The FSA's solvency requirements will ensure that insurance company managers are aware of the risks their businesses face Joe Monk explains how to identify and assess the risks.
All UK insurers, including Lloyd's managing agents, will by now have completed their individual capital assessment (ICA) models to ensure that they meet the new FSA requirements.
These models must include a detailed evaluation of the risks that the insurer is facing, both in terms of their assets and their liabilities.
This has been a new and at times stressful experience for most insurance companies, so what can be learned from it? After all, the exercise now has to be repeated annually.
For some insurance companies, this will be the first time that they have formally evaluated the relationship between the risks their business faces and the capital they hold.
In the past, some insurance companies have considered it appropriate to simply hold free capital as a multiple of the statutory minimum solvency requirement, regardless of whether the insurer is underwriting a stable direct property account, or a highly volatile US catastrophe excess of loss book of business.
The FSA now requires insurers to conduct a systematic review of all risks that could potentially threaten their solvency. This was outlined in the consultation papers CP190, which became policy statement PS04/16, and CP04/7 for Lloyd's.
It has deliberately avoided prescribing any one approach. Instead it has laid down minimum standards that it expects any capital assessment to reach, including that senior management are actively involved in the risk management of their companies.
Individual assumptions
One of the most straightforward approaches to creating a risk-based capital model is to create a model based on a series of individual assumptions, which can be varied to generate different outcomes. These models are often referred to as deterministic models.
While these models can provide useful results, a significant draw back of this type of approach is that these models will only be able to provide a single result for a given set of assumptions and will not be able to provide any indication of the probability of the resultant events actually occurring.
This can cause difficulty for those reviewing the model for reasonableness and determining what is a 1 in 200 year event, as required by the FSA. Often a wide range of scenarios must be separately generated to test the sensitivity of the model and to demonstrate a range of different outcomes.
Even after a number of inputs have been run through the model in an attempt to analyse the potential range of possible outcomes, deterministic models do not provide any indication to the insurer, or the regulator, of the likelihood of those events occurring.
A much more powerful approach is to construct a stochastic risk-based capital model that considers the entire range of possible outcomes at the same time.
These models include a sophisticated simulation process where each simulation selects a different set of assumptions based on pre-determined probability distributions.
These simulations can be performed tens of thousands of times resulting in a range of possible outcomes, together with the associated probability of each outcome actually occurring.
The real strength of this type of approach is the availability of the probabilities associated with the potential different outcomes.
Insurers that adopt a stochastic approach can take some considerable added comfort that they are able to more accurately quantify, and demonstrate to the regulator, the level of confidence that the capital they are holding will prove to be sufficient.
The assumptions are critical to the model. No matter how good the methodology applied, if the assumptions are wrong then the results will be wrong.
In short, if you put rubbish in, you get rubbish out.
It is also important when designing a risk based capital model that everyone who has an input into the model firmly believes that all of the model's underlying assumptions are appropriate. This applies as much to the underwriters and claims handlers as it does to the actuaries.
Sophisticated stochastic models are of little use if the insurer's management does not have an input or does not understand the model.
Whichever method of calculation is selected, the importance of senior management playing an active role in understanding the risk-based capital model and its underlying assumptions should not be understated.
This will ensure that the model can be used as a fully integrated business risk management model rather than purely a response to a regulatory requirement.
Types of risk faced by insurers
Insurance risk
For an insurance company it is no surprise that this is usually the most significant risk by far. This is split further into two categories, underwriting risk and reserving risk.
Underwriting risk
Insurers require capital to protect against the risk that premiums and other investment income may not be enough to cover claims expenses and other liabilities that emerge from new business written.
Reserving risk
As well as holding capital in respect of ongoing underwriting business, insurers will also require capital in respect of its reserves for outstanding liabilities for prior underwriting years. Owing to the uncertainty surrounding the cost at which outstanding claims will ultimately settle, capital is required to the extent that reserves may prove inadequate.
Using statistical modelling techniques that look at the variability in claims development patterns, it is possible to simulate the potential range of future claim settlement costs and, hence, to quantify any potential deficits in reserves.
Market risk
An insurer may suffer an adverse financial effect as a result of market forces outside the immediate control of the insurer. For an insurer this is most likely to be a fall in market premium rates and a slackening of terms and conditions. This would have the effect of having greater exposure to risk than was originally anticipated in the business plan.
Liquidity risk
An insurer may suffer financial losses as a result of income or capital growth from its asset portfolio being lower than expected. By looking at an insurer's investment portfolio and likely future cashflows, it is possible to simulate the range of possible future investment returns and the extent of any potential asset losses.
Insurers face the risk on their assets that they may not be able to realise full asset values as liabilities fall due, for example on a lump sum award by a court.
Credit risk
Most insurers hold significant assets in the form of reinsurance. However, depending on the financial security underlying the reinsurance programme, there may be significant risk that some recoveries may not be received if reinsurers become insolvent in the future.
Operational risk
An insurer will face various other internal risks in its operation such as fraud, failure of management control systems and failure of data systems.
A "scenario-based approach" can be adopted where the potential frequency and severity of a range of specific operational risks are simulated.
It is also important to identify any past operational risk in data that may have been already implicitly modelled within the insurance risk.
Group risk
This covers risks for an insurer that may come from other companies within the group. For example, the implications for an insurer if its parent company had financial difficulties.
