In this extract from The Geneva Association’s Insurance and Finance’s Newsletter, Mr Shigeyuki Goto of MS&AD Insurance Holdings lists three rules that insurers should follow when adapting their ERM systems and processes to the new regulatory landscape. These are: favouring scenario-based capital requirements; a decision making process that addresses fallacies, biases and cognitive traps and a collective effort for certain risks which are not insurable commercially.
Insurers face a daunting task of coping with new and enhanced regulatory requirements. The financial crisis in 2008-2009 and the so-called black-swan events of 2011 led to a major overhaul and/or refinement of regulatory requirements around the world. And now that there are new sets of requirements specific to systemically important insurers and internationally active insurers, we are seeing a further change in the rules of the game, particularly in the field of ERM.
New times call for new thinking. New generations in insurance firms should follow three rules when adapting their ERM systems and processes to such ever-evolving times.
Rule 1: It’s all relative
Insurers should favour scenario-based capital requirements. Such scenario-based assessments should be backed with stress tests, and also complemented with emerging risk identification and monitoring processes. For cases where setting out a factor based requirement is a must, the calibration at least should be based on scenarios.
Do away with purely stochastic-assumption based requirements that are not linked to scenarios (eg VaR assessments simply run on a spreadsheet). That is becoming a thing of the past. Instead of trying to guess in vain what our capital requirement is in absolute terms (ie what our 99.5% value at risk is), we should focus on understanding it in relative terms (ie look at how an outcome under one specific scenario compares to another in the overall context of things).
Having to understand multiple outcomes based on scenarios and stresses and make a decision on them leads to substantive discussions on whether/why one outcome is more meaningful than another (if at all). It is essentially a built-in thinking process.
As for models, they should be seen as mere tools to generate and run scenarios. They are perhaps advanced tools at that, but nothing more. The best models are marginalised ones.
Rule 2: What you think you know could hurt you
This then needs to be backed by a process that addresses fallacies, biases and cognitive traps in our decision making process.
All too often, insurers base their risk management on past experiences and fail to account for risks they cannot (or will not) perceive. As the moral of the story of the blind men and the elephant teaches us, what we (think we) know often is at best a very small part of a much bigger picture and at worst totally off the mark.
Secondly, individual heuristics (instinctive thinking based on rule of thumb) will always be inherent in decision making. Any firm’s risk management is only as effective as the least effective individual involved in running the process.
Recognising these two features, there needs to be a qualitative process that forces us to go beyond our perception(s) and/or rule of thumb thinking and consider what lies outside the bounds. In terms of scenario-assessments, stress tests, and emerging risks, there has to be a step that systematically asks us whether we have missed anything or whether we got something wrong.
Rule 3: Some risks are beyond individual firm’s ERM
Market-wide problems require market-wide solutions. There are just some things that are beyond the control of individual risk management and hence require wider, collective risk management.
Recognising insurance may not be commercially viable for certain risks, markets – led by individual firms with high awareness of ERM – must consider putting in place special systems such as pre-event CAT reserves, firm-specific counter-cyclical capital buffers, mandatory pooling schemes, state co-insurance (or government support phase-in schemes), or a combination thereof
.
Examples could include extreme risks with unknown return-periods and/or highly public lines of business. This rule applies regardless of how advanced the market is or not. Non-competition areas can exist in highly competitive markets. Sustainability of affordable insurance coverage is ultimately what an insurer’s ERM should achieve. It just sometimes requires a collective effort.
Creating a more resilient and accountable system
The underlying common feature of the three rules is the differentiation of risk and uncertainty.
Having a seemingly sophisticated risk measurement and management structure in place can lead people to believe they have taken the “uncertainty” factor out of risk assessment.
History has shown us that is not possible and nothing could be a more dangerous perception. In fact, it is better to have uncertainty constantly hanging over our risk assessment. Only then can we take account of and deal with uncertainty in ways mentioned above, thereby creating a more resilient and accountable system.
Mr Shigeyuki Goto is Head of Group ERM at MS&AD Insurance Holdings.