Once an insurance company has written a policy, we need to think about how the claims against the policy are recognized as an expense, which in turn will affect the amounts owing to policyholders.

So this might seem obvious if the policy starts at the beginning of an accounting year and ends at the end of that year.

But what happens if the policy runs from the midpoint of an accounting year? Or what happens if the policy runs more than a year? Well, as you might expect, the accounting rules deal with this complexity by having a very clear principle, which is that companies must recognize claims only that have been incurred by the balance sheet date.

But what on earth do we mean by incurred claims? Well, incurred claims mean that the insurance company has reasonable evidence that a loss event has actually occurred.

This means that the claims expense, and therefore the amounts owing to policyholders, will increase as the insurance cover is provided. This is important, as it's a common misconception that the insurance company must record a liability for all the expected future claims as soon as the policy is written.

So how does the insurance company estimate the claims incurred to date, especially if it takes time for claims to be reported and assessed? Well, a very simple way of doing this is to say that the incurred claims is the amount of premium earned multiplied by the expected loss ratio on the policy. Let's use an example policy where the premium is, let's say, 100 and the loss ratio is 80%. So total claims are expected to be 80 on this policy.

Well, halfway through the policy, 50 of the premium will have been earned, and when we multiply this by the loss ratio of 80%, that would give incurred claims of 40, half of the total expected claims.

Now, this approach is appropriate when the claims against a policy won't vary much over time, and that's, for example, with a car insurance policy.

It's also used when an insurance company has very limited historical data on how the claims would develop. However, when it comes to the amount actually owing to policyholders, clearly this needs to be adjusted for any amounts already paid. So we can show this as a formula, which is that the claims reserves, that's the amounts owing to the policyholders, are the total amounts of claims incurred, less any amounts already paid.

So if we revisit our example that we just used, where the claims expense halfway through the policy was 40, well, if the insurance company has already paid out 10 on these, then the claims reserve will be 30.

Now let's think about the claims reserve in a little bit more detail, and this table shows us how the claims reserve at the end of the year reflects the reserve at the start of the year, plus any claims actually incurred in the year, less any claims already paid out.

So in this example, at the start of the year, the claims reserve is initially nil, which presumably means that the policy had just been written, and the claims incurred in the year are 40. So that's 40 of claims expense, which we'll be showing in earnings.

Then this is added to the reserve, but then we deduct from this claims paid out, which are 10 in this example, giving us the ending reserve of 30. However, we can also rearrange this to work backwards to the claims expense. This would just be the increase in the reserves plus the amount of claims actually paid. But why on earth would we want to do this? Well, actually, we've just seen how an insurance company can calculate claims incurred based on the amount of time elapsed since the start of the policy, and typically, this calculation is done on a cumulative basis.

So after three, six, or nine months, they know the cumulative amount of claims. Also, clearly, the insurance company will know how much has actually been paid out to policyholders over that time.

So this means that they can use the reserves at different points in time to calculate the claims expense as the residual item.