Will AI be the catalyst to drive wider bank IT transformation?
19 April 2017
Widespread adoption of artificial intelligence and its machine learning sibling could be the catalyst that finally drives more strategic and sustainable transformation across banks’ ageing and complex IT infrastructures.
The impact could be enormous, with technology now able to deliver significant benefits for banks and customers, despite - and not just because of - the large number of jobs forecast to be displaced.
For many beleaguered banks the focus on taking costs out of unwieldy and inefficient IT systems has so far been mainly motivated by releasing funds to bolster battered bottom lines. But real and lasting benefits would become more obvious if a more strategic approach was adopted.
It is certainly not an easy decision for bank executives to take, particularly when meaningful tangible benefits might be some way in the future. But there is increasing confidence that capabilities like machine learning (ML) and artificial intelligence (AI) can now be applied with longer-term revenue growth in mind instead of just delivering tactical cost reductions.
It is not that AI and ML are particularly new, as both have been around in some form or other for decades. But the emergence of new powerful databases and more sophisticated analytics, both operating at much faster speeds and at vastly lower costs, are finally starting to deliver the kind of outcomes that have long been promised.
Some estimates suggest over 200,000 jobs could go in the next few years, across buy- and sell-side institutions, as a result of AI automations. However, new jobs will also be created, or individuals retrained and re-deployed, in roles that complement and strengthen the capabilities being offered to customers by the new technologies.
It is no surprise that JPMorgan Chase & Co. (JPM) again appears to be showing the industry a clean pair of heels as it not only outspends its rivals on technology but outsmarts many of them as well. Its near USD 10 billion annual IT budget, which at some 9% of revenues is around double the industry average, is being re-balanced to use automation to take costs out and redeploy it to areas where it improves competitiveness and helps customers. The banks’ executives figure that in the longer run the bottom line will look after itself.
Examples of AI are spreading across the industry from east to west and from banks to fund managers via trading venues. Wells Fargo & Company (WFC) said it is developing AI technologies across its payments business to improve financial product recommendations for customers, while also protecting against fraud or misconduct. In its wealth management division Wells Fargo is also currently testing on staff the use of robo-advisers to provide investment and savings advice to younger customers.
The Tokyo Stock Exchange is working with vendors to use AI capabilities to determine and prevent abusive trading practises. It expects the benefits to allow more specific and detailed investigations by compliance staff. Many more examples now proliferate.
Across the buy-side asset managers are adopting robo-advisers with almost competitive speed, particularly for asset allocation and stock selection in passive funds. But the boundaries are going to be pushed even further as more sophisticated algorithms are employed to extend computer driven investment strategies at lower price points to wider investor audiences. It will be interesting to see how regulators respond.
There are many more examples. But in JPM’s case its efforts are focused on two AI initiatives, COIN and X-Connect, both supported by its private cloud initiative, Gaia, and which together aim to drive swathes of automation across previous manual activities. It will also enable clients to engage directly with the JPM technologies via the Gaia cloud to access a range of services and trading tools.
JPM has made no secret of the fact that only a third of its $10 billion annual IT budget was available for new initiatives. Most other banks have much less than that. But the focus of these initiatives, as well as cost savings from the retirement of old technology, will rebalance budgets and enable funds that can be directed at competitive innovations to rise to above 40 percent of the total.
And if AI fulfils the promise many now expect, those transformative capabilities could have much more significant budget implications than these incremental percentage changes. There is no reason why the cost of IT operations to run the banks should not be below 50% of the total.
In fact, like blockchain (if it too ever reaches critical mass), any transformative impact on operations and jobs will not be confined to bankers. The legal profession is already cutting headcount in its lower echelons and those cuts could go much deeper as the twin impact of distributed ledger technology (DLT) and AI remove the need for layers of independent authentication and verification of transactions.
Capital markets customers are recognising the benefits of these technologies and are expecting financial software vendors to integrate these capabilities into solutions moving forward. Like JPM, Misys sees this development as just one aspect of the wider transformation of the investment banks’ IT landscape. Eventually this will embrace the full spectrum of cloud, open architecture and application programming interfaces (APIs) that enable everything to be delivered “as-a-service” to banks.
The prospects have been trumpeted by the data scientists and innovative developers for some time, but have often either fallen on deaf ears or received with healthy scepticism. This time there does appear to be some momentum building that banks might not only be on the cusp of buying into this with more conviction, but are ready to start making more significant adoption commitments.
The other burden banks have had to shoulder has been the massive increase in regulatory compliance. There are signs that this too could be peaking and the longer-term trend could be towards more regulatory consistency and harmonisation, particularly as more regulators are being openly more collaborative both with their banks and with each other. It is an area where more widespread use of automation could start to reduce some of those cost pressures and begin to reverse the explosion in compliance teams that have been such a recent feature.
All these new initiatives, again like blockchain, will also require more consistent global standards. The industry has invested so much recently in better levels of governance, risk management, client relationships and technology that it is unlikely to want to reverse it. In fact, as the JPM reference suggests, industry leaders will continue to spend more.
So if bank balance sheet restrictions are relaxed and risk appetites can reverse the aversions of recent years, the cost of capital should decline. The concurrent combination of rising margins as interest rates edge higher, and as other costs fall, should mean more sustainable profitability will develop. Some banks will be better positioned to take advantage of this.
But there are clearly still too many banks and an eventual consolidation sector must also surely occur. For example, it is not practical for Germany to be home to over 3,000 bank. Therefore, those that have prepared for the new ecosystem with greater automation that delivers sustainably lower operating costs, along with the digital platforms that produce smarter outcomes for customers will clearly be among those best-placed to lead the survival race.