Back to their roots

You may remember sometime ago I wrote about the innocents abroad in Europe (mainly smaller co-operative banks) investing in derivatives, not being quite sure what they were investing in or how much risk they were taking on.

Interesting, therefore, to see that Austrian bank Bawag has agreed to pay at least $675m and to co-operate with prosecutors who are investigating the collapse of stock brokerage Refco.

The failure of Refco was in its own way somewhat tedious. Its CEO had "allegedly" hidden $430m in losses built up by a company under his control. To make matters worse, the company actually had an initial public offering, before collapsing somewhat spectacularly two months later.

Bawag, it transpires, lent Refco's CEO Philip Bennett between $250m and $300m every year between 2000 and 2005 to help him cover up his losses. When a company is floating and when the CEO received nearly a billion dollars from a leveraged buyout in 2004, even the most innocent bank abroad might think that loans of more than $200m a year might be considered relevant by the regulators. But not, apparently to Bawag, whose silence in the lead-up to the IPO might be described as "deafening".

The question that I can see forming on your lips is, "why?". Why would a bank hand out this kind of cash to help a CEO conceal losses? Ahh, well, it seems that some of Bawag's own investments were not going that well, and Bennett helped Bawag conceal the extent of these losses.

More will be heard of this, with several underwriters of the IPO likely to be facing investigation. But the whole caboodle makes you wonder how much you can trust any company's "official" figures. Bawag is no baby; it's a significan Austrian bank.

So, if a big institution seems to be making a lot of its money from a single place, make sure that single place is well-audited. After all, remember Barings.

Exchanges

I love probability. I mean, I really love the philosophical side of probability, which is often what causes many poker players problems.

Take this situation. You have an open-ended straight draw with one card to come. What is the chance of you hitting your straight?

The answer, you might think, is 8/46.

However, one could equally reassonably answer that, since the position of the deck has already been decided, the probability is either 1, or 0.

Well, you might say. That's silly. OK. Suppose the dealer looks at the river card and then looks at you, but does not show it to anyone else. Now, what is the chance that you have hit your straight?

It's at this point that many poker players start tying themselves in probabilistic knots. If you then say that the dealer shows the river card to everyone in the room but you, then what is the chance that you will/have hit your straight? At this point most poker players, instead of confidently saying "8/46", collapse under the weight of the fact that everyone else knows the card, but he doesn't, and says "I don't know". Most poker players, for some reason, find it hard to cope with the fact that an event can have several different probabilities simultaneously, depending on who is doing the looking. It's like Einstein's own theory of relativity, but applied to the poker deck.

I was thinking of this when the exchanges opened up on the chances of Rooney playing in the World Cup. Now, there's a great probability question for you. Bloomberg happily reported: "Exchanges say chance of Rooney playing in cup rises from 35% to 71%". But this was just what the money line was saying. Did it reflect reality? Because here was a case where a probability was being assigned to something for which there were no actuarial tables and no dealer holding an already decided deck (unless you believe in God as dealer and the deck as predestination). All there was was a picture in the Sun of a scissor kick and film of training. Not, when you come down to it, a lot to go on.

I admire people like Tony Bloom who are prepared to sit down to make an educated guess on things like this. It goes so far beyond simple Maths, into I-don't-know-where.


++++++++++++++

I've been four-tabling at $2-$4 this month on Party. At the beginning it was hard and I know that I have made mistakes worth about 20 big bets. I'm still up, marginally, after a large number of hands for this time of the month (4,000 or thereabouts). As I get more accustomed to the four-tabling, I should move closer to my estimated 1.5bb per hundred. I had been clocking up 2.2bb per 100 over the past couple of months, after some successful tweaks. And I was happily trucking along at 1.8bb per hundred until today, when the rails came off and I got cold-decked and rivered to death for a $140 loss in 400 hands. As they say in the trade, no blame. But the $80 or so that I lost to mistakes on other days now makes up more than my profit for the month. However, it's all good r&d. For a start, it makes three-tabling a lot less tiring! Standard deviation looks to me to be about 16 big bets an hour and 16 big bets every 100 hands (i.e., about 25 minutes).

I've also dabbled in the $5-$10 games on Virgin on the occasions when they have seemed good. Sometimes they look like the Bellagio $8-$16 tables on a Thursday morning -- the same old faces. But other nights there are some goldmines of fish. Obviously this puts volatility through the roof, since I'm not playing there that often and I'm only playing when the game is good. This led to a $190 loss last weekend about which I could do nothing, and a bit of a gain back tonight before I jacked it in through tiredness. Well, I do have to be up early in the morning.


__________________