KuU - Known, unknown and Unknowable risks

 

One way to stop the needless noise associated with measuring risk is to develop frameworks on how to look at the risk problem. All risk situations are not the same. A conceptual framework can help with analyzing problems and improving the precision of our thinking. Within risk management, a good framework was developed by Deibold, Doherty, and Herring in their under-appreciated book The Known, the Unknown, and the Unknowable in Financial Risk Management. The KuU framework focuses on three types of risk, the known which is measurable or countable, the unknown which is what many have called uncertainty, and the unknowable, what we are not able to imagine or a function of our ignorance. All are focused on our knowledge which is either based on a measurement problem and theory issue. In KuU framework, we an look at risk in three dimensions. 

K, the known, refers to the probability distribution for an asset. This is the classic definition of risk. The outcomes and the probabilities for a situation are known. Knowable situations are well understood. There is a model, and the model has broad agreement among users.

u, the unknown, refers to a situation where the probabilities cannot be affixed to a set of outcomes. This is what Frank Knight would call uncertainty. If there is an unknown situation, there may be competing models which only offers conjectures and not clarity on what is possible.

U, the unknowable, would be any situation where future events cannot be identified. The events and probabilities are not known. Under this situation, there is no underlying model that can address or be associated with a market situation. 

The KuU framework can be associated with risk, uncertainty, and ignorance. By looking at any situation through this lens, we can better frame possible solutions. Is the issue a measurement problem? Is it a situation where we cannot get a count or measure? Is this a situation of ignorance?

We can solve or reduce ignorance through deeper research of the risk problem. We can also work at better measurement of risk, so that we can place bounds on the downside. If we can control risk and improve measurement, we can focus on the real problem of uncertainty. 

Unsupervised learning - Clustering and dimensionality reduction

 

Unsupervised learning can be a useful tool for finding relationships or grouping that may exist with large data sets. This can be extremely useful for finding deeper relationships than what may exist from just looking at correlation relationships. It can also be useful at finding links between a large set of assets and exogenous factors. More finance work has been done using PCA as a way of generating simple dimensional reductions. It is an east way to eliminate a primary common factor across stocks. 

I have been using unsupervised learning to help better gain diversification within a portfolio of futures markets especially within the commodity space. 

What is the right amount of investment staff?

 

The article "The unusual thing about Citadel versus other multi-strategy hedge funds", provides some interesting insights on the composition of investment staff associated with the largest multinational-strategy funds. The investment staff vary between 39% and 56% as a percentage of the total staff. These percentages tell us something about the firm strategy. A higher percentage of investment staff suggests that there is a greater focus on strategy pods while a lower number suggest that there is a more singularity of thinking about investments. It also tells us something about how first think they can create an edge. The lower investment staff percentage suggests that there is the view that an edge can be created by technology and not just investment acumen. There is no right answer, but the cost of running a multi-state is much higher than what some may think. There is a clear need for strong infrastructure to support all the trading and research that is being conducted.

Marty Zweig's rules - still useful

There is the assumption that new ideas are best. Follow what is the new thinking because ideas of the past just are not that valuable. I do not subscribe to this thinking. We can learn a lot from the best thinker of the past and in the investment area, Marty Zweig was a great technical equity trader. He developed a list of 17 investment rules, and I can say that most of these are still useful. The first two are classic trend-following along with numbers 5 and 12. The behavioral issues are embedded in rules 9, 10, 13, and 17. Fundamentals are associated with rules 3, 4, and 16. Macro investing is embedded in rule 6 and 15. Risk management is associated with 7, 8, 11, and 14. 

There are good take-aways from the Zweig list.