Monday, March 05, 2007

Crashes, randomness and coin tosses

by Veryan Allen

Hedge Fund

5.3.07

Crash or correction? The early signs were in place for a global drawdown. Smart money has been selling to dumb money for a while. With equity volatility and credit spreads at their lows but financial arrogance and market myopia at highs, a steep correction was only a matter of time. With Sam Zell cashing out, Warren Buffett searching for a successor and a few bottom tier hedge funds saying they couldn't find any shorting opportunities (sic!), the canaries in the coalmine could not have been singing any louder.

It is custom on the 9th day of the Lunar New Year (27 Feb in 2007) to make an offering from the previous year's prosperity to appease the gods, and clearly such profit-taking in China was overdue. However, it is a stretch to blame Chinese day traders for recent market turbulence. Clearly the eclipse over the weekend wasn't the only syzygy influencing investor behavior. Business magazines conveniently lined up in a classic long biased convergence to signal the end of the euphoria; when Forbes implies the bull market might just be getting started, Fortune runs a glowing advertorial for private equity and Businessweek says we are in a low, low yield world, you know it could get real bad, real soon.

Many "reasons" have been offered for the "correction". Ben Bernanke said there was no single trigger but that is ALWAYS the case anyway. Yes sub-prime mortgages are a disaster but that is not "new" news. The possibility of recession? What else could Alan Greenspan have said? He had three choices, 1) "Don't know, don't care" which wasn't really an option for him in his position 2) "No way, there is never going to be a recession ever again" (wrong obviously) or 3) Sure there is a chance. Choice 3 was the only realistic answer, but he gets unjustifiably blamed for moving the market. Then there was the China sell-off "catalyst" which had more to do with a simple desire to book some profit after a huge rise, and what better day to do it after touching 3,000 and then paying an 8.8% tithe to the Emperor of Heaven? The markets went down because there were simply more sellers than buyers.

Some say it was just a "fluctuation" in the random walk. Really? Why did the drunken man suddenly take such a big step backwards? A corollary to random walk assumptions is that there can be no such thing as investment skill! It is amazing how this nonsense persists in the face of such overwhelming counter evidence. The Dow makes a 3.7 sigma move which randomly "should" occur every 18 years or so. Since index inception there have been 130 worse sell-offs than last Tuesday, or more than 1 per year on average. Since we hadn't had such a move in ages and volatility tends to cluster, perhaps the year of the Golden Boar will see several more. Chinese stock indices made a 6 sigma move which, according to the random walkers "should" happen every 2 million years, so we can now safely pile back into China, right? Wrong. Meanwhile, the VIX and other implied volatilities made moves that "shouldn't" have occurred since the earth was formed. Sorry, but there was nothing random about last week. Buying the VIX on the extremely rare occasions it drops to single digits always pays off, incidentally.

Then there is this "fundamentals are still great" contention. Don't people realise that volatility CHANGES the fundamentals? George Soros has never received due credit for his reflexivity theories. The reminder that long only is risky will change investor and business behaviour. Formerly imminent private equity deals may not now emerge, with leverage funding possibly harder to get and more expensive. Roach motel illiquid assets (you can check in but can't check out) are going to be evaluated much more closely. Some say the "Bernanke Put" or "Private Equity Put" are floors on any sustained market drop but these are myths. Some who would have qualified for a mortgage in 2006 will not in 2007, which will impact real estate. Emerging market fundamentals and capital flows may change because newly risk averse investors may now elect to bring their money home. Funds inexperienced enough to be heavily playing the yen carry trade are going to be quiet for a while.

Truth is the first casualty of war, and liquidity is the first casualty of market volatility. The biggest casuality however is rising asset correlations. Credit and equity both dropped last week, almost everywhere. The popular fear guages of equity implied volatility and credit default protection blew out massively. Yet another confirmation that in the flat world of global capital markets, it is STRATEGY diversification more than ASSET diversification that is most likely to protect portfolios and make money when most risk assets drop in price. Also worth noting is that the stock markets impacted the most were the one's that make onshore short sales complicated or illegal. Every market needs such natural BUYERS during sharp corrections and taking profits is more pleasurable than cutting losses. Short sellers REDUCE the severity of market drawdowns through buy to cover purchases.

Alpha capture is really an alpha redistribution game. I have no idea which market offers the most beta, but the USA has the most alpha available. I know it is trendy for pundits to talk about the US markets being too "efficient", too "analyzed", too much "smart money" for anomalies to remain, but that is just plain wrong. Even Dow Jones and NYSE couldn't add up and divide the prices of 30 major stocks. The best money making opportunities are in the US markets because it is so inefficient, so liquid and has the the widest range of tradeable securities, derivatives and options. I am as globally diversified as much as anyone, but the best source of alpha will continue to be the USA, if only because it has the biggest and deepest markets. You have to be sceptical of funds moving into new areas because they are not making money at home. If a hedge fund can't make money in its "own" country how can it possibly do so elsewhere? Greed and fear and investor emotion drive all markets, therefore the biggest market MUST contain the MOST alpha.

There may be some "smart" investors around but there is far more "dumb" money playing the markets. There is not much connection between intelligence and financial savvy. Whether it was Isaac Newton getting blown away by the markets in 1721, or LTCM in 1998, or the "can't find any short sales" bozos that got "fluctuated" last week, there is plenty of money for smart investors to extract from other market participants in bull AND bear markets. It does not particularly matter what the reasons are for a move, what matters is that alpha was made and risks were ANTICIPATED.

Markets are not random just like coin tosses are NOT random. It would not be particularly difficult, theoretically, to construct a hedge fund strategy around coin flips. Coins "seem" random because most of us only witness single or very low sample sizes. If you toss many coins using the same mechanism and starting conditions the bias is 51/49 for heads which, while sounding slim, is an exploitable edge. If the coin is spun instead, the bias is hugely in favor of heads, 70/30 according to empirical work by Persis Diaconis. Look at any coin; is the centre of gravity EXACTLY half-way or isn't there a tiny bit more metal, usually on the head side? Going back to China; if you got every Chinese citizen to toss a coin each day, had knowledge of the starting face, which coin and modelled a few thousand sample flips from each flipper, the non-random bias would be obvious. Bet $1 on each toss you would have an over $1.3 billion hedge fund generating a high annual return from a supposedly random process called coin tossing. It is possible, but difficult, to develop predictive edges and probabilities in ANY process involving human bias and behavior. NOTHING is random!

I am neither competent to offer a diagnosis for last week's correction nor forecast economic growth. It is however noteworthy how the smart money has been selling out while the dumb money has been buying. I generated some positive alpha from the sell-off not because of some amazing insight but primarily because I have stress tested for anything, was sufficiently diversified and am ALWAYS long of options. Vega, vomma, volga and vanna are often ignored but were major factors in the markets last week.

If you have prepared for the chance of the sun not coming up tomorrow, -3.3% drops in the Dow don't cause a sweat. All I "know" is that there is bias in EVERYTHING, that neither prices nor volatilities are stochastic and that the risk-reward equation had swung over to the negative outlook/short biased side well before last week. It will be interesting to see how damaging this volatility cluster gets. Maybe it will soon blow over, maybe it won't.

No comments:

Lunch is for wimps

Lunch is for wimps
It's not a question of enough, pal. It's a zero sum game, somebody wins, somebody loses. Money itself isn't lost or made, it's simply transferred from one perception to another.