A previous post focused on IFRS reserves. This post is focusing on Solvency II reserves. Some of the Solvency II requirements for technical provisions are challenging compared to the IFRS requirements. The regulation is spread over the Directive – Level 1 and Level 2 and 3.
The aim of this text is to provide tidy and useful insights on the subject.
Technical provisions are usually the largest item on a general insurer’s balance sheet. This fact remains under Solvency II implying that this will remain an essential component.
Then from the balance sheet stems the import number called the excess of assets over liabilities. Technical provisions will be an important part of this number. Since this number is also affecting the SCR ratio through Own Funds the technical provision will also, obviously, affect the SCR ratio. And more implicitly also the denominator in the SCR ratio, the SCR itself. And it affects the risk margin too. And the MCR. To conclude it’s an important concept.
The below picture illustrates how technical provisions (TP) fit within the Solvency II balance sheet.
From a calculation point there is generally two approaches to use. Either a deterministic approach or a stochastic approach. The deterministic one is still the more common. This is described further in the IFRS Technical Provisions. With the increase amount of data, more AI and faster computers we might see a shift towards stochastic approaches in the future.
In addition most general insurers chose to build on the IFRS reserves rather than creating the Solvency II technical provisions from scratch.
There will always be an element of judgement involved. For instance on what model or which assumption to use. This also links to what is proportional, technical provisions calculation will always be subject to time and cost restrictions and the method chosen has to take this into account to.
Under Solvency II (and for general insurers) the technical provisions (TP) are made up of
Claims Provision + Premium Provision + Risk Margin
The claims provision is the discounted best estimate of all future cash flows (claims payments, expenses and future premiums) relating to claim events prior to the valuation date (e.g. Year End).
The premium provision is the discounted best estimate of all future cash flows (claims payments, expenses and future premiums) relating to future exposure arising from policies that the (re)insurer is obligated to at the valuation date.
The risk margin is intended to reflect the additional amount another insurer would require to take over and meet the technical provisions. It is calculated using a cost-of-capital approach.
The discount rates to use in the calculation will be provided (monthly) by EIOPA for all major currencies.
Solvency II requires that data used in general, and in particular, when calculating the technical provisions is appropriate, complete and accurate.
- Data should be suitable for the purpose
- No estimation error arising from the nature of data
- Consistent with the methods and assumptions used
- Reflects the underlying risks
- It should have sufficient granularity and information to identify trends and characteristics
- Available data for each homogeneous risk group
- No relevant data are excluded without justification
- Free from material errors
- Data from different time periods are consistent
- Recording of the data is consistent
Current liabilities for insurance contracts plus a risk margin. This needs to be calculated separately gross and for reinsurance. The premium provision may be negative if the expected cash inflows is greater than the cash outflows. The premium provision therefore takes into account expected future profit.
Future policy behaviour such as lapse should be allowed for. If claims give rise to annuities these should be calculated using appropriate life techniques.
Both claims and premium provisions needs to consider low probability high severity (and other) events usually referred to as ENIDs.
The best estimate should allow for the time value of money, i.e. discounting.
Existing contracts to use in the premium provisions is defined on a legal basis. Meaning that as soon as the insurer becomes part of the contract it should be included.
Claims in Claim Provision
This is similar to the IFRS methods used historically. The key changes are removal of any margins and consideration of ENIDs. Methods commonly used are chain-ladder or Bornhuetter-Ferguson or Average Cost per Claim).
It is good practise to clearly justify the quantification of the margin removed from an IFRS based estimate.
The present values can be calculated by applying a payment pattern to the undiscounted reserves to simulate a cash flow which is then discounted and summed. The payment pattern is likely to be based on historical payments.
Claims in Premium Provisions
For the claims part of the premium provision a common approach is to apply a loss-ratio to the unearned premium to come up with the undiscounted reserves, then applying a payment pattern and discount to reach the present value.
How to choose the loss-ratio is another interesting subject, these could be based on pricing, plan or URR calculations for instance.
The pricing loss ratio is usually available but it may cover a different period of exposure than the one covered by the premium provision. For annual contracts written uniformly over the year, for instance, the premium provision will be weighted towards the first few months.
Planned loss-ratio could be used if they are on a best estimate basis, not out-of-date and covers the relevant period.
URR loss ratios may be used if these calculations are available. Adjustments are probably required though, these would possibly include:
- Rating environment
- Legal environment
- Policyholder behaviour
- Business mix
- Inflation and other trends
- Reporting and settlement days
- Lapse rates
- Reinsurance coverage
- Investment return
- Exceptional events
- Terms or policy wording
- Socio-economic trends
Attritional and large losses should be calculated separately if this is warranted because of the nature of the risks. Some risks have a seasonal effect, for instance hurricanes or floods, this may also have to be considered. URR loss-ratios are usually updated less frequently than the technical provisions so caution may be needed.
