Banking automation and an accelerated transition to a new business model

Banks prepare for the automated future with cognitive computing and robotics designed for no touch customer interactions

Banking’s labor force moves to higher-value, customer-focused interactions
A new industrial revolution is happening around us, and the financial services industry will change to reflect it. Bank business models, workforces, processes, and services will need to transform radically to reduce costs, reduce risk, and enhance experience and value for customers.

Robotics, machine learning, and blockchain and their place in the automated future
The mission is to deliver digital customer journeys and eliminate unnecessary interventions in the most routine, repetitive tasks. After all, manual processing can create delays, lead to error rates as high as 30 percent, and increase rework volumes due to employee fatigue and poor training.

Why is this transformation necessary? Banks must respond quickly to economic and technological advances that are closing their jaws on the industry benchmark cost/income ratio. Revenues are under pressure, interest rates remain low, and new entrants and fintechs are eroding valuable customer segments, as high-value services–historically the preserve of long-standing global and regional bank brands–are stolen by months-old startups. The impact of open banking is being felt as the API economy matures, and regulators drive competition.

On the cost dimension, there’s a new sense of urgency in the boardroom, focusing on cost takeout and productivity. Regulation continues to expand, adding cost. In some markets, resurfacing inflationary pressures point to additional cost pressures. At the same time, customers expect increasing digital capabilities such as virtual self-service and immediate and transparent execution of transactions. These expectations, coupled with the increasing ubiquity of mobile, often challenge typical bank operating models.

New technologies–such as robotics, machine learning and blockchain–have C-level executives asking, “What should the bank operating model and ambition for a competitive cost base be for 2020 or 2025? What is achievable with these technologies?”

Cognitive process automation breaks through the productivity ceiling
Until now, banks have pushed against a productivity ceiling, with some retail banks achieving around 52 to 47 percent cost/income ratios (although lower in China and higher in Brazil). Cognitive process automation (CPA), which combines cognitive computing and robotic automation, can help raise that ceiling significantly.

However, the question remains: What is the winning model for a large retail bank? Is a cost/income of 37 percent achievable and sustainable with these advances in cognitive computing and robotics? The answer is probably yes, but is that enough? Or should banks set a more ambitious objective–perhaps 25 percent?

The cost equation, of course, has many components. Staff costs are significant, but property/asset costs and the cost of capital are also considerable. An aggressive goal for the cost/income ratio drives the imperative for greater efficiencies.

Historically, banks have been good at “engineered efficiency.” Strategy and key performance indicators are cascaded through budgets to lines of business (LOBs). In turn, LOB executives execute projects in areas such as processes, channels, domains like IT, and compliance. We believe that many banks will continue this approach by executing cognitive automation programs through these mechanisms in an attempt to sustain a general industry culture of control through product profit and loss.

Bookmark this report  

Meet the authors

Mike Hobday

Connect with author:

, Vice President, General Manager, IBM Automation, Europe, IBM Consulting

James Cook

Connect with author:

, Global Practice Leader, Talent Development Optimization, IBM Consulting

Nicholas Drury, Global Banking & Financial Markets Leader, IBM Institute for Business Value

Download report translations

Originally published 21 August 2018