Why does an insurance company need claim reserves?
When an insurer agrees to provide cover under an insurance policy, it undertakes to pay all claims covered under this policy. In the liability insurance segments in particular, however, some claims are only discovered with a certain time lag, or the claims settlement process takes more than one financial year. To ensure that it can sustain its non-life business over the long term, an insurance company must therefore be able – in the year in which a policy is issued – to make as accurate an estimate as possible of the total claims that will eventually need to be paid. Because only some of these claims have been paid for at the end of any given financial year, the figures used are always forward-looking estimates.
To this end, the insurer analyses the preceding years to ascertain the payment settlement pattern, i.e. after what period of time and in what amounts the relevant payments occur. The payment pattern calculated from this data can then be used to estimate the reserves required for all anticipated payments. Calculations of claims reserves are based in the first instance on accounting rules and on all information available and trends discernible at the time the calculations are carried out. Baloise, for example, applies International Financial Reporting Standards (IFRS).
«Claim reserves under IFRS»
IFRS requires all claims reserves to be best estimates that are calculated using generally accepted mathematical-statistical methods. In the case of rare and extremely large claims such as natural disasters, however, this is especially difficult and only possible to a certain extent.
How do insurers arrive at the necessary payment pattern?
Insurance companies have to process their payment data in an appropriate form, which we call claims triangles. Let’s take the chart above as an example. We’ll take 2012 as the starting point and assume that one insurance claim occurs. The first payments for this claim are then made during the same year (B1 / B2). One year later, in 2013, payments are made for new claims arising in 2013 but, at the same time, payments are still being made for claims that occurred in 2012 and which are allocated accordingly (B3 / B4). Another year later, in 2014, payments are still being made for the claim that arose in 2012 while, simultaneously, amounts are being paid for claims from 2013 and 2014. If all payments are aggregated in a single line, this gives the final amount of claims incurred in a given year after all the claims that occurred in this year have been paid. This information can now be used to estimate all payments that will be incurred in future. The claims triangle is expanded to form a rectangle (C3 /D4). Many different types of calculation can be used for this purpose. The most commonly known of these is the Chain-Ladder-Method, which results in a best estimate.
Calculations of this kind are highly complex and are therefore performed by specialists. This procedure is then repeated year after year – with the parameters for estimates being adjusted in line with prevailing circumstances – until at some point all claims from the year concerned have been settled. Ideally, and assuming there are no unexpected loss events or developments (such as new legislation), all reserves will then be used up exactly. While claims are being settled, adjustments to estimates cause reserves to be released or strengthened compared with their original levels (profit or loss on claims reserves), depending on how accurately the actual trends have been estimated. These adjustments also result in an improvement or deterioration in the combined ratio for the year concerned.