Risk analytics

The Risk Analytics Unit is responsible for providing industry-leading actuarial, quantitative and catastrophe modelling services to the StarStone Group. Our work supports the Group’s ability to make superior decisions in the presence of uncertainty.

Our key functional areas of involvement are as follows:

Technical pricing and risk selection
This function undertakes transactional-based activity and aligns itself with the underwriting organisational structure, by line of business. Actuarial and catastrophe modelling elements are combined to model risk and derive technical prices as a foundation for superior risk selection.

Capital and risk management
The Group faces risks in many forms and guises as it conducts its business, from underwriting exposures, counterparty credit risk, operational issues and cash flow management, to fluctuations in the market value of the Group’s investments. The capital and risk management function examines the full spectrum of risks faced by the Group, with a focus on identification and quantification of these risks. The function is an integral part of the wider Enterprise Risk Management framework of the firm, providing information and reporting to help the company meet its financial and risk management objectives.

Loss Reserving and Valuation
This function’s primary goal is to value, monitor and report on the Group’s operating liabilities. Loss reserves are quantified for statutory, GAAP and fair value reporting.  Claims emergence over time is monitored against expectation, with feedback loops in place to continually improve the modelling of loss development patterns.

Research and Development
The Group’s R&D activities span the full gamut of actuarial and catastrophe modelling applications in a bid to develop and refine existing modelling capabilities. Work is typically project-based and ranges from the scoping, design and development of new models, the research and application of new methodologies and theories to quantify risk, and the parameterisation of existing models using data sets.