November 2006

Hub-and-spoke versus Peer-to-peer.

Is it really as simple as that ?

Fashions come and go. It's a law. I saw some spectacular pink platform boots just last week surrounded by a pair of flared jeans you could camp under. Not a year goes by without one thinking "whatever happened to….", or "look, they're back again". Paper bags, 35mm film, 4x4s, bakelite, Duran Duran. Everybody has a different list of candidates and, that's the key here, few people can ever agree on what should be in or out of fashion unless the media is telling them.

We, of course, ensconced in our ivory towers here in the City are completely immune from such things. Or are we ? I rather think not.

For some time now, the long-awaited revolution that is occurring in the London Insurance Market has been gathering pace, accelerated to a great extent by the availability of feasible, cost-effective business communication technology that, among other benefits, enables Peer-to-peer communications.

Peer-to-peer, by way of explanation, is the opposite of Hub-and-spoke. Recall, if you will, the London Market bureau operations that were the LPC and the LPSO - now Xchanging Ins-Sure Services - which were, and still remain, classic Hub-and-spoke operations whereby market participants connect to a central hub which undertakes the vast majority of the complexity and unpleasantness of the business, IT and communications processes. Another example you use almost daily is the internet where most things are, or heavily rely upon, some form of Hub-and-spoke architecture - you connect to ISPs, aka Hubs, and use functions on, or upload to, or download from their servers.

Peer-to-peer does not require a central service Hub, making internet connections instead directly between the trading partners who, in the London Market, can be as diverse as Brokers, Underwriters, actuarial firms, outsourced claims managers and policy processors, contract certainty checkers, and so on.

It all seems perfectly sensible in a world where reliance upon the central hub is unneeded or unwanted. Some elements of our Market have never completely warmed to the presence of a single service provider in the middle and when the doomed Kinnect proposition offered a similar Hub-and-spoke solution, it was soundly dismissed. Architecturally comparable offerings, albeit in specialised market segments, from RI3K and e-Reinsure are slowly taking up members but it seems to me that if the Market wanted to adopt Hub-and-spoke across the board, then Hub-and-spoke offerings would be experiencing huge take-up, which is not happening yet.

So, rightly or wrongly, Hub-and-spoke has most certainly gone out of fashion and Peer-to-peer is "in". However, to suggest that the market is "going Peer-to-peer" is a gross over-simplification. There remains a real need for platform boots, flares, 4x4s and paper bags in this market. I shall explain.

First of all, there is the matter of complexity. A completely Peer-to-peer market consisting of, say, 100 brokers and 100 underwriters, ignoring all the other service providers for simplicity's sake, could have 10,000 different interchange paths and that's a lot of spaghetti - each strand has to be established and managed and this raises the question of cost-benefit.

It is not a simple job getting a Peer-to-peer relationship established - business process and technology changes need to be implemented and tested, some essential business processes that are undertaken at the hub will have to be taken in-house or outsourced, accounting and legal interchange agreements have to be reached, and so on. None of this expenditure would necessarily benefit a small account relationship to a cost-justifiable extent. Worse still, some smaller firms may not be able to muster the resources needed to get involved in Peer-to-peer at all.

Then there is the market architecture: London is not Bermuda less the sunshine. It is still a subscription market and this gives it particular strength in certain types of business. It also still has a diverse range of firms, from the very large to the very small indeed, which makes it hard to manage, and one size cannot necessarily fit all. Subscription markets are highly frictional, mandate considerable duplication of process and technology, require central service hubs to make them more efficient. Large-scale consensus is needed to change and move on, so revolution in this environment is not an easy task.

Now, it's certainly true that the average number of lines on a slip these days is far lower than in days of old - down from 20 plus to around 3. But 3 is not 1 so, although many big accounts may be based on 100% singletons, many more remain resolutely subscription-based. How can the market manage this with solely Peer-to-peer ? The answer is that it probably can't unless every participant is prepared to run its own subscription handling systems, that don't yet exist, which is straying into the realms of fantasy.

So someone needs to stay in the middle and offer this sort of service to those relationships that cannot go Peer-to-peer. Whether that's XIS on their own, or a selection of competing and/or complimentary providers including XIS, RI3K, e-Reinsure and other potential players, remains to be seen but one thing is certain: Hub-and-spoke is not going away yet.

So, just as the current fashion in the motoring world is for Hybrids, so must the fashion be in the London Market - a hybrid solution will evolve as the way forward. The Market is almost certainly going to change beyond recognition in the next 5 years but it won't be all Peer-to-peer. Hubs will be run by service providers and Brokers, surrounded by parties both small and large, taking advantage of the niche benefit offered by those hubs. Peer-to-peer relationships will be established carrying large volumes between individual players with 100% lines, or by small groups of players handling their own subscription mini-markets. Service providers will be peeking in along the Peer-to-peer routes, adding value and passing the data along. An interesting vista indeed.

There is one big area of common ground in all of this, though, and that is the communications technology. XML messaging communication to ACORD standards is pretty much the only cost-effective way to connect your systems to either a Hub or a Peer. The EDI standards of old are far too cumbersome and expensive to implement and maintain. Equally, parochial standards established between bi-lateral partners or within small trading groups are horrendously difficult to establish and can only serve to cut that group off from the surrounding market and/or force everyone else's costs up by making them develop yet another non-standard protocol.

Interestingly, few market participants outside of the G6 have yet undertaken any serious research into this new technology and fewer still have made any investment it. So, if you're currently unsure as to how your organisation is going to be positioning itself in this new London Market vista, there's one piece of investigation you can get to grips with right now - how are you going to communicate with your Peers and/or the Hubs ? You can't use semaphore.

And next time anyone suggests that the market is "going Peer-to-peer", let them know that they're probably only half-right. Peer-to-peer may be fashionable and it's coming to the market right now, and Hub-and-spoke may well be unfashionable but it's going to remain a necessity for years to come. Flared jeans have come and gone at least twice a decade since the 70s, 4x4s can now be hybrids too and, quite soon, I'll wager that paper bags will be back with a vengeance.

Jeff Ward

TriSystems Software Limited