Make sure your reserving techniques are right while market conditions are good – or you may come unstuck in the next soft market, says Raj Ahuja of EMB.
No one should be surprised that reserving has returned to plague the insurance and reinsurance industry (see IQ Summer 2004 edition). Throughout the soft market that ended in 2001 the impressive results announced by sections of the company market generated a great deal of scepticism that the figures just did not add up.
Then September 11 came along and made it acceptable to set aside billions of dollars of additional reserves. Yet this atrocity – uniquely costly though it may have been - was not the main reason why so many companies had to find so much extra capital. The 90s had been a decade of technically deficient underwriting, camouflaged to some extent by inadequate reserving that was bound to be exposed sooner or later.
Right now Converium are feeling the heat, but they are just the latest high profile company to admit to getting things badly wrong, and they may not be the last. Independent analysts estimate, for example, that the US primary market currently has a 15% - 20% deficit on its reserving, equivalent to up to 30% of free assets. The knock-on effects for reinsurers are likely to be at least equally as severe.
In theory, this should not matter if you are an underwriter. As long as you are making healthy profits now, why not let the past take care of itself? The reality, though, is different; just wait until the next soft market. We will then discover that the future belongs to those underwriters who pay attention to the downside of their business, as well as the upside, and that means devoting time and resources to reserving.
Accurate reserving is the cornerstone of any successful re/insurance business. It tells you which areas of underwriting are doing well and which ones are doing badly and it underpins such crucial areas as performance measurement, pricing and financial modelling.
The information it provides is essential when making strategic decisions about capital allocation, outwards reinsurance, the markets you are in and relationships with ratings agencies. Getting your reserving right does not guarantee success in these areas, but getting it wrong makes the rest a lottery and will ultimately undermine confidence in your business.
Although some companies and Lloyd’s managing agencies have reserving processes that bear close scrutiny, too many continue to rely on luck. Even worse, some use reserving to smooth out the good years and the bad. Most get away with it most of the time, but the consequences of failure can be dire.
We work in a difficult enough business without jeopardising our operations unnecessarily in this way; and we do not have to. This article looks at the question of reserving (and account rating) from three different angles: human, technical and management.
Some observations on human behaviour
Overconfident, overoptimistic
Human beings are inclined to be overconfident and to overestimate their ability to use information, according to behavioural psychologists. And we are not as objective as we think. To quote the songwriter Paul Simon; “A man hears what he wants to hear and disregards the rest.” The international re/insurance market is living proof that they are right.
Averages mislead
Humans like to quote averages, even when they are misleading or meaningless. Hence the observation that the average person has 1.9999 legs, therefore most people have an above-average number of legs. So it is with insurance and reinsurance. On average, your quotes to brokers may be realistic, but which ones will actually be written? Too often, those that err on the side of generosity are the ones that end up being accepted by insureds/brokers.
The herd instinct and anchoring
Once an underwriter has stepped out of line and made a low quote, there is huge pressure on the rest of the market to follow or compete. It takes a very strong or well-managed person to resist. Brokers like to use previous quotes as anchors. Of course, they are only too happy to ‘up anchor’ whenever prices fall. But when they are on the rise, the anchors are used to prevent increases. Good data derived from accurate reserving and benchmarks stiffen resolve because you know when you are being led down the wrong path.
Moral blackmail
If all else fails, try a bit of moral blackmail. No underwriter likes to be branded ‘inconsistent’. Again, good data and benchmarking is the most effective defence.
Why reserving fails – and how to ensure it does not let you down
The basic aims of any reserving exercise can be described like this:
Data + Assumptions + Methodology = Reserving
If all three functions on the left of the equation are properly executed, the rest should be easy. Yet all too often we find that data is poor, assumptions poorly thought out and methodology weak or inappropriate.
Old reserving methods fall short in soft markets
Typically, it is not just the price that becomes more generous in a soft market – so do the deductibles, limits, terms and conditions and sometimes also the duration. One of the basic rules that insurers ignore is that a soft market requires new data, new assumptions and new methodologies in their approach to reserving.
Gathering your data
It is a truism that you can only reserve and forecast accurately if you have sufficient high quality data, yet many insurers and reinsurers continue to fail at this first hurdle. This may be attributed, in part at least, to various key departments not communicating with each other and not fully appreciating each other’s data needs. For example, the collection of a uniform exposure base per class of business, new and renewal rate indicators and coverage indices from the underwriting department.
The ideal data will be detailed, comprehensive and reconciled with audited accounts. The necessary steps can have implications for the whole operation. There must be an appropriate system for capturing and interpreting data – and this, in turn, will depend in part on adapting IT systems, the way that claims are logged and, in some cases, how policies are recorded.
Rubbish in, rubbish out
Successful reserving also depends on accurate assumptions. This is, perhaps, why certain lines of business tend to be more difficult to reserve than others, such as any Liability class, especially emanating from the US. Assumptions about trends, for example, can make or break a reserving exercise. The adequacy or otherwise of case reserving can also be critical; this type of information should, where possible, be checked.
There is now a pretty comprehensive range of benchmarking tools and software products to help you test your assumptions. This should make it possible to extrapolate development factors, loss ratios, frequency and severity distributions, how changes in terms and conditions might affect profitability and the likely long-term effects of rare but intense catastrophes.
Shifting sands
How many times have finance directors been caught out by losses emanating from underwriting sources the business had never even considered or on a scale that defies credibility? We may not know where the next asbestos or 9/11 will come from, but we know it will come from somewhere. Conventional risk loading is not enough to understand the cost of a long-tail contract. If you really want to understand what you are being offered, maybe build in a new category: EBNR. “Exposed But Not Reported.”
Interpreting your data
Once you have taken these steps it is relatively easy to analyse your exposures using the various different software packages available. This will help you to achieve a credible aggregate reserving requirement, supported by a good understanding of variability. It will tell you where you are most and least profitable, where your key vulnerabilities lie and whether your risk is sufficiently balanced.
Pulling it all together - some strategic and management questions
Actuarial methodologies alone cannot determine management strategy, but they do ask some questions. In light of the above comments, here are some relevant questions to senior management (in no particular order):
• Are your underwriters, actuaries and claims managers in the same camp and communicating well, understanding each other’s needs and data sources?
• Do you have the systems and software to gather, store, retrieve and analyse data and is your data recorded consistently and regularly cleaned?
• Have you identified your most vulnerable classes of business?
• Is your reserving a genuinely independent exercise? Is sufficient independence built into the process to protect against the inevitable pressure to be seen to meet earnings/solvency targets or to justify previous quarter end estimates?
• Is your reserving treated as a strategic function that genuinely supports the Board’s business objectives? Or is it treated as simply another task?
• Have you added sufficient loading for risk premiums using stochastic methods (ie those that give you a full range of possible outcomes and their probabilities)?
• When rates are falling, do you allow sufficiently for the soft market?
• Have you allowed for possible new claims sources?
• Using the analysis gathered in your reserving exercise, can you demonstrate that your portfolio is sufficiently diverse and/or do you have the critical mass to withstand exceptional losses that don’t penetrate the reinsurance?
• Have you considered reserving quarterly so as to quickly catch adverse trends?
Any insurer or reinsurer who can answer all these questions in the affirmative – and maintain these disciplines as the soft market approaches - is well placed to be among those that prosper whatever the conditions.
This article appeared in London Insurance Insider in Autumn 2004
