How insurgent financial services institutions are overtaking the giants of the sector

In a hyper-competitive environment, it is very often a company’s ability to adopt new technology and turn it to profit quicker than their competitors which makes the key difference. A company’s ability to deploy blockchain, migrate to the cloud, and develop effective Machine Learning (ML) or Artificial Intelligence (AI) models often indicates who succeeds and who falls behind. Financial Services are now in the midst of just such a struggle.

Financial services have long been early adopters and pioneers of new technology. It’s a sector which combines the voluminous budgets, tight regulation and high intensity competition that often forces them to the forefront of innovation. They were among the first to really seize hold of technologies like Mainframe computers in the 50s and 60s to handle their oceans of transaction data. Moving ahead, they pioneered customer and data analytics to gain competitive edges in serving their customers. Technologies like Blockchain, mobile banking and digital wallets have marked recent years in the sector. It should be expected then, that automation and AI would be pioneered quickly and adeptly.

Yet many financial services firms are struggling to do so. That’s not the case across the board. In fact, its larger well-established firms are struggling, while smaller organisations are leading innovation efforts in the sector. 

Size and success is a double edged sword

The main cleavage in financial services between those that can successfully make use of AI and automation and those that can’t is sheer size. The reality is that larger financial services are loaded with age-old processes, departmental silos, back-dated data and legacy technologies which make automation and  AI development exceedingly difficult for many. That unwieldy size creates a whole range of downstream problems which prevent many larger financial services institutions from rolling out projects. 

Really big data

Financial services live off of data - actuarial data, customer data, personally identifiable details, market projections and so on. In larger firms, different kinds of data will be used for different metrics, collected across different departments - and possibly geographical regions - for different purposes. It was likely collected at different times and stored in different places with different technologies, for different time periods. This is the first major problem that many will come to understand in developing their own AI and automation deployments. AI and automation require good, reliable and consistent data in order to perform and the fragmented operations and practices that characterise larger firms make that significantly more difficult. 

Legacy technology

That fragmentation acquires yet another aspect when we start thinking about the legacy technologies which collect and use it. Despite their reputation, as pioneering early adopters, larger financial services firms still deal with a lot of legacy. This could be because these are load bearing legacy tools which many other parts of the organisation depend on, or it could merely be the favorite piece of kit of a particular specialist or department. In any case, this furthers the fragmentation that is hobbling AI or automation. These pieces of technology will have their own dependencies and metrics and often don’t integrate with other technologies, much less the frameworks and models that automation and AI consist of. This ultimately has the effect of forcing “small pocket implementations” and deepening the silos which hold larger financial services companies back.

This just leads to a situation where you have a large organisation, with multiple different departments, collecting different kinds of data with different tools and technologies. On top of that those tools will often not integrate together and the department will often note share data. It will be prohibitively difficult to build an automated system on top of that, much less comprehensive AI systems. 

Compliance 

Ironically, one of the key reasons financial services want to start using AI and automation is to help them comply with regulation. The financial services industry is closely watched by all manner of regulations which cover fraud, insider trading, data privacy, market fairness and more. There are national regulations set by individual governments - such as the UK’s Financial Services and Markets Act - along with sectoral regulations like PCI DSS. Given the international nature of financial services, it's likely that any sizable firm has to comply with regulations in multiple jurisdictions and across multiple borders. These requirements are both wide-ranging and thorough, threatening heavy penalties for non-compliance. Not only do these regulations have to be complied with but efforts and records have to be extensively documented which is both yet another difficulty that financial services firms have in complying with regulations, and another reason they want to start using automation and AI. 

On top of all of that, larger companies have to deal with both the regimes they’re already compliant with as well as those which they’ll have to in the future, paying special attention to those which look at AI such as the EU’s forthcoming Artificial Intelligence Act. 

Clearing House

The difference between larger, old financial services institutions and smaller, faster ones isn’t hard to imagine. They are essentially dealing with brownfield sites in which they have to fundamentally restructure some of their most basic processes. Smaller counterparts - on the other hand - have greenfields to build on top of - allowing them to build faster and leaner. This doesn’t just go for AI and automation, but all kinds of innovations too. Online banking and digital wallets were first rolled out by smaller start-up firms because they could do so without the incredible overhead that accumulates with time and, ironically, long-standing success. 

Larger financial services firms will have a much bigger job ahead of them than their smaller counterparts, but they also need it more. They need to compete with a new generation of firms who can move with an agility that’s currently hard for them and regulations are bearing down on them which are forcing them to innovate. That’s going to involve tearing out and replacing some of the most basic systems, processes and technologies that they rely on. From the root, those larger firms will need to change some of the most basic ways they automatically collect, validate and manage data from across their organisation. This will be especially important for compliance, which often has to be thoroughly documented to prove that compliance. 

Technologies like automation and AI can’t just be bolted on to pre-existing systems and processes. They offer profound benefits and they’ll require profound restructuring in order to realise.