Cat bond performance ahead of expectations, as losses continue to lag: Lane Financial

In a fresh analysis of the catastrophe bond market, consultancy Lane Financial LLC takes a look at loss expectations and concludes once again that actual cat bond losses are lower than the amount of principal losses risk modelling suggests should have occurred over the market’s history, as a result of which the market has outperformed.

Previously, Lane Financial looked at insurance-linked securities (ILS) (catastrophe bond) market actual losses based on indicative pricing, versus modelled expected losses and global industry views, and found that the actual may underplay the expected, leading them to conclude the market remains inefficient.

A new paper from the company looks at similar issues within the market, analysing the model output expectations for losses to the catastrophe bond market over a 23 year period and then comparing them to the actual loss experience of the market.

The analysis also looks at the differences between the short and long-term views of risk that hurricane cat bonds come with, so the base modelled expected loss and the sensitivity case, typically the warm sea surface temperature (WSST) version of the same model output.

While Lane Financial note that the WSST version of the risk metrics for a catastrophe bond shows the market incorporating a view of climate risk into its modelling for ILS deals, it seems that both the base and sensitivity case expected losses for the cat bond market are higher than the actual losses experienced.

One conclusion Lane Financial makes from its latest work is that, “the ILS market has retreated somewhat on its measurement of the effect of warm sea surface temperatures in the North Atlantic.”

Over time, the base and sensitivity case view of the expected severity of storms has moved in tandem, the researchers found, stating, “The two lines in the graph lie on top of one another. Their correlation is 99.3%. In other words, the modelers still believe that the sea surface temperature could lead to higher loss – but not because the severity of storms is affected by “Climate Change”.”

Adding, “The implication is that the dominant “Climate Change” effect is on increased frequency of loss-causing storms. But the modelers’ assessment of that has changed over time, in surprising fashion. One might expect given the recently announced record busting sea surface temperatures of 2023 that it would be rising. The numbers tell a different story.”

In tandem perhaps, to a degree, Lane Financial also shows through its new research that, in catastrophe bonds, “actual market losses continue to lag expectations of loss.”

“This reinforces the notion that model expectations are too conservative rather than too optimistic. Optimistic or wrong, being the cry that is heard whenever one year’s losses exceed that year’s expectations,” they add.

What’s interesting is that expectations of loss run above the actual market experience for both the base and sensitivity case expected loss metrics, again leading Lane Financial to suggest modellers could have been too conservative.

On a standard or base modelled expected loss basis the catastrophe bond market might have been expected to take some $7.7 billion in losses over its history.

On a sensitivity (WSST) case expected loss basis, that figure of expected losses for the cat bond market rises to $8.346 billion, Lane Financial’s analysis suggests.

But, the actual, so realised and also mark-to-market, loss of the catastrophe bond market is pegged at $4.958 billion, so well below the figures suggested by the risk models.

Lane Financial notes that its methodology could be wrong, the market could be wrong, the risk models are too conservative, or some combination of all three, plus investors may just have been lucky.

The lack of expected losses is reflected in the performance of the natural catastrophe bond market.

Lane Financial explains, “In terms of the long-term performance of the ILS market, it is an average annual 90bps better than might have been expected given pricing and expectations at the issue of each bond.”

Adding, “Most of that quantitatively comes from underwriting NA Wind risk but on a per unit of risk basis an almost equal amount comes from NA Quake risk, given its experience. Other risk categories underwritten provide diversification benefits but make smaller contributions to the extra 90bps of average annual profit. Their pricing is tight.”

Finally, Lane Financial also looks at the components of the very high catastrophe bond market returns of 2023, breaking out elements related to value recovery after hurricane Ian and also a compression of multiples that was witnessed.

Commenting, “While actual losses in a single year may have dropped from one year to another, 2023 is the first year we can recall where actual cumulative losses dropped from the previous year’s cumulative total. That drop, and its effects through the secondary market of market hardening at the start of the year, followed by multiple compression during the year, accounts for the very high total returns experienced in 2023.”

As ever, Morton Lane and Roger Beckwith of Lane Financial LLC raise some very interesting points regarding the market’s use of risk models and how risk within cat bonds is measured and expressed, as well as how that links to loss experience and returns.

It’s a detailed paper with some unique thinking as to how to analyse the catastrophe bond market, so we recommend you read it in full as our summary can never do it justice.

You can download a copy of the paper here and see the rest of Lane Financial’s excellent work over on its website.