Event risk comes in a number of forms, with perhaps some of the more difficult risk management challenges being posed when market participants split into two divergent camps associated with strikingly different views of the world. That is, there are two conflicting scenarios for how the future may develop, and both have meaningful probabilities. In such cases, the risk probability distribution best describing event risk may have two modes or be highly skewed and not symmetric – definitely nothing like a typical bell-shaped curve.
Our task is to develop a systematic and quantitative approach from observed market activity which allows us to imagine hypothetical risk probability distribution that is far from normal. One might want to adopt strikingly different approaches to financial risk management if faced with a two-humped distribution instead of a bell-shaped curve, even if the expected volatility was the same.
This research report takes us through our journey to develop the Market Sentiment Meter, which allows for a quantitative examination of how market risk expectations may evolve as economic environments shift from complacency, to more balanced risks, to anxiousness, to conflicted event risk scenarios.
First, we set the stage by describing our philosophy of financial risk analysis. Volatility is not risk. Starting points matter. Event risk has special characteristics. Then, we turn to our quantitative method of imagining an unobservable risk probability distribution using a carefully selected set of metrics and an innovative distribution-independent process that is fully capable of handling a simple normal distribution or a highly complex mixture distribution, which may have more than one mode or be highly asymmetric.
Finally, we present several case studies to illustrate how our risk probability distributions evolved during some well-known historical volatility episodes. We examine the drought of 2012 and how it impacted the corn market. We look at how equities responded in late 2017 and early 2018 to the large US corporate tax cut. We also study the US-China trade tensions in the spring of 2018, and close with some observations on the evolution of risk distributions during the pandemic of the spring of 2020. The cases give a flavor for how the Market Sentiment Meter can provide useful insights into the behavior of markets during both calm and stressful periods.
While the Market Sentiment Meter is not a predictive tool, we hope that the methodology and analysis of sentiment states and financial risk probability distributions can make a valuable contribution to approaches to risk management in stressful times.