How feasible is automated regulation?
The global investment banking industry is worth a few hundred billion dollars annually, as are both the audit and legal professions. And since the last decade or so, increased regulation has forced banks to devote around 10% of their salary costs to employing an army of compliance controllers to ensure that their transactions and processes meet the standards required by the law.
And the stakes are high. Rogue traders, breaches of confidentiality, and reckless financial positions can expose financial institutions to fines, cripplingly negative publicity, and even prison sentences, not to mention huge financial losses.
These stakes are what make banks the earliest adopters of many technological innovations. Banks are turning to Regulatory Technology (RegTech), chiefly Artificial Intelligence (AI) and Augmented Intelligence (IA) but also other developments in computing like blockchain, to reduce their spend on keeping compliant with regulations.
What is RegTech?
This field is using fairly cutting-edge developments in computing that mainly revolve around computers’ ability to recognise patterns in vast patterns of data, and then automatically and iteratively refine the models that the computers use for output (machine learning). So unlike conventional paradigms of computer programming which use a set human-defined processes to generate ouput, AI “learns” as it works, generating more refined decisions based on the vast amounts of data it processes. According to Hitachi Vantara’s Global Head of Financial Services and Compliance Nirvana Farhadi, businesses need to think about taking an outcome-oriented approach first.
Financial firms have a vast number of regulatory reporting obligations which are a burden, both in time & resources. @NirvanaFarhadi from @HitachiVantara explained to us how #RegTech can solve these challenges & reduce cost of #compliance. #FinTech #FinServ #MiFIDII #TechNativeTV pic.twitter.com/DMOjVFbps3
— TechNative (@TechNative) December 4, 2017
So what is IA?
IA is a related hybrid paradigm which uses the same mechanisms as AI but then defers to human users’ direction for its final output – think of it as AI-supported decision making.
RegTech is particularly well-suite to these concepts since it involves sifting through hordes of documents, communication, and other data sources to spot irregularities or unstandard patterns, and subsequently build on the models it uses to recognize these patterns using the observations it has made. Thus RegTech spots fraud and non-compliance quicker but also gets iteratively quicker over time.
This offers both quantitative (faster) and qualitative (better) improvements over manual compliance checking. Furthermore, IA can provide the best of both worlds in that both experts and supercomputers are consulted to evaluate compliance.
Aside from machine learning, other technologies like blockchains (distributed ledgerlike databases that ensure that different nodes on a network all have the same copy of a store of information, and even biometric data are all being used to increase compliance with regulation. However, “if you put trash in, you get trash out” as Nirvana Farhadi explains in the second part of our interview.
In #FinancialServices, the quality & residency of #data, as well as having a good data discovery process “is key” to regulatory reporting says @NirvanaFarhadi of @Hitachi_UKI. #FinServ #GDPR #MiFIDII #RegTech pic.twitter.com/PV0mQ9RPx2
— TechNative (@TechNative) December 8, 2017
Who is using RegTech?
Most notably, banks are using RegTech to cut the huge wage bills they have had to pay for professionals to keep them compliant with increasingly strict regulation. Citibank, for example, employ around 30,000 people in this function, so any dent they can put in this cost would be welcomed. In the second part of our interview with Nirvana Farhadi, she explains why RegTech is much broader than financial services.
Think #RegTech & #FinTech are the same thing? @NirvanaFarhadi from @HitachiVantara disagrees. Watch the fourth part of our interview to find out why. #compliance #regulation #MiFIDII #GDPR #TechNativeTV pic.twitter.com/1qBBnZgXcu
— TechNative (@TechNative) December 11, 2017
Aside from banks, auditors play an important role in meeting compliance standards, and since they also consult on tech implementations for banks they will be at the forefront of both using RegTech solutions for their own audit practices and also applying it to their clients’ setups.
Most interestingly is perhaps the link between RegTech and FinTech companies. FinTech refers to the broader trend of tech-leveraged mobile payments, banking platforms, and investment solutions. Over the past decade, FinTech has opened up the market to many startups competing with established players to provide both tech and end-user solutions. One area that these smaller companies struggle to compete with the incumbents is in compliance – one talented developer can code a web app but there are often still resource-intensive compliance requirements. FinTech startups are leveraging their technical advantage to implement RegTech to reduce the barriers to competition even more.
RegTech can have a big impact on the bottom line – if savings of 20% of the compliance expenditure can be achieved for financial istitutions this translates to a huge 2% reduction in their overall operational costs – in an industry as large ad finance that makes RegTech worth several billion dollars a year.
Secondly, the more hard-to-quantify savings resulting from the reductiom in risk of fines and loss of business from compliance failures could be even more lucrative. However, to get to this point you need “end to end stakeholder buy-in”, notes Nirvana Farhadi, in the final part of our interview with her.
In the second part of our interview with @NirvanaFarhadi, she explains why #FinancialServices firms need “end to end stakeholder buy-in” to get #MiFIDII compliance right, and how #automation & #RegTech can help. #FinTech #TechNativeTV @HitachiVantara #MiFID #FinServ pic.twitter.com/Y3sDcZG3OD
— TechNative (@TechNative) December 4, 2017
The rise of robot compliance officers and accountants
RegTech goes beyond modelling non-compliant behaviour. Firstly because of how it works; machine learning in particular learns based on quatities of data that one human couldn’t see in a lifetime. This will result in emergent findings whereby compliance officers actually refine their expertise based on what RegTech uncovers.
But secondly, the other aspects of RegTech like blockchains will make the input data for compliance decisions richer and more secure, meaning the whole regulatory equation will be improved from input to output.
