Severe tornado activity throughout the US this year is running twice the average, racking up losses estimated at up to $2 billion, prompting industry insiders to speculate how much Bermuda's reinsurers will be on the hook for.

Natural catastrophe modeller Eqecat reports that tornado activity for the year is running well above average with 272 tornadoes versus the seven-year (2005-2011) average of 123. More than 150 tornadoes touched down in two distinct systems between February 28 and March 3, with the majority occurring in Tennessee, Kentucky, Indiana, Illinois, and Alabama.

Eqecat estimates that the latest series of severe tornadoes and storms has produced, so far, $1 billion to $2 billion in insured losses.

The fast and early start to the tornado season is reminiscent of last year, which, according to independent reinsurance brokerage firm, Holborn Corp, produced an estimate of more than $10 billion in insured losses for the affected areas of Alabama, Joplin and Springfield.

In 2011, the Association of Bermuda Insurers & Reinsurers (ABIR) calculated that international reinsurers paid 24 percent, $2.5 billion, of the total US 2011 tornado losses, with Bermuda's reinsurers paying out around 40 percent of that international portion.

If this year continues on the same track as 2011, Bermuda's reinsurers could expect a similar bill, says ABIR.

“For some of the largest tornado breakouts last year in the US we determined from available data that reinsurers probably picked up about 24 percent of the (international) losses,” said Brad Kading, president of ABIR. “So for the bigger US events in 2012 the same might be surmised.”

Mr Kading added that a general rule of thumb is that for loss events under $5 billion, most losses stay with the ceding insurers.

The 107 tornadoes on March 2, 2010, mark one of the most active tornado days recorded, Eqecat product architect Tom Larsen said in an interview with

The Royal Gazette. But it is still below the 200 tornadoes recorded on April 27, 2011.

Comparing 2004/2005 to 201½012, he added, where 2004 was a set-up year with significant losses and 2005 was the 'big whammy'; does that mean 2011 and 2012 will be a repeat performance?

“There's no statistical proof,” said Mr Larsen. “From a historical perspective, the activity in March 2012 looks very unusual, as did the very active April of 2011. Two unusual years in a row does not define a pattern … The combination of the randomness in the location of the touchdowns with the underlying increase in populated areas lead us to expect an overall trend of more tornado losses masked by considerable volatility in individual year results.”