Disclaimer: I write from the point of view of a tech investor and strategist but I am not an engineer and these are my personal viewpoints only. I welcome learning and growing from any challenging viewpoints!
For context I recommend you read the previous parts in this series. I have written about some of the challenges with today’s KYC and CDD processes in Part Two.
So with all the tail winds in the world to improve the KYC onboarding process, combined with an industry driven to cut costs and the technology to facilitate available to boot, why aren’t KYC utilities popping up everywhere?
First, a brief word on history of these things: the incumbent centralized KYC databases that exist today (eg. Thomson Reuters Worldcheck, DTCC, SWIFT ) actually did try to solve for this some years ago. But with centralized ownership and a competitive market came other problems. Financial institutions could not agree on a common standard, and with fragmented use of these vendor services across different firms and jurisdictions, any ability for these centralized databases to become the golden source of data, and therefore a true industry utility, was severely limited.
The next generation of KYC utilities is already a few years in the making and seems to be well under way. It’s decentralized, owned and operated by industry, often backed by government, sometimes blockchain-driven and will aim to use many of the aforementioned mentioned technologies to solve a large chunk of the operational pain.
A number of countries and continents have tried to launch a utility in recent years: Singapore (attempt number one), France and South America have all attempted, and there are signs of early stage efforts that seem to have limited traction so far in Hong Kong, Switzerland, the UK and Abu Dhabi.
We have a good window into the challenges in Singapore. The KYC utility was shelved in 2018 as the setup cost exceeded the savings the banks would have achieved, and the challenges have been well documented. Many of the challenges encountered revolved around process rather than technology. There were issues with policy harmonization — for example, the same legal entity across 5 different countries will require 5 different sets of procedures. Further issues arose from the simple fact that a standardized process can often be more expensive than one that has been well-optimized for a bank’s specific business needs. Similar issues were no doubt driven by the customization required to deal with policy harmonization across business lines internally within a firm.
Overall it seems transformation costs weren’t sufficiently being taken into account; neither was the expectation by business of achieving a margin on that transformation. While a utility is designed to simplify, it can actually end up making the entire process more complicated. Add to that the need to hire a team (expensive human capital) for intervention for handling manual exceptions, and the costs begin to add up quickly.
Simply put the fall of KYC utilities to date has not been a tech problem but rather a mismatch of business requirements. Participants couldn’t agree on scope and commercials fell over on the lofty transformation costs.
So why are others now beating down the same door? Attempts for the trophy of first successful country level KYC utility in 2020 include: Singapore (attempt number two) and the Nordics. There are some signs to suggest that lessons have been learned and these initiatives are trying a different approach. Rather than try to solve for all shapes and sizes, they will start with standardization of simple processes common across banks. Rather than focusing on a utility to automate the entire end to end KYC process, they will focus on specific common problem sets and industry or country-level wins — for example bringing ownership of credentials back to the individual, such as with the use of digital passports.
As an strategist I look at the painful parts of the KYC process that are common among financial institutions, and try to match these against new technologies that have solved for that specific problem set. If the utility will introduce more complications in the form of transformation, it will fundamentally cost more. Instead, a successful utility might be built from focusing on automating common processes and small but significant, and commercially viable, wins. All this while always taking into account the end customers’ requirements, which in this case includes both the financial institutions and the individuals. Remember, if just 75% or even 50% of the KYC operational pain can be solved with a utility that’s a huge cost saving and therefore a huge opportunity for the market!
There’s a lot of tech out there that solves for other parts of an end to end KYC and CDD process that a utility might adopt: machines learning and artificial intelligence for risk scoring, risk identification, and automating investigations, as well as data aggregation and network analytics tools. Assuming we have a fundamentally useful and commercially viable foundation from which to launch a utility, we can then look at the application of these technologies within the banks themselves to root out financial crime, identify new risks and reduce false positives.
In the next part of this series, I will dive into the wide world of Machine Learning and Artificial Intelligence an it’s looming impact on all areas of financial crime.