Data turns into a nuclear power

I-Say

technology column

David Birch explains how new ways of refining and processing data are transforming business models into ones that can offer better services and more efficient financing

The “data is the new oil” aphorism is usually attributed to Clive Humby, who famously helped to create the then revolutionary Tesco Clubcard loyalty scheme in 1995. In Humby’s view, given in 2006, data resembled oil because “it’s valuable but if unrefined, it cannot really be used”. He thought that, just as crude oil has to be refined to create petroleum products such as gas, petrol, diesel, kerosene and so on, so must data be broken down and refined (ie analysed) to create value. It isn’t really like oil though.

The data-is-oil meme continued to spread for another decade. In 2014, Wired magazine restated it with further historical context by saying that data is like oil back in the eighteenth century – a recently discovered and untapped asset of great value. But that view, that those who learn to ‘extract and use it’ will strike it rich, doesn’t work for me. Banks aren’t out on a frontier drilling for data, they are forming environments where data is created, managed and, until recently, farmed for their benefit.

The ability of machines to obtain insight and take action makes for a very different kind of financial services sector

James Bridle (author of The New Dark Age) put forward a much better metaphor, saying that data “more closely resembles atomic power than oil – an effectively unlimited resource that still contains immense destructive power and that’s even more explicitly connected to histories of violence”. From this perspective, data isn’t the new oil, it’s the new plutonium and, in the era of General Data Protection Regulation, personal data is the new nuclear waste.

Is there a way to refine and process the data to the benefit of the many? This is an area where the fintechs have something to offer. In an interesting paper on the subject in mid-2020, “Towards a data-driven financial system”, Nydia Remolina at the Singapore Management University began by noting: “Financial institutions have access to enormous amounts of data but, due to multiple constraints, this data is not yet sufficiently converted into useful insights.”

We know this to be true. Remolina goes on to talk about a new “data operating model” that brings together open banking, cloud computing, machine learning and artificial intelligence to support digital transformation.

Don’t roll your eyes at this point: in my view this model represents much more than the usual technology upgrade cycle, because the ability of machines to obtain insight and take action makes for a very different kind of fully digital financial services sector. As the paper also points out, some jurisdictions (including the US) allow non-bank online lenders that use these artificial intelligence/machine-learning models to play an important role by helping businesses that may not have an established lending relationship, ie many SMEs. When you combine this with automation of the onboarding and loan process, you have genuine digital transformation to the benefit of all stakeholders.

To be fair, what we might summarise as data-driven liquidity is hardly new. But what is new is the combination of innovative technology and regulatory frameworks that have come together to bring it into that SME space. Using open banking to obtain accurate and timely data on SME finances, then feeding that data into rapidly evolving machine-learning models, with emerging artificial intelligence capabilities, gives the financial services sector a better way to serve SMEs and to deliver better outcomes for stakeholders. This is another facet of the push for ‘financial health rather than financial services’ that forward-thinking strategists in regulators, fintechs and techfins are now talking about.

This puts fintech firms, and particularly data-driven lenders, in an interesting position. They will be able to demonstrate how beneficial the new business models can be for the economic recovery ahead of us. They can bring together digital onboarding, simplified customer due diligence and data on transaction histories to substitute for conventional credit scores.

A recent example of how access to data changes the range of services that can be offered comes from Liberis, which provides cash advances to SMEs. The financing is secured against customer card transactions and repaid when the SME has been paid. This strikes me as beneficial for all involved and illustrates how fintechs can use data as a practical substitute for conventional credit ratings.

The UK looks good in this regard. With a competitive fintech sector and open banking already in place, there is plenty of activity. I saw this for myself when I was asked to be one of the judges for the Open Banking Innovation Awards for SMEs. I was impressed by the businesses already taking advantage of this combination of new regulation and new technology. A couple of good examples are Fluidly, which plugs into accounting packages and bank accounts and uses machine learning to manage SME cashflow intelligently, and Swoop, which uses open banking to simplify access to all kinds of SME finance. Since SMEs are notoriously bad users of financial services, relying far too much on overdrafts and credit cards, supporting them is low-hanging fruit.

Crisis talks

As many people have observed, the Covid crisis has accelerated digital transformation. But we shouldn’t get complacent, if my own experience is anything to go by. I logged into my bank for something not relevant to this anecdote to find an attractive banner headline over my account, telling me that I had been pre-approved for a loan of, if memory serves, £25,000.

Although I didn’t particularly need the money (blogging and playing around on Twitter are much less capital intensive than most small business operations), I thought that it might provide a useful buffer against reduced circumstances in an uncertain world. So I clicked on the button labelled ‘apply online’. I naively assumed that I would be taken to a page saying “congratulations, the money is in your account” but instead I was taken to a page telling me to phone the bank, which seemed a rather 1970s version of ‘online’ to me, but whatever. Much against my better judgment, since phoning my bank is literally my last resort in any financial situation, I called. A robot answered and told me that I would be on hold for at least an hour (I’m not making this up). Naturally, I hung up and never called back.

Meanwhile, the economy began to crash and the need to get liquidity to support the UK’s small businesses accelerated. You can see the problem, especially in banks that have tried to layer a Covid response over a creaking infrastructure. Banks were getting inundated with loan applications, all of which had to be inspected and assessed in time to stop businesses from going under. SMEs struggled to obtain finance as concerns about fraud and defaults caused banks, which were in any case taking between four and 12 weeks to process applications, to restrict access to the scheme.

Something had to be done. In the spring of 2020, the government announced the Bounce Back Loan Scheme. This was a project to provide individual loans of up to £50,000 to SMEs much more quickly. It immediately went into meltdown. In autumn 2020, the National Audit Office estimated that the scheme will lend £40bn-£50bn, of which £15bn-£26bn will never be repaid. I am certain this is an underestimate because the less strict eligibility criteria, which relied on businesses self-certifying application details with limited verification and no credit checks, must have been a magnet for the usual miscreants.

The loans scheme became ‘a giant bonfire of taxpayers’ money, with banks handing out the matches’

Know your borrower

The UK is not alone in seeing an increase in fraud but one particular problem here is the lack of a digital identity infrastructure. Not only does no one know who anyone is, but with businesses there is the additional twist that Companies House has not yet been reformed to introduce proper checks on whether directors are people. Yes, that’s right: there is no check on whether a company director is even a real person, let alone whether he or she is actually the director of a company.

Somewhat predictably, then, the Bounce Back Loan Scheme turned into, as was said in the Financial Times, “a giant bonfire of taxpayers’ money, with banks handing out the matches”.

Perhaps in the next pandemic, the government and regulators will make more use of the data-driven liquidity approach. Even the most rudimentary artificial intelligence, given access to bank accounts, could have deduced in nanoseconds whether I am real, whether the company exists, what the turnover is and what loan I am entitled to. At which point, it could then offer the loan for auction to credit providers.

Just as Facebook auctions advertising slots, this could be done in fractions of a second and might mean that the big five lenders don’t get 90% of the business again. The bot could then take a small cut for its trouble and sit back and watch the economy get moving again in a fraction of the time and at a fraction of the cost.

Data isn’t the new oil, but transaction data – properly channelled – could be the new nuclear fusion.

David G W Birch

David G W Birch is an author, advisor and commentator on digital financial services. An internationally recognised thought leader in digital identity and digital money, he is a member of the CSFI Governing Council.