Inspired by the great book Red-Blooded Risk, Aaron Brown. How to manage risk to maximize success. Requiring the understanding that risk is two-sided, dangers and opportunities are one sided. Dangers and opportunities are often not quantifiable, we have limited ability to control them. Whatever techniques you use, analyze risks, dangers and opportunities, optimize risk, arrange things that make danger and opportunity a positive contribution.
Consequently, think different
- Duality - split the analysis into 2 parts: the known (for normal, common events you have plenty of data and methodologies for quantitative decision support) and the unexpected (for Black Swan events you need what if analysis and conclusions)
- Boundaries - take VaR with a specification of its property. For a 1 percent one day VaR, the property is that about 3 days in a year a portfolio will lose more than VaR over one day. Optimize the risk for the days in which markets are "normal". Analyze the 3 VaR break days separately (the analysis of the VaR break day patterns is essential). See VaR of the Jungle
- Optimization - the idea traces back to John Kelly and his criteria. With all pros and cons (see The Problem With Computing Expected Returns in Finance). Risk optimization only works inside the VaR limits. Risk optimization creates growth.
- Money - IMO, the money system is universal in the sense that it is programmable (as media for transaction and storage for values). This makes it not so easy to understand. But, you cannot optimize risk without money and to understand money you need to understand risk. Especially outside the VaR boundaries, say, with some special instantiations of the utility theory.
- Evolution - it is an important example of good risk taking. Not surprisingly genetic programming is a general way to optimize. It may help to make (investment) decisions that fit better to future challenges.
- Superposition - it is about true randomness. And it has an impact on understanding financial "optimization". Michael (physics Friday) started a post series about this with Billiards, Chaos and Portfolio Optimization
- Game Theory - an addition to the theory of randomness, the mathematical study of uncertainty caused by actions of others (this leads to the question whether actions are "rationals" - see the "prisoner's dilemma" paradox and the TIT for TAT strategy). Game theory should influence thinking about risk in a different way than evolution and quantum physics. However, I see an analogy to co-evolution.
UnRisk FACTORY does produce and aggregate information automatically to better understand duality, boundaries and optimization. UnRisk-Q is the system for doing computational finance atop this and other informations - in any mathematical paradigm. UnRisk-Q is bundled with the FACTORY, but you ca get it individually.
Picture from sehfelder
Picture from sehfelder
