Nov 062009

I just read a very interesting article by Malcolm Gladwell. He is a very interesting author. He has written 3 books in the last few years, all best sellers. The Tipping Point, Blink, and Outliers and now has a book of his best articles from the New Yorker magazine called What The Dog Saw. His is my favorite kind of writing in that he seeks out the uncommon wisdom, the hidden connections in a story.

I read two of his articles on finance, and the one about Ceasar Milan, the Dog Whisperer in his latest collection. The first article on finance was about Enron. It turns out that Enron failed for the very same reason AIG did. They allowed MIT PhD physicists to make their financial decisions. These guys used very complex financial instruments that required mark to market accounting, which means that what they put in their books is what the estimated value of an investment will be in ten years, based on its current value. Their predections turned out to be wrong and the books in real life went bad.

The executives, when asked, admitted that they did not understand the deals, or the implications of the deals they were making. But they had confidence in the genius of the mathematical models and the wizards that explained them.

I was very intrigued to learn that Enron did not hide as much information as the media would lead you to believe. The transactions were so highly complex, however, that there were literally hundreds of thousands of financial and legal documents “disclosing” the details. They had these third party entities that would go between them and the bank for the purpose of making the loans seem less risky and for each of these transactions, the legal documents would be thousands of pages long. Enron had thousands of these third party entities set up, so the sheer amount of documents that were publicly disclosed were quite literally in the hundreds of thousands. But they were public, and it is on record at Harvard, where students reviewed Enron as a class exercise that they were able to use very standard accounting review methods and were able to determine that Enron was in trouble. This was way before the collapse. So we cannot say that the cause of the Enron debacle was a lack of information. The information was there. In fact, when the wall street journal did a similar review, they had to question the executives as part of the audit. The executives flew out to answer all questions. The WSJ concluded that enron was overvalued. This was before the collapse.

Now back to Gladwell.

Gladwell, goes on to discuss the difference between a puzzle and a mystery. Wow, that sounds familiar! But his defining difference is that a puzzle can be solved by finding the missing information. If this information is missing due to human malevolence or deception, we blame the person who was holding or hiding this information and the “crime” is solved. This works very well for solving puzzles. Mysteries, are quite different. In a mystery, it is not a lack of information that is the source of the problem, and it can even be too much information that is the problem. Gladwell says that failure to distinguish which TYPE of problem you have is disastrous, for these two different problems require very different solutions. If you misdiagnose, you can make the problem much worse.

Solving a puzzle requires more voices, more sources; solving a mystery often requires less voices, better ears. This is a common mistake in software development management. A development project is running late. The manager decides to double the number of engineers working on the project. This slows the project down. As the lines of communication grow exponentially with each new developer, progress slows.

The genius of open source development is in how it solves this problem. There are thousands of developers working on any given open source project, so how does the project not get mired down by the sheer weight of the overhead? The pieces of the project are designed to be orthogonal to each other, meaning perpendicular, at right angles. In software development, this means that if part C fails, it does not affect part A. A bad driver developed by a third party software vendor does not blue screen the operating system, for example.

OK, back to finance. the lesson is the lesson of the mystery. Of recognizing when the puzzle model solution is not appropriate. Think of our current crisis. They are trying to solve the problem as if it is a puzzle. It is not MORE regulation that is needed. It is the CORECT regulation.

Gladwell also notes that in the case of Enron and others, there is a very important element of overconfidence – this is what the third party entities are designed for: to give the banks more confidence of their loans.

But in that meeting where the executives flew in to talk to the wall street journal auditors, the MIT mathematicians were explaining why their models were so brilliant, and while they were admiring their own genius at being the masters of these complex entities, the auditors asked an interesting question. Who do you think is going to be elected President? The geniuses said it did not matter, then started to go back to their formulas, but the auditors interrupted. (Remember, Enron, is about energy, oil). The auditors asked, “One candidate is an environmentalist and the other an oil man. Do you not think that it will make a difference which gets elected?” The executives, deferring to their masters, looked to the mathematicians. The mathematicians blinked a couple of times. Then swallowed.

These complex mathematical models can be destroyed by the simplest assumption gone wrong. In Enron’s case it was simply a mis-prediction in the price of oil ten years out.

And yet there is nothing magical about seeing this. The IRS does not use mark to market accounting. They tax you when the purchase is made, when the money actually comes in. Thus, Enron paid no taxes for years and years, because the fact was always the same: They were not making money. There was nothing to tax. The hundreds of millions of dollars on their accounting books were from the start, imaginary. The IRS knew this. The investors should have known. It was not a lack of information that sank the ship. It was too much confidence. We let ourselves be dazzled by mathematicians and wizards.

It was the same mistake we made with Long Term Capital management 10 years earlier and the same mistake AIG made ten years later.

It is common sense that has gone perverse. Proof of this can be seen in the way the media, the financial gurus, Obama, and most others have treated the simple, common sense solutions of Paul Volker. His voice is almost alone the voice of common sense and reason in this current crisis. His advice is simply to separate hadge fund activity from banking. Let the hedge funds gamble in a sandbox. If they fail, they fail. Relocate them to Vegas, and let us get on with the business of banking. We can watch the neon flicker on and off from afar. Volker’s point is simple. Banking has become what it is not. Banking is not supposed to be short selling and derivative swaps. This is what hedging is. Gambling, in short, on the future value of paper. Until we learn the difference, as long as the hedge fund gurus are the presidents of our banks, this problem will not end.

Gladwell’s genius is in being able to see through the chaf, to the heart of the problem. Is it a puzzle, or a mystery? Too much information or too little? Are we a bank, or are we a hedge fund? You have to understand, philosophically, what a thing is, first and foremost. These are philosophical questions: it is the question socrates asked, “what IS it?” What IS a puzzle? What IS a mystery? What IS a bank? Get these answers wrong and nothing else that follows will be correct.

Sorry, the comment form is closed at this time.