There have been several times in the past where I’m explaining ‘XYZ strategy’ to someone (hopefully they asked me beforehand), and the concept of skewness comes up. A couple examples:
- Several (successful) strategies lose far more frequently than they win. It’s not always like playing the lottery…
- Sometimes ‘the sure thing’ trade, which has made money every day, suddenly blows up.
Thus loops in the concept of skewness – how big are losses relative to gains? On the lottery side, you’re almost certainly going to lose USD 1 on a game with a (highly improbable) gain of USD tens of millions. But other examples abound in financial markets:
- Long-only (just about) anything: this is a negatively skewed strategy. Most months/years you will have a gain, but some months will be TERRIBLE. Don’t think about the little correction we just had…think about 2008. It can take years to recoup the losses from long-only: for example, notice that the NASDAQ is still about 10% below its 2000 peak.
- Venture capital: this is a positively skewed strategy, in its most basic form. The VC fund manager selects (say) 10 companies at an early stage of development. Financials don’t really mean much at this stage – they could do anything. The hope is that, out of 10, there will be 1 big winner and maybe a few small winners. The others are expected to be written off. So, one gain outweighs the many.
- Volatility selling: this is a classic negatively skewed strategy. VERY negatively skewed, epitomising ‘picking up pennies in front of a steam roller’. After premium selling funds lost about 50-70% in 2008 (or went completely bust), several actually hit high water marks in the past couple years. So it’s a sustainable, if nerve-wracking, strategy. By the way, insurance products and market making are roughly the same as option premium writing, in terms of performance characteristics.
- Momentum trading: a classic positively skewed strategy. Frequently momentum is classed as ‘long volatility’, which it is…kinda. More long gamma…but anyway. This is a ‘pain trade’, in that most of the time you’re losing money as markets oscillate back and forth and you’re trading with the trend. Only occasionally do the big trends come; you can’t really forecast them, and you MUST be in the market when they come. Otherwise this is a losing strategy.
I leave you with the following track records, harvested from Altegris’s managed futures website. Interesting place to learn about volatility and momentum offerings.
Extra credit: those seeking more technical info on return skewness, and particularly how the time-variance of skewness is a function of strategy design, should look at this wonkish paper.