The HKMA will explore the potential conversion of regulations into machine-executable codes and the introduction of a system that automatically creates regulatory reports.
The HKMA (Hong Kong Monetary Authority) has issued its seventh and final issue of the Regtech Watch series, outlining a three-year roadmap to integrate suptech into its supervisory processes to enhance their effectiveness and forward-looking capability.
The focus of the initial stages of implementing the roadmap is placed on conducting pilots and PoCs to test the feasibility of suptech solutions that will enable the development of a centralised platform that supervisors can use to view and access information about banks in a single location.
The HKMA is also looking to build a knowledge management system for storing structured supervisory information (e.g. banking returns) and unstructured information (e.g. board minutes, internal audit reports).
A PoC will be conducted on the use of RPA (robotic process automation) to automate repetitive tasks and streamline existing workflows. “It is envisioned that RPA can help perform routine supervisory work by mimicking how users interact with applications such that the entire end-to-end processes can be performed by software robots with little human intervention,” the HKMA says.
A separate PoC will be conducted on the use of an audio transcription tool to help the HKMA consolidate information onto the centralised platform and knowledge management system. The speech-to-text tool will also be used to help analyse transcriptions and identify specific keywords, tone and sentiment that will enable the HKMA to identify potential misconduct.
The HKMA will also be looking to adopt advanced analytics techniques, focusing on enhancing its ability to detect early risk signals, which in turn will improve its forward-looking capabilities. This will include the use of automated intelligence gathering and machine learning techniques to help the HKMA capture and process structured and unstructured information from proprietary sources such as banking statistics and publicly available sources such as social media posts.
“This will enable the HKMA to develop a holistic and comprehensive understanding of the institutions that it supervises and the risks the institutions face,” the HKMA says, adding that the new capabilitieswill enable more timely supervisory actions.
Two PoCs will be conducted to facilitate the gathering of emerging risk signals.
- Network analysis will help to improve transparency on corporate shareholding structures and banks’ exposures to their customers, as part of efforts to improve early detection of risks within the entire credit network.
- Sentiment analysis tools that can categorise and assign scores to large quantities of bank-related news, allowing the HKMA to stay more alert to negative news and gauge banks’ exposures to conduct risks from the financial products they offer.
The HKMA is also considering the use of other advanced analytics techniques to improve the efficiency and effectiveness of banking supervisory processes. An example cited was the potential use of text tokenisation and summarisation, which splits text into smaller units to quickly summarise and obtain insights from sets of documents.
The HKMA will also investigate methods to make policy assimilation and dissemination more accessible through the use of electronic formats and an open API framework. This will help to improve banks’ ability to manage and process supervisory policy communications and maintain compliance with evolving regulatory requirements, it says.
Part of this work will centre around the potential conversion of regulations into machine-executable codes and the introduction of a system for the automated creation of regulatory reports.
The full newsletter is available here.
The HKMA’s Regtech Watch series will be succeeded by a Regtech Adoption Practice Guide series. The inaugural issue provides guidance on cloud-based regtech solutions, covered here.