Connect with us
Archive
THE IMPORTANCE OF PORTFOLIO MANAGEMENT
Friday 17 August 2012
Author: Russell Group
Within most specialty (re)insurance companies the actuarial and risk management departments often focus exclusively on building models and embedding processes to technically evaluate and price individual risks that are being considered for underwriting. Many companies will have sophisticated spreadsheet tools with advanced statistical techniques employed to calculate expected loss profiles and corresponding expected capital returns, utilising various proprietary and market data sources. But with almost every (re)insurance company claiming to rate every policy in its portfolio, does this really give them the competitive advantage they think? Alternatively, would their resources be better invested in also understanding higher level portfolio dynamics either deterministically, such as Realistic Disaster Scenario’s, or stochastically by generating portfolio loss distributions of underwriting risk? Is it possible that the market is spending too much time pricing while ignoring the bigger picture?
Pricing
There is no question that the ability to price a risk is a minimum requirement for the successful operation of a (re)insurance company. The continued investment in actuarial and statistical resource within companies in recent years is a trend that looks set to continue, particularly within new Solvency and underwriting control regimes coming in to force. However the degree to which some pricing tools now model underlying risk seems to have overtaken the usefulness of such analysis, particularly when this is done to the exclusion of the same analysis at a portfolio level. This is particularly true of markets where subscription is the norm and pricing across the line of business will tend to be reasonably consistent due to market forces. It should not be forgotten that a (re)insurance organisation will never be judged on the results of a single risk, and that it is always the results of the entire portfolio that matter.
Exposure Management
While pricing has taken centre stage, aggregate control is often neglected or ignored. One common approach to exposure management is to take a “worst case scenario” approach, managing the portfolio and associated capital on the basis of the policy limits that are accepted. Although this approach, is simple, cheap to implement and very conservative, it does not really provide an answer that has any practical use – other than a public statement regarding the extent of the company’s loss from any market event, or series of events. The worst case scenario provides limited information about the portfolio we write as we have no understanding regarding where this event sits on the loss distribution curve, and therefore no way to identify the capital position, or capital requirements, for the portfolio as a whole. Where a worst case scenario approach is taken, any return on capital/equity pricing that is done becomes fairly meaningless as we cannot accurately estimate the capital required for a single risk in isolation due to lack of data concerning diversification benefits. If we are working in an environment where the organisation has a fixed amount of capital for the portfolio, accuracy in our calculations is not really improved as the sophistication used to estimate the expected loss and volatility for a risk becomes diluted used with a capital amount that has limited statistical basis. Either way some of the resource used in pricing would have been more effectively utilised in understanding the portfolio.
The overriding impact of taking the conservative worst case scenario approach to aggregate management is that it results in an inefficient use of capital as the company is writing less business than their capital would otherwise allow. Even where market conditions are such that the company doesn’t want to write more business then this approach does not identify the surplus capital that could be put to other uses inside, or outside, of the organisation.
Market Perception
Having an accurate understanding of exposures is not just important from a capital management point of view, but it is a key element in managing how the company is perceived by clients, competitors, regulators, and most importantly rating agencies. What we have increasingly seen in recent years, following market level loss events, is focus on the company’s relative market share of the loss and not necessarily the profitability following such events. It is during these times of high loss activity that the inner workings of a company are under most scrutiny, with the company’s ability to manage it’s exposure prior to the event, and report on exposures after the event, under particular focus. To most stakeholders, a company’s ability to pay claims following a major event and to be in a position to continue trading is more important than the marginal benefit gained from sophisticated approaches to pricing.
From the investor’s point of view aggregate control is vital, not only in understanding the losses that their assets are exposed to, but also in having the necessarily level of transparency for them to understand how this investment fits with the rest of their investment portfolio. The dangers of not fully understanding the underlying risks were exposed when investors lost money on property derivatives and this is not something that should be repeated in the (re)insurance sector.
Reinsurance and Retrocession
Reinsurance purchasing decisions are often much more important than any single inward decision, particularly to the long-term sustainability of an organisation. To ensure that any reinsurance or retrocession that is purchased is optimal, it is vital that the organisation has an understanding of the distribution of losses from their inwards portfolio, and understand the different types of events that may give rise to these losses. With the market often behaving in a reactionary way, and products constantly evolving, a historical review is likely to be much less effective than a complete understanding of the current portfolio and analysis of possible losses for the coming year. Although most organisations will perform some form of outwards pricing this tends to be done in isolation, perhaps using the reverse of a pricing model used to evaluate inwards treaties. Although this can provide a reasonable estimate of the likely impact of purchasing, or not purchasing the treaty, it is not as effective or as reliable as a thorough portfolio level analysis of the underlying portfolio.
Strategic Decision Making
Finally, for senior management to make pro-active strategic decisions, it is vital that they understand the exposures from the various portfolios under management and that they can assess the overall return levels and not just the accumulated loss ratio across all accounts. In addition the most successful organisations are those that can respond when things do not go to plan. When deals turn out to be less profitable than expected, this can be managed during the cycle and corrections made swiftly to minimise the damage to the company’s capital base. When losses turn out to be more than expected there may not be the opportunity to make the necessary corrections as the damage done may be too great and the company may not be able to continue trading. In this respect aggregate management could be seen as being more important to the long-term health of the company than pricing.
Conclusion
The current market focus on inwards pricing, I think, misses out on the opportunities that can be gained by focussing on portfolio level analysis first and foremost, and deal pricing second. There is no denying that basic profitability is vitally important for an organisation, but having a narrow view of this analysis does not provide the maximum benefit from the resources invested in it. The basic concept of (re)insurance is firmly rooted in diversification and portfolio benefit and not that any one account in isolation should be seen as worthy of investment. The key to running an ultimately successful risk management function is using sufficiently sophisticated pricing models that can also scale-up to a portfolio level view .
