With COVID-19 still an evolving story (given the variants) and physical (social) distancing here to stay and lockdowns/voluntary lockdowns continuing (on and off), supervision by central banks is undergoing a sea change. Regulation too will undergo change to (permanently) allow for board meetings and board sub-committees to meet and function virtually, as may be required.
The question is how will (regulation and) supervision of DFIs and FIs (i.e., commercial and investment banks, financial conglomerates, microfinance banks, NBFIs, MFIs and FinTechs etc) now be conducted by central banks and/or designated supervisors and examiners.
One, central banks will now be forced to rely more on off-site and online supervision. On-site supervision is going to be much lower. This will at least be true of the short-medium term (at least 3 years going forward).
Two, given the (phenomenal) size of DFIs/FIs (and especially, the nature of MSME, microfinance and agriculture loans and other loan assets), off-site supervision is going to be a challenge across the board. The silver lining is the fact that board meetings will be virtual and recorded. Hence, central bank supervisors will be able to see governance and risk management processes in action. Overall, reduced off-site supervision will impair effectiveness of supervision by central bank supervisors and examiners, but can be greatly enhanced by use of technology and better directives regarding governance to DFIs/FIs as described hereafter.
First, let us look at the use of technology through RegTech. Given that off-site and online supervision are going to play a greater role, central bankers will require new skills sets and innovative RegTech tools. RegTech can play a meaningful role as outlined below, especially in the absence of on-site supervision and given the past crisis situations {like the U.S. subprime crisis of 2007/8 or the Indian microfinance crisis of 2010 in Andhra Pradesh (AP) and the recent financial sector crisis in many parts of the globe in 2018-19}.
It becomes evident that in all past financial crisis situations, central banks (most often the regulators and supervisors) were extremely ill-prepared for the impending crisis. While their incoherent responses in great measure added to the uncertainty and panic in the financial system, a major problem was that (most often) early warning signals were either not there or, at best, not clearly discernible. Sometimes these early warning signals were available but ignored and shrouded by conflicts of interest that were at play. This happened, in particular, because of the seamless exchange of people from the financial services industry to the regulatory domain and vice versa, through what is commonly called the revolving door and reverse revolving door phenomenon.
In effect, we had what can be called regulatory and supervisory failure coupled with a serious breakdown in accountability and ethics, from the grassroots level to the corporate boardrooms and regulators and supervisors, who (all) sat and watched as Icarus continued to fly high. In turn, this had an impact not just as (disastrous) financial consequences, but it also led to a serious erosion of trust in the financial system and all of its constituents, including regulators and supervisors, by the public at large (including customers, depositors, investors, end users and others).
Having set the context, we can safely argue that, as compared to the past financial crisis situation (especially the U.S. subprime of 2008), today’s institutions are larger and riskier. Therefore, we cannot afford to have regulatory and supervisory failure going forward. Part of the problem is that with the burgeoning growth of institutions and their portfolios, as well as clients, the information coming to the regulators and supervisors is what we would call information overload. The larger the bulk in compliance reporting, the lesser the chance of spotting exceptions and the greater the chance of mismanagement (or fraud) leading to a full-blown financial (systemic) crisis.
The larger point here is that, with many more institutions today and a lot of information flowing in, off-site and online supervision will become an even more difficult task. The continuance of COVID-19 makes central banks less able to use on-site supervision and so, they have to rely almost entirely on off-site supervision. Given this situation of central banks having to rely more on off-site supervision and with the maze of information coming in, off-site supervision will become difficult. Therefore, what can central banks do?
They can use smart RegTech tools that will help them pick up early warning signals about impending crisis situations in the financial services sector and information on the key aspects that caused the past financial crisis in the first place—facts like whether the compensation policy is rewarding the quick deal (short-term) when the risks are medium or long-term, or whether conflicts of interest are causing a DFI/FI (and other stakeholders) to go rogue, or what the real (hidden) leverage of a DFI/FI is, after taking into account all aspects such as off-balance sheet items, etc. All of these are applicable to the past financial crisis situations including the recent banking crises in many parts of the globe (2018–19), 2008 U.S. subprime and the 2010 Indian microfinance crisis in AP (these are just examples and not meant to be an exhaustive list).
To realize the full potential of RegTech (especially machine and deep learning), two crucial things will have to happen. One, central banks, as well as other regulators and supervisors, will have to remove any ambiguity in rules and circulars to their constituents. The discretionary power of the regulator and supervisor will also have to go so that the rules governing the game are crystal clear and not subject to interpretation and the whims and fancies of the regulator and supervisor. Two, machine and deep learning (AI), on their part, will have to develop the ability to discern causality (as opposed to mere correlation and association) to have greater predictive ability for regulators and supervisors. It is the fusion of deep domain knowledge of the financial sector and strong technical knowledge of machine and deep learning that can lead to practical, usable regulatory and supervisory tools with enhanced predictive ability.
RegTech is in its infancy and requires significant effort, from both regulators and supervisors and specialized information technology (RegTech) firms, before greater strides can be made concerning use of machine and deep learning in regulation and supervision (of financial services) and to specifically achieve the objectives set out above—i.e., provide early warning signals of key aspects going wrong and periodic information on exceptions in key factors, both with a view to pre-empt and prevent a financial crisis. Some areas for potential Regtech applications are outlined in part II of this article.
[1] RAMESH SRIVATSAVA ARUNACHALAM, a board member of the Financial Inclusion Advocacy Centre (FIAC), Ghana and UK and is also a partner is ASCENSION ADVISORY (India), under incorporation. He is the author of 14 critically acclaimed books. Ramesh also provides strategic advice on a wide variety of Financial Sector, Financial Inclusion and Economic Development issues. He has worked in over 314 assignments with multi-laterals, governments, private sector, Banks, NBFCs, DFIs, regulators, supervisors, MFIs and other stakeholders in 31 countries across 5 continents and 640 districts of India during the last 31 years. He can be contacted at [email protected] and +919962815615.
