InsurSight

Automated reserving

InsurSight automatically sets first-cut reserves for every reserving class and cohort in minutes.​

InsurSight automatically sets the assumptions used in traditional reserving approaches. For example, pattern averaging periods, method selection etc.​

InsurSight also shows you where to prioritise your time, using the ‘universe of reserves^TM’ to show where there is the most uncertainty in the automatically selected reserves.

Key assumptions prioritised

InsurSight shows you which reserving assumptions are the most important for each reserving class.

This gives you clarity on which assumptions to focus on right at the start of the reserving process, giving more time for wider business input on key judgements.

Easy-to-review diagnostics

Easy-to-review data visualisations of key assumption selections and diagnostics are set up automatically right at the start of the process.

• For each key assumption, the selection made is clearly shown alongside alternatives that could have been used.
• Reserving diagnostics used by regulators, such as weakening of case estimates, are embedded into the reserving process.
• Dashboards are fully customisable, set up in seconds, and highlight the key features of each reserving class.

Assumption fine-tuning

You can easily adjust any assumption set automatically by InsurSight, and fine-tune specific assumptions where necessary. For example:

• For development patterns you can: make exclusions, set averaging periods, set tail assumptions, and set patterns by cohort.
• For initial expecteds you can: on-level past ULRs for rate change and inflation, use various trending approaches, and adjust weights used in averages for specific cohorts (eg where there are unusually large or small losses).
• For method selection you can: apply custom weights by cohort to a range of methods across paid, incurred, frequency-severity, DFM, BF and Initial Expected methods.

Machine learning enhanced selections

InsurSight's machine learning powered assumption setting separates the signal from the noise in your data, helping you set assumptions with more confidence.

Dashboards are fully customisable, set up in seconds, and highlight the key features of each reserving class.