The Bank of England’s (BoE) fintech accelerator has kept good speed and completed the third round of various proofs of concept (PoCs).

As reported last year, BoE launched its own accelerator to harness innovation for central banking.

The latest PoCs covered four areas of the bank’s work: analysing large-scale supervisory data sets; executing high-value payments across currencies and borders; identifying and applying cross-cutting legal themes from regulatory enforcement actions; and measuring performance on the bank’s internal projects portfolio.

Andrew Hauser, BoE’s executive director for banking, payments and financial resilience, says it has learnt a “great deal through these latest PoCs, both in terms of what fintech can do, but in also in terms of how it can help us work, think and communicate differently”.

It worked with Mindbridge AI, a machine learning and artificial intelligence (AI) firm, to explore the analytical value of using AI tools to detect anomalies in supervisory data sets. Using a sample set of anonymised reporting data, BoE says it found Mindbridge’s user interface to be intuitive, allowing the user to explore a time series of each variable, whilst comparing the results to industry averages. This PoC enabled the bank’s internal team of data scientists to compare and contrast their own findings and the underlying algorithms being used, providing a “complementary layer” to the bank’s work.

In the bank’s PoC with Ripple, it looked into how distributed ledger technology (DLT) could be used to model the synchronised movement of two different currencies across two different ledgers, as part of the bank’s wider research into the future of high-value payments. Although the bank has “already concluded that DLT is not sufficiently mature to support the core RTGS system”, the learnings from this exercise with Ripple have “reinforced the bank’s intention to ensure its new RTGS system is compatible with DLT usage in the private sector”, and has highlighted areas where it would like to conduct more exploratory work.

With Enforcd, BoE gave a group of staff from its regulatory action division access to a cloud-based database of regulatory enforcement actions with supporting commentary and trend analysis. According to BoE, this PoC demonstrated how technology could potentially facilitate compliance and the development of best practice in some key areas of regulation.

Lastly, the bank ran a PoC with Experimentus, using its ORB tool, to analyse historic BoE projects and visualise how they had performed against a range of standard key performance indicators (KPIs). This PoC allowed the bank to explore whether its existing test data were sufficient to carry out effective KPI reporting, and where further data collection might be needed.

The accelerator invited applications for its fourth round of PoCs in April and expects to announce the successful firms shortly.