Progress report
An assessment of hedge fund YTD performance in the face of renewed fears for a potential eurozone crash Read More
Against the backdrop of difficult market conditions and growing investor…
20/07/2011
The death of the News of the World has wiped Sundays clean of the weird and wonderful, but recent events in the financial sector can be construed as far stranger than the celebrity intrigue of the red tops. As we have discovered from events like the demise of Long Term Capital Management, from authors like Nassim Taleb and the recent financial crisis, unexpected stuff happens all the time – especially if you pretend it can’t happen. In fact, to momentarily sink into tabloid-ese: you couldn’t make it up.
It goes like this: a clever trader finds a new source of alpha in a low-risk trade, judging by historical patterns. But as the strategy becomes successful, and the size of the trade increases, so
the dynamics surrounding the trade are altered. In other words, the trade is theoretically very low risk but in practice, when you place it, the risk increases.
Concepts like the normal curve and metrics like Value at Risk (VaR) are dependent on price history, and neither the trade nor the risk has been observed before, so these metrics are totally
incapable of providing a meaningful assessment of the trade’s risk.
The real risk is a function of changing market forces. Often, the magnitude of the positions relative to the size of the respective markets results in such a small free float that market supply and demand has an inordinate effect on the price. This is a common theme in large losses: the potential for adverse changes in liquidity has been grossly underestimated.
Of course, as a trader becomes iconic or a strategy becomes popular, the buying interest can be self-fulfilling, until there is no one left to buy, and nothing to support the price. Did anyone say ‘bubble’?
And once the large positions start a forced unwinding, market prices tend to accelerate the wrong way, leading to serial correlation and a vicious circle of deleveraging. Of course, the mere knowledge or suspicion of a large and unwieldy position can cause other market participants to take advantage of it. A lot of commodity trading is done through futures on the Chicago Board of Trade (CBOT), and they publish volume and net open interest in different markets.
Sometimes, the knowledge that the market is positioned one particular way is enough to drive a bet that at some point the positioning will be reversed, with consequent money flows influencing the price.
Clive Capital apparently reported a 8.9% loss when the oil price plunged recently. Ignoring the joy that the press derives from the sheer schadenfreude of it all, this is hugely unlucky for two reasons. Firstly, coming at the end of the month, it was somewhat difficult to sweep under the carpet as an intra-month blip. But secondly, Clive Capital said that the oil price drop was a “five standard deviation move” which was somewhat unexpected.
In reality, a 8.9% loss is not very big. Monthly losses of that sort of size have happened in far too many track records. Also, if you invest in a commodities fund, you have to expect losses. This
is one reason that hedge funds are rightly restricted to professional investors who diversify at the level of their portfolio, and don’t rely on each manager to
provide diversification.
The real loser is the credibility of risk management. Let’s hope that the author of Clive Capital’s comment about the ‘five standard deviation move’ was the marketing guy, not the risk management genius.
The normal curve and VaR do have their uses, but they are mostly for when things are working normally. Conversely, they are misleading and useless when you start getting into the tail ends of the distribution. Beware managers who use statistical buzzwords, but don’t actually understand their relevance.
Christopher Miller is CEO of Investment Quotient, a financial research consultancy
07/06/2012
Join us and our panel of experts for HFMWeek's Subscribers' Club June's UK breakfast briefing, 'Impact…
31/05/2012
The next US HFMWeek Subscribers' Club breakfast, will take place on Thursday May 31. Join us and…
02/02/2011
HFMWeek's European Hedge Fund Services Awards are designed to recognise companies that have outperformed...
Be the first to comment on this article!