Moving our focus away from the business-centered issues regarding regulatory intelligence, we would like to face some more technical aspects related to this business sector, while our latest blog posts were meant to present the market’s situation and its numbers.
In our path so far we have talked about the RegTech market, focusing on its current dimensions, its growth prospects and its bright future, suggested by the constant increase of the compliance costs, coupled with the growing interest in technological solutions for this business area.
The status and future of the RegTech market are obviously related to the technical side. An essential issue in this field is the one regarding MDMER (Model-Driven, Machine Executable financial Regulations) which has recently been a topic of the Response by EBA to EC Consultation (June 2020). “Regulations would benefit from the translation into machine-executable form. A clear benefit is the elimination of the need by institutions to interpret the legislation if not sufficiently clear, or the possibility of misinterpreting it otherwise” – we read. Everything started in November 2017, when the FCA (Financial Conduct Authority) and BoE (Bank of England) enlisted experts within the UK Government and in the private sector and academia to create a proof of concept model to implement Model-Driven, Machine Executable financial Regulations (MDMER). This Tech Sprint has so introduced the MDMER approach as an answer to the need of monitoring and analyzing thousands of legal sources in real-time, in order to provide financial companies with a tool to keep up with the ever-growing regulatory changes and updates, but also to support the authorities in their supervision work.
The adoption of machine executable regulations would bring many benefits.
- Clarity: they cannot be as ambiguous as typical natural-language regulations.
- Temporal efficiencies: they decrease the time it takes to update, implement and monitor regulatory enforcement.
- Cost efficiencies: they generate immediate savings on the private sector hand (decreased need for expensive and time-consuming disambiguation) and a decrease in time of overall costs of monitoring and enforcing regulations on the government side.
- Change Management: enable regulators to more efficiently distribute changes in regulations to market actors and will enable regulated entities to more quickly adapt to those changes.
So, while currently the regulatory interpretation and disambiguation takes place after a regulation has been promulgated – in the “back-end” phase, so to speak -, the MRER aims to place this burden on the “front-end” with the regulators, with a complete disambiguation and immediate machine executability.
Obviously the adoption of machine executable regulations presents also many risks (incorrect disambiguation, errors in the code base, lack of flexibility, code base opacity, versioning challenges).
Apart from these hypothetical benefits and risks, it is essential to say that the implementation of this kind of approach would require “to rethink the way regulations are conceived in order to be able to achieve the standardization/automation needed for a machine executable regulation” (source: EBA Response 2020). The first step in this path would be to define precise rules and logics to use in writing regulation themselves, so that machines would receive a code ready to be automatically executed – in fact -, without any work on it. The same EBA writes that “a clear benefit is the elimination of the need by institutions to interpret the legislation if not sufficiently clear, or the possibility of misinterpreting it otherwise” (source: EBA Response 2020). In fact, in November 2017 already the FCA and the BoE identified significant benefits both for the regulators and the public, also introducing two possible approaches to the implementation of the MDMER:
As we have already highlighted, both these approaches would need to face many technical difficulties and risks. According to the EBA Response 2020, “a way to overcome this would be to already translate into machine-readable form the underlying legislation” (second approach), but also this operation would be critical and – maybe – utopian in the real world. Besides, it must be said that these technical issues need to be put together with some other legal and even philosophical aspects, which will be analyzed in our next blog post.
For the moment, at Aptus.AI we have developed an AI tool which has been specifically created to answer the RegTech market needs in terms of clarity, temporal efficiencies and costs reduction: learn more about Daitomic!