Month: June 2018

It is perhaps an overwrought cliché these days to begin all serious contemplation through the digital lens. Nowhere is this truer than in the case of the BFSI sector, which has seen it’s fair share of upheavals in the last several months. Most notably (and perhaps tellingly), ‘Digital’ Fraud and Risk Management features high on the agenda of bank boards, as they grapple with new threats and new realities everyday. As monolithic structures threaten to crumble and further test the patience of the savvy investor and the regular consumer alike, it is prudent to try and understand the elements of the digital wave that can perhaps avenge the metaphorical Thanos in this new ecosystem.

Financial Risk Management (FRM) has historically been a core focus area of most progressive financial institutions that view fiduciary responsibilities as more than mere lip service. Deep in the bowels of the banking system, the Risk Management Committee often votes ‘aye’ or ‘nay’ on a potential loan/mortgage/advance by relying largely on years of honed intuition and a helpful dose of supporting data. What’s changed in this idyllic setting is the rapidity with which banking and digital boundaries have blurred, thereby throwing into the mix, difficult problems posed by Big Data and innovative methods of technological skullduggery that have evolved virtually overnight. To stay in business therefore, donning armor tailored in the digital factories of hyperaware tech giants and fintech firms may be the most meaningful solution.

At the frontline of this newly minted armory is Robotic Process Automation (RPA). While RPA’s home turf is in the area of repeatable and rule-based structures that require little human imagination, it has largely been implemented in operational efficiency improvements (purportedly, with some critics still insisting that the jury is out on this) and in back-end roles that have always had step-motherly treatment meted out to them. That said, RPA seems to have become more than just a digital buzzword, with several bankers viewing it as a potential panacea to the ills of human-controlled FRM. But is it?

Can it avenge FRM and save it from becoming a gangrenous canker that threatens to sabotage the system as a whole? That is a question that requires use cases to justify its utility, and forward-looking banks that can allay fears and doubts by boldly going where their predecessors feared to tread.

However, as is the case with every decision, one must remember that enthusiastic pros are often weighed against somber cons. RPA, by its very definition works its magic in structured, rule-based, repeatable scenarios that require little or no imagination. Brute force therefore, trumps creativity and ingenuity. But in the arena of FRM, increasingly weighty odds are being placed on swiftly changing scenarios, bringing with them new mountains of unexplored data and hitherto uncharted analytical territories that require intrepid and decisive action on the part of the mightiest heroes that the institutions can bring to bear on these problems. Thanos therefore begins to shrug his mythical cloak and becomes a clear and present danger that RPA may be ill equipped to handle alone. Where then does one turn to for additional ammo?

Just as the digital cliché is a convenient bandwagon to hitch prima facie arguments on, Artificial Intelligence (AI) and Machine Learning (ML) provide ample fodder for the ruminating analyst to chew on while contemplating solutions to problems mired in complexities that ‘mere automation’ is ill-equipped to handle.

AI/ML not only addresses the blind spot that an RPA solution to FRM causes, but also builds long-term resilience into the overall setup to ensure ‘future-proofing’ as far as such a concept is realizable with today’s technology. Complex scenarios can be mapped and the ML algorithms can learn from the data, develop parallel scenarios, understand underlying nuances between hitherto unconnected variables and develop a meaningful FRM business solution that banks can slowly pilot and then as technology, support and the ecosystem develop, deploy as a backbone to their overall digital offering.

Once again, it is prudent to look at all this in the context of the current scenario and try to delineate a pragmatic way forward. Presently the sector is turbulent, with flagrant customer and employee frauds, corporate governance lapses and an increasingly wary set of investors whose ROI, combined with the government’s political will could just be the decisive factors in the short to medium term fate of the Banking vertical in India.

Macroeconomic and political considerations aside, the microeconomics of prudence, choice and utility can best be encapsulated by Peter Drucker’s timeless pronouncement: “ The best way to predict the future is to create it.”

To predict the future therefore, we must look increasingly at deploying an RPA-first, AI/ML-next strategy that can bring back much needed credibility to a flagging FRM situation.