Reserve releases are keeping many insurers and reinsurers profitable, but have they gone too far and will the practice come back to haunt them? Some companies will soon have to make difficult decisions, says Steve Mathews of EMB.

We read a lot about the underwriting cycle, but much less about the reserving cycle. The underwriting cycle is the observation that insurance company profits and losses tend to move in cyclical fashion over time. However, the profits disclosed by insurance companies are based on current reserve estimates. These may not be robust. The reserving cycle is the observation that reserving models tend to under-estimate current-year reserves during soft parts of the cycle and over-estimate current-year reserves during the hard market.

Rating indices are often used to indicate where we are in the insurance cycle. Even for the longest-tailed classes of business these are typically available on a quarterly basis giving an up-to-date picture of the market. Indeed, there are plenty of pundits who will tell you whether or not the market as a whole is writing at a profit, and word quickly gets around when a particular company appears to be pricing aggressively.

However, it takes time for losses to actually develop to ultimate so as to test how accurate the reserving was. As the market softens, policy conditions are often relaxed, exclusions are removed and deductibles reduced. Traditionally, ratings indices have struggled to reflect the effect of these changes and have consequently under-stated the true rate change. In addition, as the market softens it has often been observed that loss development patterns can start to lengthen, this has been partly attributable to multi-year deals being written to ‘lock in’ the soft rates. Both of these factors are difficult to build into standard reserving models so loss reserves can easily be under-stated in soft markets – hence the reserving cycle. And this is without considering the effect of pressure that management may put on the reserving function to support a business plan loss ratio or profit target.

The realisation that reserves are under-stated will happen quickly in short-tailed lines but can take years or even decades for long-tailed lines. When business is written in a hard market reserves may initially be overstated and as the market turns soft current year reserves may be understated. By happy coincidence, for some long-tail classes this is about the same time that the over-reserving in the back years becomes apparent and these back year releases can be used to offset marginal or even loss-making business in the current year.

This can be a sensible, strategic way to manage the business provided you get it right, but it often involves fine judgement. We know that the history of our industry is littered with apparently successful entities that failed because they pushed this too far and wrote business they considered was marginally priced but only later realised had been seriously under-reserved.


Many well-informed commentators are already sounding the alarm bells. AM Best had this to say about the Bermudan market earlier in the year, and one could make similar observations about other territories: “Although improved significantly over the most recent five-year period, accident-year loss reserves are deteriorating. AM Best sees the sizable reserve releases of 2006 and 2007 as unsustainable. These factors point to concern over the long-term operating performance of the rated entities on the island.”

In theory, it should be much more difficult to under-reserve than it was during the last really soft market ten years ago. Since then, many companies have created Enterprise Risk Management (ERM) frameworks. These will typically ensure rating levels are monitored, business plans are derived from robust models based on historical performance and that pricing experience is fed back to reserving and vice-versa. However, as with any process, the results are only as good as the assumptions that go into them, and there is a lot of pressure on companies to produce good news for their shareholders.

AM Best clearly suspects that there may be a divergence between theory and practice: “As much as enterprise risk management is touted, the coming years will clearly test the execution of this discipline.”

Let’s now briefly take a more detailed look at one particular market, UK Motor. Although narrow in its focus, I have chosen it partly because the data is plentiful and partly because the recent record level of reserve releases has raised a few eyebrows. The issues that emerge apply to other markets as well.

During 2007, insurers subsidised the UK motorist by more than £1bn as competition drove down premium rates to uneconomic levels. The back-year reserves released cut the combined ratio by 13 points to just over 102% - respectable, but hardly mouth-watering.

On closer inspection it is misleading to look only at the market average. Some insurers actually strengthened their reserves during the year, whilst others depleted theirs by the equivalent of 30% of premium income.

Are these insurers being responsible and how long can they keep it up? In fact, the reserve releases are not as dramatic as they seem because, during 2007, the Fourth UK Bodily Injury Study revealed there had been a sustained and sharp drop in the long-term rate of personal injury inflation. This meant that claims departments, having reserved prudently in earlier years on the assumption of 10%+ inflation, could legitimately let go a little.

Nonetheless, after several years of releases, the market is clearly heading for a crunch. At this rate, we estimate the pot will start to empty in about two years, but it will not be as simple as that. Some companies can keep going longer than others; there are ‘haves’ and ‘have nots’. And those that run out of cash first are going to face some really tough decisions.

They could put up their prices and accept a sharp drop in market share or they could try to raise capital to stay in the game or they could cross-subsidise. In practice, some of them will have no choice but to withdraw and may even struggle to keep going as independent insurers.

Exactly the same forces are in evidence within the wider international market. Some insurers and reinsurers are in a stronger position to take part in this game of poker than others. Unless the reserving cycle suddenly ceases to hold (and there is little sign of that happening as prices continue to head south), those less well capitalised players will be forced to show their hands and leave their particular markets.

That is when a third, little talked about cycle may come into play: the run-off cycle. It has been an unfailing rule that, as re/insurance prices go down, under-reserving is later exposed and then the number of companies going into run-off rises after a short time lag. Maybe in this new world of ERM that particular link will be broken, and the concerns expressed by analysts will prove unnecessary; we shall see.


For the reasons already discussed above, the eventual outcome of the cycle for re/insurance companies will depend to a significant degree on how successful they are at avoiding the perils of under-reserving whilst satisfying the demands of shareholders.

In these circumstances it is vital that they maintain the integrity of their reserving process and are aware of the presence of the reserving cycle. In the soft markets of the late 90s some in-house actuaries came under tremendous pressure to be generous in their assumptions, to base their decisions on the more optimistic scenarios. In fact, rigorous reserving and an awareness of the reserving cycle can be a life-saver for re/insurance companies. It forces senior management to make difficult decisions that they might otherwise put off, to deal with weaknesses before they escalate into something worse. If this means doing more to educate shareholders and their representatives to take a slightly longer-term perspective of their interests, especially in so-called ‘Anglo-Saxon’ economies, then so be it.

Looking for more fundamental, longer-term remedies, the best way to avoid these pressures is to link the pricing and reserving functions rather than treating them as separate exercises. Sometimes we overlook or ignore obvious points, and it is clearly the case that profitable underwriting makes it much easier to reserve adequately.

Pricing processes are moving forward rapidly on all fronts: commercial insurance, reinsurance and personal lines. Not only are they much more technical, but they increasingly take the behaviour of the market into account. Even in relatively data-sparse commercial lines, it is increasingly possible to price accurately, or if nothing else, price more accurately than your peers. Above all, pricing should go to the heart of a company strategy. It asks questions about capital allocation, risk appetite and resilience.

By the same token, a modern reserving exercise will provide detailed information about those segments that are more or less profitable than others, and so feed into the pricing and other management functions. Both exercises demand the same detailed information, although the pricing also requires a feedback loop that enables speedy reaction to adverse developments. And both make demands on data storage and retrieval systems.

It is exactly this kind of joined up management thinking that comes together within ERM to provide a framework that should enable re/insurers to ride the peaks and troughs of the pricing/reserving cycle. This does not mean that underwriting or senior management will ever be driven by the numbers coming out of these models; far from it.

Faced with identical conditions some firms might make a strategic decision to price aggressively for market share, while others would opt for a more cautious profit-driven approach. The important thing is that, whatever the approach, it will have implications for your reserving. Through ERM, it is possible as well as desirable to anticipate and provide for the consequences of your pricing decisions.

We shall soon see whether companies have the culture and the will to make it happen.

Steve Mathews is a director at EMB, the actuarial and business consultants.