The materiality of this will depend on two things mainly. The type of business, i.e. long tail or short tail, and the current yields. Lately the yields have been very low, even close to zero, which means that the effect of discounting is less than otherwise.
Allocated and unallocated claim management and other expenses are part of technical provisions.
Future premiums – claims and premium provision
This part should be known from the “premiums receivable” on the balance sheet. Technical provisions should include premiums that are not passed due. Premiums already due are part of the “insurance & intermediaries receivables” on the Solvency II balance sheet.
The potential confusion between finance and actuarial departments are considerable here. It might not be clear what relates to overdue premiums and what doesn’t. There is obviously a risk of double counting or omission here.
Further issues may arise because of the need to split future premiums in the expired and unexpired buckets. Usually the amount of future premiums relating to past exposure is low, possibly immaterial. The future premium receivable should be net of any expected premium defaults.
Existing contracts to be included in the technical provisions relates to the concept of contract boundaries.
Contracts should be recognised as existing when the (re)insurer becomes a party of the contract. And derecognised accordingly.
The boundaries should be defined as per below:
- When then insurer has unilateral right to cancel the contract, reject premiums or amend premiums or benefits to reflect the underlying risk at some future point in time, premiums after that point should not be included.
- Any future premiums relating to options and guarantees belong to the existing contract.
The assessment should in principle be made on a per contract basis. If this is not practical some grouping may be applied.
One of the impacts of the above is that the premium provision may need to include cash flows from policies with a starting date after the valuation date.
If there is a contract with an intermediary thing might be more complicated. In general “a look through” approach should be used to identify existing contracts. The views on this subject might be different depending if the business is profitable or not. As with other areas that involves judgement the insurer should use a reasonable approach and justify it accordingly.
Another area of judgement is where the insurer aren’t necessarily obliged to re-quote the contract but will in almost every situation still do so because of market practise or from a moral obligation.
It is useful to compare the premium provision under Solvency II with the unearned premium reserve under IFRS. It will be a good sense check and an opportunity for reconciliation.
Under IFRS the UPR is reduced by the Deferred Acquisition Cost (DAC) on the asset side of the balance sheet. The DAC refers to upfront costs paid at inception of the contract but should be spread over the lifetime of the exposure period. In other words, the unexpired part of the acquisition costs. In most cases UPR is shown gross of DAC, and DAC as a separate item.
In cases where the UPR is assumed to be inadequate to cover future losses (and expenses) an Unexpired Risk Reserve (URR) should be held. This is therefore a prospective calculation. A key difference between Solvency II and IFRS in this area is that the need of an URR for one group of non-profitable business cannot be offset by another group of profitable business.
Consider a potential claim that will be either 0 or 100 with a probability of 50% for each outcome. Under Solvency II the technical provisions would be calculated as 50. Under IFRS insurers will probably come up with different numbers. Some will reserve 50. Others will reserve 100 because they would prefer to release a reserve of 100 than be in a position with a reserve deficit of 50.
Summary of key differences:
|IFRS||Solvency II PP|
|URR||Includes in-force contracts||Also includes Bound but not incepted.|
|ENIDs||No explicit allowance||Yes|
|DAC||Allowed for||Not applicable since part of historical cashflow.|
|Future profits||Not allowed||Included|
Removal of margins
In theory removal of margins is easy. In some cases the margin will be explicit in terms of a number or a specified percentage and these would then just be excluded. In practise there could also exist implicit margins which may be more difficult to quantify. In other cases the explicit margin includes management’s new view on the claims environment and therefore part of best estimate.
Often the management are for more happy with reserves that are sufficient than at a best estimate and generate a positive run-off than not.
Other areas for consideration include
- Tax implications
- Impact on the audit opinion
- If margins are consistent with actuarial opinion
Insurance & reinsurance receivables
Under IFRS these are assets held on the balance sheet in terms of future premiums from brokers, policyholders and reinsurers. Under Solvency II these should be part of the technical provisions but identifying which part should be moved and what to leave is not always straightforward.
Further complications may arise if large amounts of recently received premiums are sitting temporarily in clearing accounts waiting to be allocated.
The technical provisions discussed above would generally need to be calculated at a minimum per Solvency II line of business and per currency. Some considerations will be needed here as well, for instance,
- If there is uncertainty in which currency a claim would be paid.
- The premium might not be in the same currency as the claim.
- Different currencies between claims and expenses.
“IFRS reserves shouldn’t be compared to Solvency II technical provisions alone but rather one should compare the two balance sheets.”
In general, under Solvency II, reinsurance recoverables should be treated by the same principles as gross liabilities.
