Following ten years of research, two principle contributory factors to our inability to model the future accurately.
The first of these is the manifestation of trends and developments which have not been observed before. The lack of data on such events pose a severe limitation on our ability to accurately model future outcomes. An example is Quantitative Easing, for which no precedent nor data exists. How do we model its eventual unravelling?
The second is our tendency to underestimate future correlation. We tend to model the past, allow for adjustments – even complex ones such as Generalized Autoregressive Conditional Heteroskedasticity (GARCH) – but it remains challenging.
Dynamic Risk Assessment applies innovative expert elicitation scientific principles to augment the paucity of past data by processing the findings applying graph theory, and using network analyses to interpret the findings and identify potential future systemic risks. It models the potential causeways of future contagion between discrete risks in a novel, insightful manner.