Dear Friend,

I’ve been reading a great book this week. It’s called “The life of Pi” — I suppose it’s the latest in the long line of the school of magical realism, invented and popularised by Gabriel García Marquez back in the seventies and eighties and expanded upon by authors such as Patrick Suskind, Isabel Allende and Louis de Bernieres.

Anyway, it contains a superb yarn at its heart and I’d thoroughly recommend you read it. Our hero, the “Pi” of the title, is the son of an Indian zookeeper who somewhat fantastically finds himself shipwrecked in the middle of the Pacific. So far so “Survive the Savage Sea” — until I tell you that in the lifeboat with the teenage boy is a zebra with a broken leg, a large female orang-utan, a hyena and a magnificent Bengal tiger!

It struck me that this action adventure might make quite a fun allegory for the state of today’s market. Reinsurers do often seem to be cast adrift on a vast ocean, at the mercy of the elements. They’re all in the same boat, yet they also compete with each other for food and water. Occasionally one lashes out and takes one of the others out of the game — others die of their wounds, or succumb to dehydration or starvation.

I won’t spoil the book by telling you exactly what happens, but the end result is perhaps inevitably one animal on top and a lot of blood and bones at the bottom of the boat.

Do try not to have nightmares!

Dear Friend,

It’s strange how fashions come and go in business. The only certainty is that once every generation fashions do an about turn and go back to where they came from.

Today’s Swiss Re takeover of GE insurance solutions only makes the whole process more fascinating.

In the post war era — corporations were all about building diversity — big groups all aimed to become conglomerates. The Bill Gates’ of the day all set out to own vast swathes of non-correlating industries, with the thinking that extreme diversification was the best way of protecting a group from putting too many eggs in one basket.

Hence an industrial machine-maker was added to a raw materials producer, which was then blended with a consumer goods group and a major retailer:- a bank was bolted on and probably an insurance company was welded on for good measure. A good year for banking could offset a poor year for the raw materials and vice-versa. Theoretically this grouping would produce predictable earnings growth.

This conglomerate trend probably reached its peak in the 1960s in the States when a few high-profile collapses knocked some of the sheen off the strategy. However, this way of organising big business still held sway until the mid-eighties, when the process suddenly swung into reverse.

Instead of looking at the positives of diversification, analysts started to concentrate on the negatives — instead of being prudent it now began to mean that you were a bit of a “jack of all trades” and probably a master of none.

The new mantra became “sticking to core competences” and concentrating on markets and sectors in which you were most successful and most likely to dominate and dictate terms. London stock market slang called this process “sticking to your knitting”.

Suddenly demergers and spin-offs became the order of the day. The insurance industry was no stranger to this trend — although, due to its innately conservative nature, it was a little slower on the uptake. The new style really kicked in back in the hard market years of 1992-94. Then pretty much overnight all the insurance composites stopped dabbling in London market business and firms started the long process of demerging their reinsurance and primary operations.

The creation of the Bermudian monoline catastrophe reinsurer also occurred at this time. The powerful logic was to be a specialist and a leader at what you do — and it makes an awful lot of sense. Allstate, Allianz, AXA, Aviva, AIG and Zurich could concentrate on doing what they do best, whilst leaving specialist reinsurers to be strong at what they did.

The funny thing now is that diversification is back on the agenda. Monoline cat players are deeply unfashionable, especially with the ratings agencies. In fact start-up supremo extraordinaire, Don Kramer, told us only last week that he thought that the monoline model was pretty much dead.

But the great thing about fashion in business is that it is there to be ignored.

Look at GE — perhaps the last emperor from the old conglomerate world — and a company whose interests stretch from aeroplane engines through to store-card finance. Only now is it is exiting the insurance business — decades after similar conglomerates broke themselves up.

Fashion doesn’t pay the bills — in the end the only thing that matters is results. If the insurance results had been better GE would probably still be in the (re)insurance game and happy with its investment.

So if you want to start a monoline property catastrophe reinsurer — I say forget what everyone else thinks and go for it! Now is as good a time as any other — and you’ll probably be back in fashion in a decade or two.

Dear Friend,

Fitch has been pumping out some high quality research of late and yesterday outlined their latest attempt at making sense of the more esoteric end of the world of (re)insurance.

