SupTech, short for supervisory technology, is the use of innovative technology by financial supervisors to support their work. Central banks, banking and securities regulators, conduct authorities and financial intelligence units use it to collect data from the firms they oversee and to analyse that data for signs of risk, misconduct or financial crime. The term was set out by the Bank for International Settlements (BIS) Financial Stability Institute in a July 2018 paper on the experience of early users.
SupTech is the mirror image of RegTech: RegTech helps regulated firms comply, while SupTech helps authorities check. The two meet at the point where data passes from firm to supervisor, most visibly in regulatory reporting. A later BIS paper in 2019 refined the idea to focus on big data and artificial intelligence, and widened it to include bodies such as financial intelligence units.
SupTech in plain terms
Imagine a building inspector who, instead of visiting each site once a year with a tape measure, receives a live feed from sensors in every building and has software that highlights the ones showing cracks. The inspector still decides what to do, but spends time on the buildings that need it rather than on paperwork.
That is the idea behind SupTech. Supervisors used to receive spreadsheets and PDF returns weeks after the period ended, then read them by hand. SupTech replaces that with automated data feeds and analytical tools that point staff towards the banks, payments or traders that look unusual.
Data collection
The BIS groups SupTech uses into two areas, data collection and data analytics. On the collection side, authorities are moving along a path:
- Automated reporting. Firms submit returns in machine-readable formats such as XBRL instead of spreadsheets or paper, with automatic validation checks on arrival.
- Granular data. Rather than summary tables, supervisors collect loan-by-loan or transaction-level data and build the aggregates themselves, so they can ask new questions without new returns.
- Pull models. Supervisors fetch data directly from firms' systems through an application programming interface (API). The Bangko Sentral ng Pilipinas, working with the RegTech for Regulators Accelerator, built a prototype API for prudential reports, described in a 2018 case study, and has since moved toward submitting prudential reports through live APIs.
- Data management. Data lakes, cloud storage and common data dictionaries let one authority combine returns that used to sit in separate silos.
Data analytics
Once the data is in, authorities apply analytical tools to it. Common applications include:
- Prudential supervision. Dashboards and early-warning models track capital, liquidity risk, credit risk and concentration risk across banks, with alerts when indicators move.
- Financial crime. Financial intelligence units use network analysis and machine learning to sort large volumes of suspicious activity reports (SARs), linking accounts and people across filings.
- Market conduct. Securities regulators scan trading data for signs of insider dealing and market manipulation.
- Consumer protection. Text analytics read complaints, websites and social media to spot mis-selling or scams.
- Document review. Natural language processing summarises board papers, policies and inspection files so supervisors can find relevant passages quickly.
Many of these techniques, such as anomaly detection and behavioral analytics, are the same ones firms use for transaction monitoring; the difference is that the supervisor sees across the whole market rather than one firm.
Who is building it
SupTech work is spread across national authorities and international bodies. Central banks and regulators in advanced economies and emerging markets alike run SupTech teams, often inside innovation units. The BIS Innovation Hub runs shared experiments for central banks, and the Financial Stability Institute publishes surveys and case studies. In emerging markets, SupTech is often tied to financial inclusion: supervisors of mobile money and digital credit need automated tools to watch large numbers of small transactions and new types of provider.
For regulated firms, SupTech changes the relationship with the supervisor. Data that is submitted in granular, machine-readable form can be checked against other returns instantly, so inconsistencies surface quickly. Good recordkeeping and clean data become part of financial compliance in their own right.
Challenges and limits
SupTech projects face recurring obstacles:
- Data quality. Analytics fail if firms report inconsistent or incomplete data; standard definitions take years to agree.
- Skills and budgets. Authorities compete with the private sector for data scientists and must justify technology spending to governments.
- Legacy systems. Old reporting platforms and legal mandates tied to specific forms slow change.
- Explainability. A supervisor that acts on a model's output must be able to explain the decision, which limits the use of opaque models.
- Security. Holding large volumes of sensitive data makes supervisors an attractive target, so cybersecurity risk and data security are central concerns.
Technology supports judgement but does not replace it: the decision to intervene in a bank or bring an enforcement case still rests with supervisors and the law that empowers them.