Measuring diversity and systematic risk from derivative market data and mapping to real-world probability statements

"Never put all your eggs in one basket". Most investors are well aware of this advice and often prefer to invest in a blend of different stocks. This is generally considered as a prudent and risk reducing strategy because losses caused by some of the assets may be countered by gains caused by others. However, periods of high market stress are typically linked to high levels of co-movement, implying that the diversification benefit is evaporating when it is needed most. Having a notion about today's level of co-movement (which we will also call the degree of herd behavior) may give market participants the opportunity to take necessary cautionary actions.

In the last years, the promotors of this project have introduced various methodologies for measuring how strong stock prices will move together in the future, using available option data. In the current project, we will consider the question how this information can be used to actually manage the co-movement risk in stock markets. To be more precise, we introduce a new product, called a diversity swap, which can be used to directly trade co-movement. We also propose a new methodology for determining an optimal equity portfolio by optimizing the diversification benefits. We also investigate to what extent implied measures reveal information about the real observed dependence between stock prices.