PRA Supervisory Statement SS5/14
On 25 April, the PRA published a supervisory statement discussing their expectations regarding technical provisions and internal models. The PRA consulted the content of this statement in CP7/14. The aim is to enable firms to consider the PRA’s expectations as part of their planning and preparation for the Solvency II regime.
The statement covers some of the key areas to consider when calculating technical provisions and using internal models.
Technical Provisions
The calculation of technical provisions will significantly change under Solvency II. Given that technical provisions are usually the largest item on an insurer’s balance sheet; this is likely to have a substantial impact on an insurer’s financial strength.
Here we have summarised the supervisory statement with a set of 'do's' and 'don’ts'.
| Technical Provisions | ||
| Do's | Don'ts | |
| Risk margin | Realistic assumptions and adequate methods. |
Approximate the future SCR used to calculate the risk margin as proportional to the projected best estimates (unless this has been shown to lead to a material misstatement of technical provisions). A hierarchy of five levels of simplification has been suggested (in EIOPA guidelines): make a full calculation of all future SCRs without simplification. Approximate some of the individual (sub) risks |
| ENID (events not in data) |
ENID must be taken into account when calculating technical provisions. An allowance should be made for extreme values that cannot be projected using historic data. Consider outliers that have been removed as part of the reserving process. |
Use the terms ‘binary events’ or ‘extreme events’ – ENID are not necessarily extreme or rate. Apply simple percentage uplifts without justification. |
| Premium provisions | Produce a best estimate as required by Article 77 of the Directive. | Use optimistic business plan loss ratios – this will not produce a best estimate as required by the Directive. |
| Approximations (if used) |
Consider the cumulative materiality of approximations within technical provisions, as opposed to individual aspects. Quantify the materiality. |
Ignore the impact of the approximations with other simplifying assumptions. |
Internal Models
It is important that an internal model covers all material risks to which a firm is exposed. Documentation and validation will be a crucial tool, allowing firms to discuss how they have treated the key risks and to justify assumptions and methods selected to the PRA.
| Internal models | ||
| Do's | Don'ts | |
| ENID (events not in data) | Concept of ENID must be applied to the data used to set the parameters for the internal model. | Assume that historic data alone will take into account all quantifiable risks (unless stress and scenario testing shows that an adjusted distribution captures the full range of possible future events). |
| Risks covered by third party models | Demonstrate third party models cover all material risks in their own risk profile.
Understand and monitor the limitations arising from the use of external models. |
Discard the limitations arising from the use of external models. |
| Technical provisions in the internal model, e.g. to calculate the movement in basic own funds over one year |
Use a method that produces similar results to a full technical provisions calculation. The method adopted to calculate the technical provisions must produce a probability weighted average of future cashflows. |
Use a method that only produces results in benign circumstances, i.e. don’t ignore the rest of the probability distribution forecast. |
| Uncertainty around parameters | Allow for parameter estimation error where this is material and it is practical to do so. | |
| Calendar year effects | Consider and allow for calendar year effects. | Utilise any models or methods that do not account for volatility introduced by calendar year effects. |
| Improvements in performance | Assume future improvements in performance without evidence to back this up. | |
| One-year emergence of risk | Assume that insurance risk emerges according to historical development patterns. | |
| Industry standards | Justify assumptions on the basis of your own specific risk profile. | Set assumptions solely based on the fact that they are generally accepted market practise. |
| Default options | Justify assumptions on the basis of your own specific risk profile. | Set assumptions on the grounds that they are selected by default |
| Own data | Data used for the internal model must be accurate, complete and appropriate. | Ignore any data that can have an impact on the outputs of the internal model. |
| Risk mitigation | Consider the possibility of reinsurance exhaustion. | Fail to ensure that the risks arising from the risk mitigation techniques are not properly reflected. |
| Management actions |
Take into account management actions. Treat the renewal of reinsurance as a future management action unless it has been shown that the renewal will not rely on a decision made by the firm. |
|
| Validation standards | Have a regular cycle of validation to review the ongoing appropriateness of the internal model. | Perform validation that does not relate to your own specific risk profile. |
| Data from third party models | Demonstrate that the third party model satisfies Articles 120-125, and not the data alone. | |
The PRA wants to ensure that general insurers set an adequate level of technical provisions and hold sufficient capital.
As part of their preparation for Solvency II, firms should be considering the key issues and establishing processes to calculate their technical provisions. In addition to these issues, firms should also consider the required documentation and validation.
The statement sets out the PRA’s expectations in relation to using internal models. IMAP firms should be in regular contact with their review team and should be discussing these key issues.
The full statement can be found here.