Discover: AI powered covenant

Complexity and cost often put traditional covenant assessments out of reach for smaller schemes or scheme sections. This can be a problem given the importance of covenant to
member outcomes, increased focus from The Pensions Regulator (TPR) on covenant, and
draft DWP legislation requiring trustees to assess covenant strength.

Data and modern artificial intelligence techniques can be a solution for this problem, providing the type of high-level covenant input and analytics more appropriate for these smaller schemes or scheme sections. Such an AI covenant tool should:

1. Start with a comprehensive database of covenant assessments and all the variables underlying each covenant rating, and use machine learning to analyse which factors are typically the key drivers and need to be considered for a streamlined approach;
2. Use a model to replicate rating classifications not only based on quantitative metrics, but also the forward-looking qualitative metrics that are crucial to understanding covenant strength; and
3. Illuminate the ‘black box’ AI model, to provide not only a covenant rating, but also identify the underlying drivers – these are often key to informing next steps, negotiations, or monitoring.

Developing Discover

At Cardano, we have followed such an approach to develop Discover – our AI powered covenant analytics solution. Discover provides an employer covenant rating, examines the
main drivers underlying the rating, benchmarks rating metrics, and provides corresponding recommendations.

Discover has been designed for schemes with assets less than £50m and simple covenant structures. Or for sectionalised or non-associated multi-employer schemes, including local government pension schemes (LGPS).

Discover provides covenant ratings and rating drivers

Launching Discover is a significant and exciting step towards modernisation within the pensions industry. Harnessing our wealth of data (capturing 300+ years of combined covenant
experience) with new technology to build cost-effective tools and solutions for schemes. Discover gives comfort that employer covenant views have been externally validated,
supports triennial valuations, scheme return submissions, ongoing monitoring, and can also be utilised to explore ‘what-if’ scenarios, for example, an improvement in a sponsor’s trading or an increase in its debt.

Read the full Covenant Quarterly Q4 2022 which also includes the following articles:

• Datawatch: UK GDP trends
• Ask the Analyst: How should I be thinking about covenant in the context of recent gilt yield volatility?
• Regulatory developments: Covenant front and centre within draft Occupational Pension Schemes Regulations 2023

For more information contact Felix Mantz, Director