Enzio's Clock

Background

1. What is VAR?

"VAR"is an initialization for "value at risk". In the words of Gillian Tett's lucid Fool's Gold, p. 38, VAR is"... how much money the bank could expect to lose, with a probability of 95%, on any given day."  Next to suggesting how much money a bank may lose, the flip-side of VAR is that it, therefore,  indicates how much capital a bank must set aside in order to prepare for these possible losses. VAR was developed by the boss of my professional  alma mater, Morgan Guaranty Trust: Dennis Weatherstone and his team of quants developed this in 1989. 

To this day, VAR is "the" way that funds of all shapes and sizes measure how much of their money is at risk.  The only problem is its assumptions: first, that all risks are normally distributed (which they cannot be in a chaotic world), and secondly,that all risks repeat themselves in a straight line fashion. We all know that whilst patterns of history repeat themselves, historical events do not. But the VAR assumption is that historical events repeat themselves. 

Crucially, VAR models are based on the data of the last 365 days.  So any data before that is omitted from the model. So what? Well, if any data exceeding 365 days is omitted, then any history before that is omitted, is left out of the thinking of those analyzing the model's results. 

According to the Wall Street Journal Asia of 14th - 16th August 2009, pl M1, "From the middle of next month, the market volatility resulting from last year's collapse of Lehman Brothers Holdings Inc. will stop showing up...  The result is that the models will give a green light to start taking chances again. If banks act on this, a range of markets could see a boost in both trading volumes and in volatility."  So what going to happen as of mid-September is that past historical data will "look" much more stable - but that's because the Lehman's turmoil got truncated from the series on which the model's projections are based. So with less perceived "risk", the props boys can put on more Value at Risk again, correct? 

Moving beyond the middle of next month, however, "Regulators around the world are planning to impose more rigorous requirements concerning the way banks use the models (i.e. VAR, etc.), but these aren't expected to be in place before the end of next year." Indeed, Basel's Bank for International Settlements wants, by the end of 2010, to "...introduce rules forcing banks to introduce stress-tested VAR models .  Britain's Financial Services Authority already is urging banks to review how they use their VAR models - and wants them to introduce separate stress tests.

If you believe that in the banking world, greed has overtaken fear, then my conclusion is that as of the middle of next month, market volatilities will rise, for two reasons:

• first, particularly the more junior VAR analysts blindly will remove any Lehman-related risks from their 364-day history from which they are measuring market volatilities, so the "past" will look very stable. This means that - based on the last 364 days' worth of data - props desks and the likes can assume more risk, and
• secondly, there will be a scramble for the door before regulators impose onerous rules: greedy people will want to make the easy money before the game gets tougher.

We are not offering you a straight-line projection of higher volatility. We are just basing our notion of more volatility on the fact that newcomers to the risk game will know little of what happened a year ago, and that they along with old-timers will want to make the money and run before the lights go out, i.e. before regulators put the clamps on next year.

How much this ties in with props desks deciding to emphasize short positions and thus cause a crash is a moot point at this stage, but one that I already have warned you of...

2. How to Make Money Off This Idea

1. Always consult your financial adviser first.
2. Go long the yen.  If it is true that carry trades flourish when there is market stability, this is when the yen weakens. That is because people borrow yen, sell them for dollars, and put on "low risk" trades". But then the opposite must be true, if the most rudimentary rules of logic still apply: when volatility - instability - increases, the yen strengthens because people unwind their  carry trade, sell the currency in which they did this, buy the yen back and re-pay the bank. So up goes the yen. 
   
3. Short the stock markets.  Do this if you believe my crash of '09 scenario.