Black Swans

So last night on the train home from not so sunny Paddington I finally finished reading Black Swans. It's took me a month (originally purchased in Edingburgh on the JP Roadtrip) but in summary its been a worthwhile and... necessary read. Strange word to chose I know but it seems apt. Its one of those illuminating books that pulls back the veneer of conventional wisdom and shows you that some assumptions you make: that things are just the way they appear, are actualy constructs that have implicit implications.

So what's the book about?

It's another of these socio-economic, philosophical books like Freakanomics etc. The difference here being that I think it sets out to debunk a lot of the wisdom peddled in such books.

So what is a Black Swan?

A black swan is a random unforseen event. And the books basis is really critiquing anything and anyone who sets themselves up as an expert on pretty much anything! Nassim Taleb, the author in making such bold claims is as I'm sure you can gather, not backwards in coming forwards! What he really rallies against is mans desire to try and establish models around things that are essentially "unknowable" (risks). The chief culprit Taleb identifies as the Financial Services arena. Ironically Taleb himself comes from a financial services background.

To explain this in a bit more detail, Taleb argues that we suffer from a gross narrative fallacy. That is we look back at history and try and compose narratives (models, controls etc.) that make sense of the past. This whilst understandable isnt inherently dangerous - just deluded. They become dangerous when we take these tools (that are constructed based on a highly questionable view of the past) and project them forward, and then present this as proven, factual, and scientific - when it is absolutely not.

A model that is created based on historical data clearly can not be criticised or objectively evaluated on that same past historical data. However when these tools are focussed on the future they inevitably break. The alarming thing here is that most fund managers do operate on exactly this basis. Likewise Hedge Funds achieve their profitability on the basis that they "understand" the risks. There is no real acknowledgment of the real risks. And the real risks or black swans are those things that happen that are outside the norm - or more correctly outside the expectations of the founding historical view.

I'm starting to realise in trying to explain this why the book itself is quite so long and spralling. It's very concept seems to ebb and flow between a multitude of concepts (some tightly related some would appear light years apart).

So what should we do?

Taleb argues that we should view anyone who presents themselves as an expert with suspicion. This is good with me! :) Second he argues that both in life and financial matters we should be very mindful of the effect of black swans upon us. And as a result of this we should protect ourselves from their negative impact, whilst at the same time maximizing our exposure to positive black swans. I appreciate this may sounds contradictory! So how to explain this?

So an example of a negative black swan is 911. A financial black swan is the present credit crunch. Most financial companies will wash their hands for any culpability regards either of these situations because their "perfect" models could not foresee them. They present it as being outside the bounds of their responsibility. Yet in the real world there is no such thing as pure finance. Things evade strict definitions and boundaries. Stock prices etc. whilst they can be modeled to some extent are inherrently very limited in the range of factors they can incorporate.

Fortunately to balance this we have positive black swans. Examples of this are Google, Microsoft, The Internet etc. Companies that from nothing grow exponentially to create almost inconcievable wealth.

So following on from this what Taleb recomends from a investment perspective is to go ultra defensive on about 85% of your assets and on the other 15% go crazy and open yourself up to the most risk and therefore the most black swans possible!

Mediocrastan and Extremistan

So I've kind of run ahead to the conclusions and missed out a lot of the founding theory behind these conclusions. One of the big concepts that's important to understand is why do the models we have break down. Why do they fail us so completely? The answer to this Taleb puts forward is that these models live in the world of Mediocrastan - where things can differ but generally things lay comfortably within the limited range of the bell curve and therefore can be understood as such. An example of something that can live comfortably within Mediocrastan is height. The tallest person on the planet might be 8 feet and the shortest 1. But over thousands of people it is safe and fair to make forcasts and predictions on what someone's height may be. If you have to guess someone you havent mets height you cant be that wrong - a few feet here or there.

In Extremistan this does not hold true. And today most people will acknowledge we live in a world of extremes. Perhaps the grossest extreme we can talk about is personal wealth. If we compare the wealth of an individual in the UK to a person from Africa then we have a big differential that starts to stretch the plausability of any predictions made on the data. Then once we combine that with individuals like Bill Gates the data can get horifically distorted. Even across a large sampling of data such disparity makes using tools from Mediocrastan dangerous. You do not understand just how wrong you can be. Its not too much of a stretch to see the worlds financial system as belonging to Extremistan!

The book is choc full of anecdotes and ideas along these lines. And it makes good if dizzying reading. One particular chapter title I'll leave you with, that I particularly like, is"Information Is Bad for Knowledge". In the persuasive world of Mr Taleb he kind of has a point!