The insurance underwriting cycle is important from a financial stability perspective, as insurers are more prone to face underwriting losses during a soft market. The forecasting underwriting cycle could help insurers improve profit. But there are some pertinent questions that need to be addressed, including:

• are the underwriting cycles real and explainable?
• even if they exist, can they at all be objectified through a forecasting model?
• can an auto-regression time series model be appropriate to explain the documented cyclicality?
• are there any models, other than the auto regressive model, that capture the cycle more appropriately?
• which variables can be used to quantify this insurance phenomenon?

This interactive presentation will attempt to address these questions, discuss some documented techniques to forecast the underwriting cycle, and assess the impact of unusual (catastrophic) events.

All in all, the session will make a comparative assessment to see if it is worth trying to forecast the underwriting cycle. How can we navigate the challenges and successfully forecast the cycle? This will help insurers take proactive actions to deal with the underwriting cycle.

This will be a high-level session with a touch of technical application. It should be of interest to actuaries at all levels with an interest in general insurance market dynamics and a knack for statistical modelling.

Speaker: Dhrubo Banerjee, Swiss Re