Putting the tech into RegTech (part 1): text and semantic analytics

Updated: Mar 2, 2019


#regtech #compliance #regulatoryreporting #textanalytics #semanticweb #semanticontology


Introduction


Consider this -a series of new regulations are published and as is usually the case governments, central banks and regulatory bodies mandate adherence to them. Compliance teams in most supervised institutions are already weighed down with the burden of existing regulatory compliance. They must now go through the typical process of responding which will initially entail something similar to the following steps:

  1. understand the regulation; and then

  2. outline a plan to become compliant.

In order to understand the regulation an institute will likely draw up an interpretation internally (via in-house legal and compliance) and sometimes solicit the assistance of an external consultancy. This takes time and of course, comes at a significant cost especially when we consider the volume of regulatory change that currently takes place.


In order to define a plan one must understand, at a high level to begin with, the extent of change required within the organisation. What aspects of organisation, process and technology need to be changed and how quickly can that be done?


This is where text analytics and semantic matching can be of help.


What is it?


First some basic definitions:


Text analytics is the process by which text (typically unstructured and written in natural language) is structured in a way that enables further analysis such as classification, tagging or annotation.


Semantic analytics combines text analytics with technologies traditionally used to power web search. It effectively takes input text and interprets it by matching and sorting the text against a set of terms (usually described as an ontology).


When used in the above scenario, the combined application of these technologies is to give institutions the ability to match parts of regulatory text to a whole series of data categories that will enable implementation planning (step 2 above) to become a much easier, data-driven exercise. For example, the text can be matched to:


  • Metrics - e.g. machine executable calculations that pull in data from existing sources and generate results (e.g. Core Tier 1, Liquidity Coverage Ratio, Value at Risk)

  • Evidence sources - e.g. stores of evidence which can be pointed to when demonstrating compliance against a given part of regulatory text

  • Processes - e.g. by mapping to a given set of processes, the organisation immediately obtains a view of what part of its activities and services to target when addressing a regulatory need


Of course, there is a lot more that comes after these initial stages but for the purpose of brevity, we will focus on just this for now (more to come in future blogs).


Why is it important?


All of this becomes important during times where low volume, low interest rates and low growth markets have negative correlation with operational costs and in particular, the high cost of compliance:


Governance Risk and Compliance (GRC) spend accounts for 15-20% of “run the bank cost”, and 40% of “change the bank costs”

(Groenfeldt, 2018)


For the traditional, universal banks with a multitude of globally dispersed applications (both off the shelf and bespoke) the challenge of responding to regulation is particularly acute as they simply have too much internal complexity to deal with. This gives the rising "neo-banks" a competitive advantage as they have built systems "clean" and therefore find it much easier to get the information they need to demonstrate compliance.


What next?


The regulatory industry is creating a world where this will become increasingly relevant. Currently, a number of regulatory driven initiatives are under way which seek to evaluate technologies that will help them to digitise regulation and more easily obtain compliance data from the institutes that they are monitoring. The Financial Conduct Authority (FCA) in the UK appear to be heading the charge in this arena with their Regtech Sandbox. The quicker banks implement tools that incorporate the above technologies, the quicker they will be able to move to a new world where this type of requirement is the norm and be ready to meet the demands of their regulators.


Regulated institutions must act now to get ahead. To address this need and challenge of the rising cost of compliance, a low cost investment in the right technology will not only provide readiness but also an economy of scale which will enable them to address the increased volumes of regulation coming their way and quickly identify which parts of their enterprise need to be changed.


How can we help?


We are actively developing the above technologies within our "MonPlat" offering. Contact us if you'd like to know more.

0 views
LINKS
CONTACT
SOCIAL

info@complianceasaservice.com

Manufactory House, Bell Lane, Hertford, Hertfordshire, England, SG14 1BP

  • Facebook - White Circle
  • Instagram - White Circle
  • White LinkedIn Icon
  • White Twitter Icon

2018 Compliance as a Service Ltd.