Casualty insurance catastrophes had long been considered too difficult to accurately model, but now those threats are being quantified.
A new report from Lloyd’s – in conjunction with Arium, Willis Re, and others – examines how casualty risk can be modelled using historic data and stochastic tools, in much the same way property risk can be mapped.
Andrew Newman, president, global head of casualty and CEO of alternative strategies at Willis Re, said developments in casualty risk modelling are crucial to the industry right now.
“Companies are [no longer] going bust by betting against the Almighty – but they are becoming impaired as a result of liability risks.
“So what does this [liability modelling] do? It begins to herald an era in which the gap between the risk management framework the industry applies to ensure that it’s making the correct bets and holding enough capital for property business, is now being mirrored in the context of taking on liability claims.
“In my little world – the world of people who manage liability threats – it’s a very big deal.”
Newman said the “the C bit [casualty] in the P&C equation” was finally catching up to its counterpart, property, in terms of modelling risk.
Property risk modelling went through its “renaissance” about three decades ago, Newman said, and now it was casualty’s turn.
“It’s significant. The P bit of the P&C industry has had an ever-improving, more accurate, more confidence-inducing approach to risk as relates to models, and casualty has been its somewhat distant sibling.
“What we’re seeing now is a big move to accelerate towards more objectivity, more precision. And clients want to know that, clients want to know how big their bets are,” he said.
Casualty modelling took longer to come to fruition than property due to the types of risks it covers – threats with less available data and more unknowables.
Want the latest insurance industry news first? Sign up for our completely free newsletter service now.
“Now the problem with casualty business – or liability business – is that the threats are very different from the threats that the industry models for property,” Newman said.
Insurers can map property catastrophes more easily because volcanologists and meteorologists, for example, have hundreds of years of data for fault line and storm activity.
“There’s a consistency in [those] threats which creates fairly predictable patterns of losses over extended time horizons,” Newman said, of property modelling.
“[But] if you contextualize that for liability business, the threats are consistently changing. There are economic forces – such as rigging Libor, the scandal that hit the insurers of banks; asbestos; pollution – those threats are consistently evolving and consistently changing and the forces that drive them are economic, judicial, legislative, societal, and human behavior.
“When you look at all those things, they’re very hard to model, so the industry generally hasn’t modelled them … because it was kind of assumed that it was too hard to do.”
That mentality is changing, though. There are several models in the market now.
“It’s not too hard, and we actually have to do it,” Newman said.
“The insurance industry, for a long period of time, has been unable to work out the magnitude of its bets on casualty business with any precision – which is why things like asbestos hit it hard and pollution hit it hard.
“A lot of regulatory capital regimes want to know what the output of a cat model is for individual companies in order to determine whether or not enough capital is being held to finance the risk. That is what regulators are primarily concerned with – to make sure that in extreme events, individual insurance companies and the industry at large have sufficient capital to pay the debts.”
Casualty risk modelling can’t know the future, but it can help clients diversify their portfolios enough survive the next potential liability catastrophe.
“We don’t pretend to tell clients what the next problem will be [in a certain industry] … what we do want to do is alert clients to where their big accumulations of risk are,” he added. “It’s not whether or not this year there’s going to be a loss in [for example] Florida but, if there was, what it would do to their portfolio?”
