It is fair to say that the “current” debate on the electronic future of the insurance industry is, in fact, the same as it was several years ago. And yet it continues to be both confused and inconclusive. On the one hand, the whole market recognises the imperative need to improve performance and lower costs, and that this can only be achieved by reducing the amount of manual effort required to process business. Everyone seems to accept that a much higher degree of automated electronic processing is inevitable, both for policy placement and subsequent claims administration. The industry is fully behind that principle.
On the other hand, there have been enough failed attempts to impose a market-wide mechanism, accepted and used by all, to place many doubts in peoples' minds that it can actually happen. There are plenty of examples of new initiatives that simply were not able to get sufficient take up by the insurance parties (brokers or underwriters) and hence have fallen away. In London, the Electronic Placing Support (EPS) is a classic example while, on a more global scale, neither World Insurance Network (WIN) nor Reinsurance Network (RINET) were able individually to get sufficient support.
Why should widespread electronic automation be so difficult to realise, at least in the general commercial (re)insurance world? In the banking sector, particularly, many millions of transactions are successfully exchanged between different organisations every day without batting an eyelid. What is it about the (re)insurance business that makes it so different from others? It is true that, in the (re)insurance industry, transactions flow from end to end of the process seemingly very much faster and with only a fraction of the amount of human intervention. There can be any number of explanations for this.
First, the participants on the ground either have to really want it or there must be an all-powerful central body with enough clout to make it happen. The down side of electronic automation has to be counted partly in employment terms - and people do not generally organise their own funeral.
Historically, banking has had a much more centrally regulated mentality and improvements for the good of the market can be imposed. This is not necessarily so in the global (re)insurance sector.
This lack of central regulation means that each insurance organisation generally seeks advantage over its competitors ahead of its responsibility to the whole. This is the inevitable consequence of an unregulated competitive market. Without a strong independent force to impose collective changes, they are unlikely to occur unless the market players can see real benefits for themselves.
Second, for people to adopt new ideas - such as automated trading - they must feel confident in the integrity, reliability and suitability of the process. We have already established that there is a good deal of ammunition for sceptics to work with to convince others that it cannot ever happen. If a new venture does not work right at the first time of asking (or even the second time), then confidence will be very low and the subsequent attempts proportionally more difficult.
But perhaps the most critical success factor for widespread automation lies in the creation of a set of data standards that will really work for the business of (re)insurance. Perhaps the key difference between the sectors that have successfully introduced highly automated electronic trading and the insurance industry is not that a single process can be imposed, but whether it can be defined in the first place.
(Re)insurance business is very complex in structure. The variety and extent of information required to fully understand an insurance proposal or claim demand is typically larger than, say, a banking transaction.
No underwriter or claims handler is going to agree to proceed unless they are totally confident that they have all the information there is available and that it is both complete and accurate. Ignoring the low value/high volume domestic insurance market (for which e-commerce will clearly be the accepted channel), it is incredibly difficult to define a process in which all the necessary information can be exchanged electronically and with a high degree of automation. While the human influence on success cannot be ignored, one of the most common reasons given for the market failures to date has been that “it does not work for my class of insurance”. The implication is that the amount of information required for one class of (re)insurance is very different from another, and no single business process will satisfy all.
Other strategies are, of course, totally viable, such as sharing a common information pool via market data repositories or the exchange of largely soft information, such as documents and e-mail. In London, initiatives such as the market wordings database (MWD) and Worldwide Insurance e-Commerce (WISe) (which incorporates the WIN and RINET offerings) are getting substantial backing.
If, however, there is to be a future for automatic electronic insurance trading via a clearing-house type system, then the creation of workable data standards agreeable to the majority is critical. When this challenge can be met, then it will be possible to realise all the performance improvements and cost reductions that the market needs. Indeed, a very significant start has been made with the XML (extensible markup language) standards from the International Underwriting Association of London/Lloyd's protocols and standards group. But it must be recognised that this will never be able to cover absolutely every piece of (re)insurance transacted. Further, the inability to cope with the exceptional few should not be allowed to prevent the process for the remaining majority.