A week after Rendez-Vous ended last year, Hurricane Ike ripped through the Gulf of Mexico, devastated Galveston, Texas, and even caused considerable inland damage. Immediately, (re)insurers noted that losses would be high … but they did not expect to spend more than half a year continually revising their estimates upward. Many carriers began to look back on the storm and question their catastrophe models, wondering how a storm of much magnitude (at least from a financial perspective) could be missed. While the models themselves do bear some responsibility, the enduring lesson is likely to be the role of the risk manager in using them. After all, people — not technology - make decisions about risk and capital.
When Hurricane Ike had finished cutting its path, few realized that the initial loss estimates would amount to little more than an opening offer of negotiation. Through the end of March 2009, the implications became increasingly severe. Many (re)insurers saw losses from Hurricanes Ike (and Gustav) relative to shareholders’ equity grow by more than 20 percent — more than 40 percent for a handful of carriers. Such a profound loss of capital was bound to result in questions, and catastrophe model accuracy became an early target.
As with any large, unexpected event, there was no single point of weakness. Modeling firm historical experience information, insurer-supplied exposure data and insufficient knowledge in regards to assumptions and possible scenarios all contributed to the impact on (re)insurer balance sheets. The catastrophe models themselves drew some criticism, as how they can be used in risk identification and evaluation — not to mention reinsurance decision-making — could be improved.
The primary gap in model usage at this time last year — and historically — involved customizing methodology to a (re)insurer’s portfolio: catastrophe modeling is not a one-size-fits-all affair. A deeper look at each carrier’s risks and financial targets would have helped them ascertain the reliability and applicability of exposure and claims data and develop risk-transfer solutions that accurately reflected carriers’ strategies.
If this seems like more than a minor adjustment to how risk-bearers use catastrophe models … well, it is. (Re)insurers need to change how they view and transfer risk, starting with whether the models on which they rely are adequate, especially for portfolios with high concentrations of risk. Deficiencies would lead to suboptimal reinsurance purchases, gaps in coverage and an increase in the likelihood of disproportionate catastrophe losses — as we witnessed with Hurricane Ike. Since each modeling vendor has its strengths, risk managers must be sure to diversify across several to narrow any exposures that would otherwise go unaddressed.
Weaknesses in catastrophe models may have contributed to the size of Hurricane Ike-related insured losses, but this is just part of a broader issue regarding how the models were applied. Only by including them in a broader risk management plan - with careful risk identification, the use of multiple models and scenario imagination — can model use be optimized. The upside is substantial: unexpected losses are magnified in market capitalization. So, protecting capital can be an effective way to grow shareholder value.
Guy Carpenter & Company, LLC provides this text for general information only. The information contained herein is based on sources we believe reliable, but we do not guarantee its accuracy, and it should be understood to be general insurance/reinsurance information only.
Statements concerning, tax, accounting, legal or regulatory matters should be understood to be general observations based solely on our experience as reinsurance brokers and risk consultants, and may not be relied upon as tax, accounting, legal or regulatory advice which we are not authorized to provide. All such matters should be reviewed with your own qualified advisors in these areas.