So far the hurricane season has not lived up to the numerous forecasts of above-average storm activity — but that doers not mean the predictions are wrong.

That is the view of David Smith, senior vice-president of technology development and consulting for catastrophe modelling firm Eqecat.

Mr. Smith was in Bermuda yesterday as Eqecat showed off its new Asia typhoon model and an updated US earthquake model to insurance industry clients at the Fairmont Hamilton Princess hotel.

The models can play out computerised catastrophe scenarios, helping catastrophe re/insurers to gauge their risk exposures. At yesterday's meeting, Mr. Smith said he'd been asked why the above-average storm season expectations were not being realised.

"At the moment, we're right on track for what happens in an average season," Mr. Smith said yesterday. "But we're still at the very beginning of the season. Typically, August 1 is the real beginning of the season. Remember that Katrina (2005) and Andrew (1992) both came at the end of August."

Eqecat is not in the business of weather forecasting, but its storm models do take into account historical data from hurricanes past. The three basic components of a model are based on considerations of hazard, vulnerability and potential financial losses, Mr. Smith said.

Eqecat claims its new Asia typhoon model is the first to offer a view of risk across the entire western Pacific basin, as typhoons can often impact multiple countries. The model factors in the direct effects of wind, storm surge, and typhoon rainfall-induced flooding, as well as local building practices, design, and building codes.

The company also demonstrated its 2010 US earthquake model, USQuakeTM, which implements the 2008 US Geological Survey earthquake model and also features the innovations of soil-based attenuation and three-dimensional vulnerability.

Catastrophe modellers traditionally give a range of likely insured losses in the days immediately after a catastrophe, a difficult job amid the chaotic aftermath of a devastating storm or earthquake, but an important one for insurers who are looking for guidance on their likely losses to help them prepare for the onrush of claims and to give guidance to investors as soon as possible.

Mr. Smith said Eqecat had a reasonably good track record of producing an accurate range of estimated losses within 72 hours of a storm. "We did pretty well with Charlie, Frances, Jean, Ivan and Rita and also with Ike and Gustav in 2008," he said.

"In fact, before Ike had even made landfall, our advice was insured losses of between $8 billion and $12 billion of onshore insured losses. The final estimate was about $12 billion.

"It's more difficult to assess the smaller events, because the uncertainties are so great at the lower wind speeds. Small differences can make a $200 million loss into a $600 million loss."

One hurricane that did Eqecat, along with many others in the industry, did not forecast accurately was Katrina, which brought "a lot of surprises", particularly the scale of the losses suffered in New Orleans.