Exploring the efficacy of higher moments in portfolio optimisation

(The avocado talk)

Main points:

• Neither word in “portfolio optimization” is very good — a much better phrase is “trade selection”
• An experiment hints at the usefulness of skewness and kurtosis for equities
• The number of parameters explodes for skewness and kurtosis for large asset universes
• Factor models and shrinkage are two possible ways of modeling skewness and kurtosis
• There is a lot that we don’t know

annotated slides (pdf)

Presented 2013 December 9 at the London Quant Group.