delivering relevant products and services. Partnerships and investments with nimbler players in the space are obvious routes here. 

3. If your organisation nevertheless decides to proceed, despite having multiple core systems, set expectations internally correctly. Your C-suite colleagues need to understand that, yes, things might proceed swimmingly. However, there is a more than reasonable chance that there will be delays, cost overruns and, worst case, system failures that could result in regulator penalties.      

If that sounds excessively pessimistic, consider that a recent survey of over 2,000 API initiations across 31 European countries found that 38% of bank APIs don’t meet EU or U.K. regulatory standards and 28% of APIs experienced downtime during integration. And, while most banks in Europe now offer “API availability”, in Czechia, for example, only one-third of banks’ APIs actually work and the European average is just 50%.

If APIs have the potential to operate as a magic bullet solution for banks seeking new market opportunities, like expansion of SME lending, they are not yet reliable enough to take you there without many additional, probably unexpected complexities and overheads.