Monday 9 March 2009

Duck, it's a Black Swan!

The Times Economic Review today (9th March 2009) eloquently described the pratfalls of the traditional stochastically based risk models so ineptly used by the banks to give balance sheet risk analysis. As Nassim Taleb regularly points out, these models have little use in predicting unusual or “Black Swan” events.

We’ve seen a flock of black swans cruise into view recently and asked “why didn’t we see the incoming fleet of Cygnus heading our way?”

In my humble opinion, the “machine” always did know there was a chance of a devastating implosion in the financial markets due to the burden of debt. I recall gasps of astonishment when a dynamic synthetic credit product using a form of leverage/deleverage to offer a yield of 2% above risk free and a ‘AAA’ rating. So, a ‘AAA’ rating (pretty much riskless) offering 2% above standard ‘AAA’ instruments such as RaboBank or EIB? Seriously…did nobody ask questions? Billions of Dollars of this product were sold despite the niggling worry.

For the vices known of traders such as hubris, arrogance, self-interest and greed, stupidity is not one of them. Any fool would, at least, wonder, what an earth was going on when two “riskless” instruments offered such wildly differing yields.

The bottom line was, if it could be rated, it could be sold. Ratings buckets guided many customers, almost entirely. Give it a ‘AAA’ rating and you could sell it anywhere it seemed. Here lies the lesson. Models may offer some form of comfort, whether for options, derivatives of debt default estimates, but without common sense can lead to catastrophic mis-allocations of capital.

Scientists can pour over statistics and provide ever more elaborate valuation models, as detailed in the Times article, but the same fate awaits the next generation of bankers as has befallen us. Here’s how;

Looking forward in time, perhaps to 2010 or 2011 and the financial system is propped up by the tax-payer and credit becomes easier to source. Rates will be low, probably 0% to .5% across the developed world economies. Models will be deployed in banks that value “shock” events, (black swans) highly and instruments that capture this risk trade at high premia.

Some smart cookie will realise that with rates at zero, one needs to generate little in the way of profit to garner a nice management fee and an annual performance fee by taking these highly priced elements of risk (perhaps engineered out of other products) and generating regular income from them. The selling of out-of-the money put options is a simple example. It may be quite easy to generate a return of 2% above base rates by selling these options or risk units.

As time goes by, more and more funds jump on the bandwagon and the supply of these risk units increases and the price begins to fall. Though the probability of the event happening that underlies the risk element is unknown, the longer it doesn’t happen the lower the premium becomes.

Many years of satisfied fund holders and highly paid hedge funds (or whatever they’ll be called in 2015 or so) will convince the world that we have, indeed entered a new period of growth and expansion and the toxic past is cleansed and in the history books “we’ve learnt from our mistakes of our predecessors”: President Palin, June 2015 (oh god no!)

The markets rise, implicit leverage climbs again, debt is re-accumulated as rates remain low. Once again the system becomes debt fuelled and ripe for another “pop”

The long and short of this is to agree with the authors of the Times article, that putting shocks and more historical data can help the analysis of what the banks hold today but as a forecasting tool as to future behaviour of investors is of little use. Behavioural biases will tend to take advantage of “perceived” over-priced assets, construct ways to exploit them until a form of equilibrium is reached once again.

The market, whether it for equities, bonds (‘AAA’ or not), real estate or volatility, is simply the consensus or “crowd” valuing the asset to its group value and risk tolerance. Modelling this is difficult, it changes every day. Its what makes the markets fun!