Why Inconsistent Views of Risk Cost You

If models underpinning these four systems are not aligned at a structural level (consistent parameterisation of past experience and future assumptions), you end up with four different views of the same risk – with no clear way to reconcile them.

Further, if these views drift apart, the business is more likely to respond poorly to changing market conditions. For instance, divergence can manifest in opportunity cost (incorrect pricing / over capitalisation) or capital adequacy problems (inadequate reserves and/or risk capital).

A single, coherent statistical framework can remove that disconnect. The same underlying model should drive reserving, pricing, reinsurance and capital management, with differences arising only in how the outputs are used.

Why Alignment is Crucial for Long-Tail Liabilities

In long tail business, risks emerge over time – and can have far reaching effects. If four teams read these risks differently, the organisation ends up reacting too late – or in the wrong direction.

Single view of data

The same underlying process underpins all four areas. Using a common model and forecast scenario avoids running multiple, conflicting versions of reality.

Speed of response

Changes in calendar trends – for instance social inflation – emerge first in reserving. If this information does not flow into pricing or reinsurance, the business continues writing mispriced risk or holding inappropriate levels of capital.

Consistency

Using consistent assumptions to estimate the full loss distributions for both reserving and pricing mitigates model specification risk and streamlines reinsurance and capital management applications.

Risks of Independent Models

• Strategy disconnect: If reserves are strengthening while the underwriting team reduces prices – due to different model assumptions – this is likely a strategic oversight.
• Lagged Capital Management: Independent capital models often lags emerging trends found in claims data resulting in capital inefficiencies.
• Inconsistent Views of Risk: If different parts of the business are using different assumptions, this ensures inefficiencies within the system. At least one of those views is wrong.

Bridging the gap

The solution is not to “build four better models”, it is one coherent framework.

Trends and volatility estimates, along with any volatility correlations between portfolios, should be estimated once for the past, and the same future assumptions applied going forward to project emerging risks irrespective of application.

Granularity may differ between Reserving, Pricing, Reinsurance, and/or Capital Management, however the view of the underlying process should not.

Our flagship software solution, ICRFS™, includes the Probabilistic Trend Family modelling framework and the Multiple Probabilistic Trend Family modelling framework. The latter provides a company-wide framework to identify one single composite model driven by the data

For more information, see: One Single Composite Model