Embracing Systemic Risk in Insurance

Much has been made in the press and on LinkedIn of recent moves made by the insurance industry, especially as it pertains to Cyber, to manage systemic risk.?None more so than Lloyd’s – through the implementation of their Cyber War exclusion - but also Beazley, and their forthcoming developments in their approach to classifying and managing systemic events in their cyber insurance product – building upon the lead taken by Chubb earlier this year.

More broadly, systemic risk has come into further focus by the (re)insurance industry through losses encountered during the pandemic. More abstractly, climate change has also highlighted systemic risk through the consideration of its impact on existing modelled views and its influence on insights drawn from historical data which has, in the most part, held a static context.

Systemic risk is proving to pose challenges to the (re)insurance industry, clients, and the capital markets at large.

The embracing of systemic risk is not something that the insurance industry has, historically, naturally built its foundations on, or which is naturally familiar. However, what is clear is that the world has become - and will continue to be - more interconnected. Through travel, digital connectivity, and broad globalisation. Consequently, the notion of systemic risk and its associated challenges is not going to abate, and the (re)insurance market must meet it head on to retain relevance and value as an industry.

It is natural for those operating in the (re)insurance sector to seek certainty (thus limit uncertainty) to drive confidence in decisions. It is human nature. In most recent times, such certainty has sought through growing confidence in modelled outcomes directly describing and capturing the probability of loss. However, even on individual risk events that still poses its challenges.

Systemic risk however, by its very nature, is complex. With complexity comes significant uncertainty. Such conditions are compounded if the nature of such systemic risk continues to evolve (i.e., is non-stationary). The pace at which a risk environment alters makes traditional techniques relying on volume and tenure of data challenging, as such data lacks material context to drive insights with high degrees of confidence. Surprisingly (to the annoyance of many residing in the sales departments of data science firms over the last 10 years) not all problems can be solved by Machine Learning – where its value is highly dependent on strong context. This is most notably the case in Cyber, where there are significant shifts in the risk environment day-to-day, reframing the lens through which observed data should be evaluated.

Whilst there are significant challenges in definitively quantifying systemic risk and the pursuit of minimising uncertainty, that does not necessarily mean it is unmanageable; or that such conditions are insurmountable for the industry to overcome. However, what is required is a shift in mindset.

Overall, the (re)insurance industry needs to embrace systemic risk and become more comfortable with the fact that achieving certainty is a fallacy. However, a more valuable approach is to move towards the notion of risk pricing and capital management as a function of confidence – i.e. a measure of certainty. Those with an increased appetite for risk are not simply managing aggregate levels of capital, but also their own confidence levels relative to a range of outcomes. Where increased uncertainty exists – as is the case with systemic exposures – there will inevitably be reduced confidence in outcomes. Consequently, such limited confidence needs to be priced for and supported by a product and market structure that makes this sustainable. Conversely, there will be elements of systemic risk that are more acutely understood, and less variable. Where more confidence exists, more risk can be embraced.

We cannot eradicate complexity or uncertainty, but we can balance confidence levels against a range of potential outcomes. Such outcomes can be enumerated and stress-tested over time, and our experience of the pace of change and variability in systemic risk will, inevitably, become more acute – driving confidence for all market participants.

Such principles and concepts are familiar to those operating in the Cyber space (if subconscious) because it is all that has ever been known. Cyber as an insurance product line remains young (and with various challenges of its own). However, a great deal can be taken from those experienced in cyber risk management in the growing need to manage and operate in a world presenting increasing systemic risk.

The internet and digital communications are so complex you cannot hope to deterministically understand all outcomes. In fact, the underlying architecture of the internet is built on the basis that such knowledge is unreasonable (i.e., TCP/IP) – but such resiliency is embedded in its structure.

The same concept needs to be applied by the (re)insurance sector, amending its market and product structure as well as the capital and pricing decisions it pursues to create a foundation for sustainable future growth.

Systemic risk is not easy, presents different challenges and may be unnatural. However, addressing it is achievable. I remain confident the (re)insurance industry will respond. It must... it is here to stay.