Why exchanges crashed: exchanges were tried in every industry and the big players like Ariba and Freemarkets are today trading at a fraction of their value five years ago, But these firms were successful mostly in the trading of commodities and raw materials, not reinsurance

Risk & Insurance, Sept 15, 2004 by Howard Green

A lot of people support the idea of reinsurance exchanges. There are some persuasive and well-meaning folk that actively promote the concept. I also know the industry has spent tens of millions experimenting with exchanges. But I must be level with you from the outset: I think reinsurance exchanges are a fundamentally daft idea.

I cringe when I hear the word "exchange" spoken in the context of insurance. Reinsurance is neither a commodity nor a standardized share certificate that can be priced and traded in an open market. The essence of our industry is being able to craft risk solutions to meet the specific needs of each client and to execute and manage these risk transfers in the most efficient way. We make custom suits, not overalls.

Exchanges have always been an appealing idea in our industry. I guess we all like to see ourselves as floor traders at heart. But what I have observed is that we confuse the concept of structuring deals over an exchange (which is daft) with the idea of exchanging data (which is good).

With the advent of the Web many people rightly saw a better way to exchange data. But these initiatives quickly got caught up in the much more glamorous concept of doing deals over the Web. At first glance the idea of a deal-making hub or exchange is alluring. Everyone submits their offers online as eager counterparties wait to take them up.

But we quickly discovered two important Nets. Exchanges only work well with commodities, where every transaction necessarily has a similar profile. This might be OK for simple property facultative, but introduce anything out of the ordinary and the exchange seizes up.

Secondly, and more importantly, exchanges only work effectively when prices are publicly posted. Otherwise they are simply used to game the price. This is what happened with Inreon. The broker would use Inreon to obtain a price and then use it to game off-line participants into providing a better price. Look at any exchange initiative in insurance and you see the same effect: the ratio of bound to quoted deals is extraordinarily low.

Insurance is nut alone in making these discoveries. Exchange initiatives were tried in every conceivable industry and the big players like Ariba and Freemarkets are today trading at a fraction of their value five years ago. And it's no surprise where these firms had any success--trading commodities and raw materials.

RETHINKING REINSURANCE

A better way to think about reinsurance is as a cross-enterprise business process. Reinsurance does not exist within a given enterprise. It is the constant interaction between the parties that gives life to and sustains a reinsurance transaction. It is a business that depends on interoperability, a feat that is still achieved today largely by exchanging pieces of paper.

Business processes of a similar scale and complexity are common in manufacturing and in many parts of financial services. The tendency there has not been to commoditize the process by forming an exchange. Rather, parties have won competitive advantages by making their version of the business process more efficient. The big industries like food, automobile and electronics may have created sub-markets for sourcing raw materials, but actually the principle supply chain is the opposite to an exchange. Think of GE or Toyota. Companies like these compete aggressively by developing business processes that are superior and more efficient than that of their peers. And that is what reinsurance firms should also do.

Reinsurance too needs to improve the way its business processes are managed. But like -all things, it needs to do it in a way that is specific to the unique dynamics of the reinsurance industry. To automate reinsurance transactions, business processes must work within and across enterprises; handle both the volume of facultative and the complexity of treaty; enable data to be added or extracted at any point of the process through links to back office systems; -embed industry standards like ACORD; and embed compliance and checking steps.

To achieve this, a business process needs three key layers of functionality: standards, processes orchestration and adaptors.

SEEKING AN EDGE

Though an easy concept to grasp, standards are difficult to achieve in a diversified global industry with numerous players and no central authority like the Bank of England or the New York Stock Exchange to lay down the law. Yet enormous progress has been made, thanks primarily to the efforts of ACORD. Adopters of ACORD XML, comprising over half the cedents and reinsurance carriers globally, already report average integration cost savings of 20 percent to 30 percent.

Individual firms should be able to take these industry standards and develop business processes that provide them with a competitive advantage. Powerful technology tools are today available to automate business processes. Because accepted standards are being used to structure these processes, a firm can extend the automation out to their counterparties without fear of rejection.