On 27 June 1967, a crowd of excited onlookers watched as the world’s first ATM (Automated Teller Machine) was unveiled outside a branch of Barclays Bank in Enfield, north London. English comedy actor Reg Varney was there, brought along to do the honours of making the first cash withdrawal.
For many of those present, it would have been their first inkling that an everyday financial transaction could be carried out automatically – without a cashier or bank official in sight. And, in the years since, the cashpoint has delivered to those with high street bank accounts the best of what automation has to offer: universal availability, 24/7 functionality and freedom from human error.
Was that June morning a new dawn for the financial sector? Were the forces of full-scale automation just around the corner? In some ways no. After all, physical banks still exist and people still queue up in them to cash handwritten cheques and deposit bags of coins, just as they did fifty years ago. On the other hand, thanks to the advent of online banking and the evolution of accounting software, a number of financial processes are typically now either completely or partially automated.
Within financial systems, as in all systems, automation most naturally recommends itself to processes that are recurring, predictable and rule-based. The establishment of regular payments into and out of a bank account is a perfect example of such a process, and standing orders and direct debits have been around for some time (the latter developed in the 60s for consumer goods giant Unilever to expedite payment from ice cream vendors).
Mortgages, rents and utilities are all now easily and preferably paid for automatically. For those who can afford it, the automation of further regular payments into savings accounts, including retirement funds, is more than just convenient: it guarantees the growth of those funds in a way that can be difficult for people, without that self-imposed structure, to sustain.
Automating more complex financial processes – especially those associated with the work of accountants – has been directly dependent on advances in specially-written software and online resources. And while an increasing number of programs are assisting people with their personal finances (in areas like the preparation of tax returns), it is undoubtedly businesses that have to date been the principal beneficiaries.
These days, for example, few company bookkeepers do not take advantage of Microsoft Excel’s macro functions. These automate the kind of simple but data-heavy accounting tasks that, done manually, consume time and are vulnerable to mistakes, from completing journal entries to reconciling bank statements.
Since the popularisation of Excel, software technology has sought to go beyond a conventionally fixed, linear approach to task-automation. Robotic process automation (RPA) in particular has been developed as a way of having software mimic human work dynamically, intelligently and across multiple applications.
Thanks to RPA and other IT technologies, a 2017 report by the McKinsey Global Institute estimated that over a third of business finance activities can now be effectively automated, including issuing and checking invoices, processing expense-approval requests, gathering data for analysis and the compilation of reports and forecasts.
Financial departments whose operations have embraced this level of automation testify to a number of key benefits – none of which will come as a surprise to the automation engineer.
In the first place, these organisations enjoy a quality of smooth functioning unknown to old-fashioned accounts departments with their weekly and monthly bottlenecks of labour. When work – making payments, sending out invoices (and then reminders of invoices), approving claims for expenses – is spread out in real-time, delays, and delays caused by delays, no longer interrupt performance.
Secondly, as financial departments become more automated so they necessarily become more integrated – and therefore both better co-ordinated and less opaque than previously. This may mean anything from eliminating discrepant styles of bookkeeping between employees to gathering together the data for a company’s total expenditure in an easy-to-understand location on a dashboard.
Which leads, as always with automation, to the standout dividend: the use of information to generate insight. The kind of comprehensive data-picture of a company’s payables and receivables that would be too time-consuming and too complex to assemble manually can, especially in combination with forecasting tools, lead to a smarter mobilisation of working capital, reduce dependency on borrowing and enhance both competence and credit.
If automation has been shown to improve the efficiency of a company’s accounting department it has demonstrated no less prowess on the stock market. So-called Robo-advisors – investment guidance programs that rely on algorithms rather than human judgement – are tasked with controlling the content of individual portfolios in line with principles derived from Modern Portfolio Theory, reducing risk through the spread of asset allocation and continuous rebalancing. These automated investment packages often outperform actively managed funds, and at a fraction of the cost.
Developments in Artificial Intelligence are behind the most recent – and, to date, the most fundamental – ways in which financial systems are becoming more automated, and less subject to direct human input. Applying for a bank loan, to take another example, once a case of person-to-person negotiation is now more likely to be a question of satisfying a complex of entirely algorithmic demands (though, interestingly, there is research to suggest that such algorithms are not as free from human bias as one might imagine).
Most agree that current trends in financial technology are making money behave more efficiently – and that this efficiency is good for the overall economy. Some have voiced reservations, though, about the possible consequences of driving forward with automation in this sphere at too unchecked a pace: from the simplification of spending mechanisms (such as online one-click ordering) in societies already prone to overconsumption; to anxieties over data security and privacy in a diversified financial marketplace; to the subtraction of the human touch from, say, a review of a loan repayment schedule’s terms.
There is no doubt that automation’s power to accelerate, streamline and integrate financial services is here to stay. But that branch of Barclays on Enfield High Street is still there and has no immediate plans to close.