Nematrian
/

Views on non-Normal markets [59]

Go to: Summary | Previous | Next   
Bullet points include: Fat tails are pretty common. Both for individual indices and for multiple (joint) distributions. Treat skew, kurtosis, co-skew, Cornish Fisher etc. with some care. They don’t necessarily give appropriate weight to the right observations. Better may be to curve-fit the distributional form directly. Fat tails can be explained in part by time-varying volatility. Both of individual return series and of joint distributions. But still some “unknown unknowns”, i.e. “black swans”, e.g. Taleb (2007). Stress-testing, implied volatility/correlation analysis, crowded trade analysis etc. Portfolio construction. Output from optimisers are typically very sensitive to input assumptions. Try to keep adjustments as simple and as intuitive as possible

NAVIGATION LINKS
Contents | Prev | Next | Library

Desktop view | Switch to Mobile