Governance of models

Governance of models and decision-makers’ behaviour in relation to these models.

Insurance companies have increasingly sophisticated in-house models which are, theoretically, at the core of the decision-making system. However, this only applies to two attitudes to risk as described by David Ingram and his co-authors. “Managers” representing the decision-makers that might follow the traditionally-accepted enterprise risk management strategy (ERM or GRE) need a sophisticated in-house model to guide risks for the entire company.  “Maximisers” prefer the so-called “risk trading” strategy that requires knowledge about whether a transaction is worthwhile in terms of price (compared to the added risk). Decision-makers’ attitudes in relation to models can be expected to depend heavily on their attitude to risk.  However, initial studies conducted by Clot, Ingram and Loisel (2014, currently being prepared) show that attitudes to the model appear to depend (more heavily than attitude to risk) on the role of the person interviewed as well as the circumstances at the time when the survey is carried out.  On average, actuaries seem, at first glance, to put too much confidence in models. Indeed, initial findings that were analysed from questionnaires showed that attitudes to risk and models are highly correlated.

Understanding how recently experienced events influence attitudes to risk and models and how these concerted attitudes can have a retrospective effect on the company’s strategy will be one of the future targets for research. Another direction of research concerns the study of collective decisions involving a combination of attitudes to risk and models with more than the average number of component parts. We can surmise that this is a sign of taking several points of view into consideration by each decision-maker to inform their choice, but this may need to be verified and understood how each of these points of view ‘join-up’ in an Executive Committee. Is it desirable to have an Executive Committee with uniform or varied attitudes to risks and models?  What levers for action do we have to result in improved risk-taking?  What effect do managers’ short and medium-term experiences and memories have on decision-making? Consideration is also being given to a link between experimental economics and neuroscience applied to risk-taking to tackle these questions using an inter-disciplinary approach.