Multinationals want value for money from their pooling arrangements. A new database can show if they are getting it, writes Roger Beech
Multinational pooling is not a new concept by any means. But for many large international companies it is still very much on the agenda. This is because of their increasing demand for greater transparency in employee benefit costs and a desire to their pools are performing at least as well as those of other international companies. In short, the question that companies want answered is whether, based on their own contracts and experience, they are obtaining real value for money from their group insurance contracts.
Pooling works by experience-rating each group insurance contract that participates in a pool and “pooling” the net results. Pooled insured group benefits can include pension, death, disability, medical and accident contracts, provided they are insured with a local carrier that belongs to a pooling network of insurance companies. Each year the network prepares for each contract an account of the premiums received and claims paid out, together with interest accrued, commissions paid, profit sharing etc. The positive results from some contracts offset the negative ones from others. A charge is made to cover the insurer’s expenses, profit and risk exposure and the overall outcome is either a surplus or loss to the pool.
Surplus is usually returned to the employer as a “dividend” and may in turn be shared with those subsidiaries whose contracts made a positive contribution. Losses are usually either carried forward to be experience-rated in future or eliminated using a stop-loss arrangement.
As a response to corporate needs to be certain of value for money, W M Mercer completed an initial pooling research study in 1998. This has since spawned a growing database of pooling experiences.
The study involved analysing the pooling experience of 69 multinational companies with pooled contracts in 78 countries. Over 6,000 separate policies were included in the 275 pooling reports reviewed, encompassing 886,000 life years.
When we put together one large multinational pooling account to cover the database experience, the average annual dividend from loss carry forward pools was 5.6% of premium, which was a little surprising. Although there is no historical data available, this is less than might have been expected in the past. One possible reason is increased competition in local markets and therefore lower up-front premiums. If true then the result isn’t in itself a bad thing but does strengthen the argument that pooling does require active management to reduce the overall cost of insured employee benefits through both competitive local terms and pooling dividends. Other reasons for the figure being less than previously anticipated may be the gradual removal of tariff premiums and the apparent increase in contracts providing local dividends.
One part of the study looked at network retentions and, in particular, how they vary with the volume of pooled premium. Retentions for the purpose of the study are defined as the total of risk and administration charges. Figure 1 plots the retentions expressed as a percentage of premiums for the latest year for each of the pools in the study.
As expected, the chart trend shows that retentions decrease as the size of the pool increases. This supports the premise that a company may achieve savings through lower retentions by combining multiple pools into larger pools.
Also as expected, the chart shows that retentions for pools operating on a stop-loss basis are generally higher than those operating on a loss carry-forward basis.
Interestingly, we have also been able to estimate retentions across a range of premium volumes for each network. This allows a comparison to be made with an individual company’s pooling retentions and a preliminary assessment made of whether its retentions appear reasonable in comparison with the database Figure 2 shows an example for company A whose retentions are higher than the regression line. Of course, this is an indicator only, as both the individual company’s experience and the particular mix of contracts will have an important effect. Nevertheless, it is a useful yardstick and may present an opportunity to negotiate lower retentions.
I mentioned earlier the need for active management to minimise overall insured employee benefit costs. One issue regularly faced by companies is whether to add an additional contract in a pool or not. From what the research shows so far, the inclusion of medical cover tends to reduce a pool’s profits. By contrast, life assurance tends to be the most profitable for pooling while retirement also makes a positive contribution. The impact of disability contracts is less clear, due to there often being large reserves created.
By implication, to improve a pool’s profitability, a company might consider either removing medical contracts or including them in a separate medical-only pool. Such a pool could operate on a loss carry-forward basis as the volatility of medical claims is fairly low and the margins relatively slim.
The database has also enabled us to establish models to simulate and compare the performance of pools using stop loss and loss carry forward techniques. For medium sized pools, say those between £0.5m– 3.5m of pooled premium and with an average distribution of coverages, the simulations have indicated that higher average dividends can be expected under a stop-loss pooling system than loss carry-forward. For ‘average’ pools of less than £500,000, loss carry-forward would generally seem to be more financially favourable: such pools tend to be more volatile, resulting in the networks charging a significantly higher retention for stop-loss. Larger pools (more than £3.5m pooled premium), are generally much less volatile and a loss carry-forward pooling system can be expected to produce superior results.
The table sets out a list of some of the more profitable countries together with the UK and US. It should be noted that the profitability for each country will be highly dependent upon the mix of coverages in each country. For example, for Ireland the database premiums relate mostly to life and disability, while Switzerland includes a mixture of retirement, life, disability and medical.
This type of table again provides an indication of which countries and which contracts to consider pooling. For example, including the UK in a pool requires careful consideration by companies and a review of their own previous claims experience to aid assessment of whether to pool or not.
With a positive contribution of 6% from US contracts, companies not already including them within their pools may wish to take another look. Countries such as Argentina and Greece may also warrant consideration.
Another useful element of the database is that it allows comparisons to be drawn between an individual company’s claims experience and the claims data included within the database for a particular country and coverage. Therefore any variance can be identified and investigated further.
With corporate demand for best value for money and greater transparency in employee benefit costs, the use of objective analytical databases is likely to grow considerably. Therefore, whilst multinational pooling is not a new concept more companies are deciding to take a fresh look at their own arrangements.
Roger Beech is a European partner of W M Mercer and a senior international employee benefit consultant, based in London.
Companies wanting to obtain more
nformation about the survey, or wishing to participate, at no charge, should contact him on +44 (0)171 963 3169 or by e-mail on firstname.lastname@example.org