
By Henry Gale
June 26 - (The Insurer) - This year's Los Angeles wildfires highlighted the need for ILS investment managers to use their own view of risk to assess catastrophe bond opportunities alongside the outputs of vendor models.
There are "very good vendor models", Mark Gibson, senior investment director at Schroders Capital, told The Insurer. "But if you're a good manager, you also have to be able to validate those models. And if you can't validate them, you have to create your own models."
He continued: "Even for the vendor models where they're very credible, i.e. they can be validated, such as for U.S. hurricane where we think the models are very good, there are still going to be reasons why you might make adjustments to those very good models to get an in-house view."
One limitation of catastrophe model outputs is that they rely on data about insured exposure at a certain point in time, which is then subject to change later. Climate change is also affecting the frequency and severity of some natural hazards, meaning models that have not been recently updated may go out of date.
In a recent article, Schroders Capital said that the Los Angeles wildfires in January were just the latest in a series of recent secondary peril losses with a return period greater than 100 years.
"Based on our model validation process, this trend cannot be explained solely by the changing environment since the most recent release of the respective catastrophe model," it said.
Schroders Capital was "materially underweight compared to the market" on exposure to the Los Angeles wildfires, Gibson said, "because we didn't invest in transactions because our view of risk was very different from that which was being presented".
If an investment manager's own analysis suggests that a cat bond's expected loss or attachment or exhaust probability is significantly higher than what is specified in the documents it is brought to market with, it is likely to see it as a less profitable investment.
"If we can't validate an external view, but we do have our own view, our own view may be that the riskiness of a transaction is multiples of what's been presented to us," Gibson said.
Without specifying the model or the peril, Schroders Capital said in its recent article that one vendor model did not pass its internal validation. After it created an in-house model instead, it found that an event the vendor model saw as a 1-in-100-year event was a 1-in-50-year event under its own view of risk.
"Given that most cat bonds have remote attachment levels, there are few opportunities to demonstrate an outperformance that resulted from the superior risk selection afforded by an own view of risk. The California wildfires afforded us that opportunity," the article said.
Etienne Schwartz, chief investment officer for liquid strategies at Twelve Securis, also highlighted the limitations of wildfire models when speaking to The Insurer.
"We will be happy to take wildfire risk if the pricing is right, but what we see is that the modelling sometimes is just quite weak in these areas," he said.
In January, before it completed its merger with Securis Investment Partners, Twelve Capital said that it had excluded from its portfolios one of the cat bonds with occurrence structures affected by the Los Angeles wildfires, while it was "underweight with the other relative to its market weight".
The potential investment advantages from developing an in-house view of catastrophe risk come with a cost, bringing an advantage to larger managers that can spread this expense over a larger portfolio.
Speaking to The Insurer TV shortly after the Twelve Securis merger was announced, Twelve Capital founding partner Urs Ramseier said the cost of doing business is going up for ILS due to the sophisticated tools required for pricing and analytics. This means the advantages of scale are rising, he said.
“These tools are expensive and have taken a long time to develop, and the bigger we are, the more we can spread the costs among more investors and bigger portfolios.”