Nearly all reinsurers (90 percent) said they apply surcharges to compensate for data quality deficiencies. Among these, 70 percent said they would include a 20-to-25 percent premium penalty.

What's more, 92 percent of reinsurers said confidence that the cedant did use strong controls when collecting data could mean premium credits between 5 percent and 15 percent. More than one-third said they would be willing to offer a minimum 10 percent premium credit for cedants with high-quality data.

This could mean big savings for insurance buyers. For example, a large insurer purchasing $10 million of reinsurance could receive $1 million premium credit, or more capacity, she said.

One reason reinsurers apply surcharges rather than going back to fill in the data is that “reinsurers renew a lot of contracts at the same time,” during Jan. 1 renewals, for example. “They have a lot to do at once, and may not go through the smaller accounts as thoroughly,” Ms. Conway said.

She added that supplying better data would benefit the primary insurer's underwriting process as well. She said critical information, such as location and number of buildings, could be better gathered with specific data fields.

The study found that more than half of reinsurers use sophisticated tools to evaluate the accuracy of exposure data received in broker submissions, while 83 percent have basic checks in place confirming that the most critical data fields are populated.

When asked which data items they consider most problematic, reinsurers pointed to insured values, complete inventory of locations and secondary characteristics.

How would reinsurers be convinced that data is, in fact, high quality? With the input of a third party, such as a catastrophe modeling company, or actuary–with the insurance broker being the last choice, she noted.

When analyzing the potential risk of a cedant with poor data quality, a majority (58 percent) of reinsurers said they directly modify their cat model results, while the remaining 42 percent of respondents make upward adjustments to the data before running their models, the study found.

To help boost their confidence level in data quality, the vast majority of reinsurers (92 percent) agreed that if the cedant used strong collection, enhancement and data maintenance controls, the risk would be more attractive to them.

While it's an important consideration, cat modeling is somewhat limited in that much recorded data for extreme events, such as hurricanes, only goes back about 100 years.

Ms. Conway said in a statement that investing the time to improve their catastrophe data quality is a “win-win situation for insurers. It helps them improve their own operations, and also makes their business more attractive to reinsurers, potentially leading to better pricing and more capacity.”

She said those reinsurers applying similar efforts in assessing the data they are given by insurance companies “will find success in the increasingly high-stakes property-catastrophe market.”

E&Y said the survey was conducted in early 2008 among leading reinsurance chief underwriting officers, heads of property underwriting and heads of cat modeling. Participating reinsurers included both domestic and off-shore, pure-play catastrophe and diversified companies that cover both commercial and property exposures.