Clark, who founded AIR Worldwide five years before Andrew struck in 1992 and now leads her own modelling firm, Karen Clark & Company, said losses would be substantially higher if such a storm made a direct hit on Miami. 

Speaking to The Insurer TV to mark the 30th anniversary of Andrew’s formation, Clark highlighted that the storm did not hit the most populated areas of Miami, having made its Category 5 landfall around 50 miles south of the city. 

Insured losses at the time totalled around $15bn but that number would be significantly higher at today’s values, while insured losses from a similar storm track and intensity would be elevated by the scale of development within the region over the past 30 years. 

“If the same storm were to happen today, losses would be on the order of $70bn. But if Andrew were to occur today around 50 miles north and make a direct hit on Miami, the insured losses will be well over $200bn.”

Clark had established AIR Worldwide – now Verisk Extreme Event Solutions – in 1987, and said the modelling firm had a “handful of clients” at the time of Andrew. 

“We had been able to convince some forward-looking underwriters that they should be using more scientific approaches,” she explains. 

“But even our own clients probably found it difficult to believe the numbers that were coming out of our hurricane model, because they were much higher than the industry expected.”

Top 10 costliest tropical cyclones to impact the US

Clark said it took the industry around six months to wake up to the fact Andrew was going to cost it around $15bn, with early estimates suggesting a bill closer to $6bn.

“Around six to nine months after Andrew, the industry realised they were woefully under-estimating the risk and that they needed this crazy new tool called the catastrophe model.”

Andrew remains a watershed moment in the development of catastrophe modelling and science and data-based underwriting. 

The storm followed close to two decades of relatively low hurricane activity in the 1970s and 1980s. 

“As we were having this lull in hurricane activity, people were moving to the coastline in droves. So there was an enormous, very fast-growing population along the coastline. 

“Property values were growing very quickly in hazardous areas and the industry was not tracking exposures in terms of what are the values of the properties they were insuring,” Clark explained.

“Most of the industry was using premiums to estimate the loss potential and to price catastrophe reinsurance, so they weren’t monitoring the growth in exposures in coastal areas.”

Watch the 10-minute interview in full to hear more from Karen Clark on:

  • How Hurricane Andrew shook up the way the industry manages catastrophe exposures
  • Why the (re)insurance industry was unprepared to deal with the storm
  • How catastrophe modelling tools have evolved over the last three decades
  • And the next steps being taken to improve these tools for the future