The expected recoveries should be adjusted to take account of expected losses due to default of the counterparty. This should be based on the probability of default and the average loss resulting from this (loss-given-default).
For many companies the approach is similar to what they have done in the past.
As with many areas under Solvency II, there are a range of possible simplifications that may be used for reinsurance.
What to include
The concept of existing contracts described above also applies to reinsurance contracts. As soon as there is an obligation it should be included in technical provisions.
There is also a principle of correspondence often applied to reinsurance contracts in a way the reinsurance recoveries (and reinsurance costs) should reflect the gross existing contracts.
What should definitely not be included are future recoveries from reinsurance contracts yet to be formalised relating to claims from gross liabilities relating to contracts that are not existing either.
The best estimate recoverable can either be calculated directly as the probability-weighted average of future recoverable cash flows or indirectly as the difference between gross and net best estimates. The indirect method should only be used though if it is expected to produce corresponding results to the direct method.
Most commonly, the reinsurance pattern is assumed to be the same as the gross claims payment pattern. But this may be a considerable simplification of cash flow projections. Justification is probably needed and as usual the nature, size and complexity of the business should be taken into account as required.
An example of an analysis on the gross vs. net patterns may result in a graph similar to the below, in this case assuming the reinsurance have the same pattern as gross and net is probably not justified.
On the other hand with increased share of reinsurance and longer delays the difference in patterns may be significant. Image below.
Under Solvency II recoverable from reinsurance most be adjusted for expected losses due to counterparty default. This is no different from past reporting requirements.
The assessment however should now be made on a cash flow basis, including therefore the timing of any defaults. This should include the whole run-off period and not just the coming 12 months.
The counterparty default adjustment should be calculated separately at least for each line of business, counterparty and where applicable separately for premium provision and claims provision.
Simplifications may be used but should be justified, some examples are:
- Asses the counterparty risk as a simple deterministic value based on mean term of cash flow
- Group counterparties by credit ratings
- Use a static estimated recovery rate in the case of default
When calculating technical provisions all future expenses that will be incurred in servicing existing insurance and reinsurance contracts should be taken into account. This includes administrative, investment expenses, claims management, claims handling and acquisition expenses including commissions. Administration expenses include salaries, property costs, IT.
This is a difference from IFRS since typically only ULAE are explicitly considered and ALAE is included as part of the projected ultimate claim amount.
The general principle here is that the expense should reflect the costs another insurer would incur if they were to take over the management of the business, i.e. the existing contracts, but on a going concern basis, unless a run-off assumption is justified.
As the other items expenses should be calculated at least on a Solvency II line of business level and in the currency they are likely to incur.
To be able to capture the true best estimate of all possible outcomes some considerations to future events is needed. This includes binary events, extreme events, specific latent claims etc. This allowance for ENID is an issue where a wide range of methods are used. The concept is still being developed and depending on the specific business it may be a more or less import part of calculating technical provisions.
A previous example was asbestos. Other examples could be:
- Electro-magnetic fields
- Nuclear waste
- Global warming
- Court rulings
- Political intervention
- Retrospective change in conditions
- Meteor strikes
Practical suggestions on approach:
- Use history (not company specific) as a guide
- Asses relevance and exposure for the insurance company
- Calculate the difference in mean between the 1 to 200 year point a the tail beyond 1 to 200 year for a specific distribution.
One way to see it is as the difference between the mean of reasonable foreseeable events and mean of all possible future outcomes.
As a minimum (re)insurers must segment their obligations into prescribed Solvency II lines of business when calculating technical provisions. Further segmentation may be needed to apply discount rates or geographical diversification.
A complication is that the Solvency II lines of business often doesn’t reflect how the business is sold or managed. One policy may therefore have to split into several lines of business. Whether this is easily doable or not obviously depends on the data and the systems the company uses.
Excluding reinsurance, the Solvency II lines of business in general insurance are:
- Medical Expense Insurance
- Income Protection Insurance
- Workers’ Compensation Insurance
- Motor Vehicle Liability Insurance
- Other Motor Insurance
- Marine, Aviation and Transport Insurance
- Fire and Other Damage to Property Insurance
- General Liability Insurance
- Credit and Suretyship Insurance
- Legal Expenses Insurance
- Miscellaneous Financial Loss
Technical provisions for most non-life insurers will be calculate as the sum of a best estimate and an explicit risk margin.
The risk margin is calculated using a cost of capital approach. The calculation uses the assumption that at time 0 the Own Funds needed is the SCR. This is SCR is however a subset of the usual SCR. For instance only unavoidable market risk has to been included.
In most cases the risk margin is calculated in total and then allocated to the different lines of business.
At the moment the Cost of Capital rate subscribed by EIOPA is 6%. Pleas find illustration of run-off below.