The work highlights the difficulty in trying to stress-test (re)insurers when looking at events that have a 1-in-100 or lower probability of recurring. As we all know, this task is ultimately as impossible as trying to nail jelly to a wall, as one man’s 1-in-250 is another’s 1-in-50, but Fitch deserve plaudits for giving it a go and trying out some a new method.

The general idea is to try and differentiate better between how low-probability events impact upon different (re)insurers. At the moment Fitch only looks at one point in the probability curve — the 1-in-100 year event — and this throws up inevitable distortions.

This is because the way that you structure your book and your reinsurance protections can have a huge effect on how less probable events hit you. Fitch cited the example of two notional companies that both suffer the same $100m loss from a 1-in-100 year event but a 1-in-500 year event might leave one with a $200m loss and the other with a whopping $1bn loss — five times as bad as his neighbour — ouch!

At the moment the two probably get identical ratings, which obviously can’t be right.

So quite rightly Fitch wants to try and take an average of the impact of all the less probable losses towards infinity, all the way down to 1-in-10,000 year events. And here comes that clever T-VaR bit — the T-VaR (or Tail value at risk) is the average of all the loss scenarios above a certain threshold.

Leaving the maths lesson aside for a minute, it doesn’t take a genius to see that if you start adding in 1-in-1,000 year scenarios, where you once measured only 1-in100s, the overall ratings effect is going to be negative and the industry is going to have to find more capital if it doesn’t want its ratings downgraded.

Anyway – be prepared — the new methodology and model is being bolted together over the next three to six months, when it will be let loose on ratings.

It’s surely another good reason to get down to Wall Street and join the queue for more capital now while stocks last!

Dear Friend,

My old boss used to observe wryly that it’s a strange and endearing human characteristic that we all end up believing what we think we need to believe when we need to believe it.

Let me explain — the dotcom investor buys shares in a company valued at a thousand times paltry sales because he is so sick of seeing the prices of dotcom shares double and quadruple that he finally gives in. Even though he doesn’t have a clue why he’s doing what he’s doing, the guy somehow convinces himself that this is a new era and that in any case, “stocks always go up in the long run”.

The same can be said of a property investor buying house at the height of a speculative bubble — believing that “property always goes up”. Even though history will show that rapid price rises are almost always unsustainable and followed by painful falls. Buying into such bull markets may be a 100% gold-plated sucker’s strategy, but it doesn’t stop people believing all sorts of strange things when it suits them.

Such popular beliefs are transient and almost always evaporate in a puff of smoke once they are shown up for the wishful thinking that they really are.

So on to the reinsurance game — and we’ve got our own fair share of wishful thinkers out there — in fact in reinsurance we probably have quite a lot more than our fair share.

In the reinsurance version of popular beliefs and delusions, a reinsurer who has been badly hurt by a large loss always says that there will be no new start-ups coming in to compete with him (but in the same breath always adds that rates are sure to skyrocket now that capacity has been dented).

Why does he do this? Because he needs to — no-one in a recovery phase, having to pay out and replace 20-30% of capital wants to see legacy-free competitors sharking in on juicy-looking business. Hence the cries this autumn of “no class of 2005”.

Well, surprise, surprise, there is a class of 2005 — four so far — Don Kramer’s Rosemont Re II, Safe Harbor (or Chubb Re II), Amlin Bermuda and Lancashire — the new reinsurer to be headed by Richard Brindle. And there are five if you count Jeffrey Greenberg’s as yet unnamed and unconfirmed new venture.

Brindle was previously one of John Charman’s star underwriters on Lloyd’s syndicate 488/2488 and was reported to have pocketed £20m when Charman’s agency Tarquin was bought by ACE back in 1998. He then retired and I have no knowledge of what he’s been doing since then.

If I ever had £20m, I’d retire too — but I suppose after seven years I’d probably get bored, especially if I was as good at what I do as all accounts of Richard Brindle’s underwriting prowess suggest.

Maybe he bumped into John Charman in a departure lounge and felt a sudden pang of regret that he hadn’t joined the class of 2001 when he had the chance?

Well, he’s taking his chance now — and good luck to him. In fact, good luck to everyone — this market is looking more complex by the hour.