This trend toward higher (median) net income persisted in all asset-size groups with the exception of the largest banks where there was no difference, said the firm.

It reported that among 470 banks with $1-to-$10 billion in assets, for instance, those with some insurance activity in 2008--about half the total--scored 23 percent higher in (median) net income, and the trend was more pronounced in smaller banks.

Andrew Singer, Bank Insurance Market Research Group managing director, said, "Overall median net income was off 36 percent in 2008 compared with 2007. Nonetheless, the data suggests that pursuing a diversification strategy--of which insurance brokerage is often a key part--may have again paid off for banks in 2008."

He noted that "at a time when banks' traditional income sources are under pressure, an insurance agency business can help smooth out earnings and act as a hedge against interest-rate volatility."

Overall, the median net income for 7,563 operating banks and savings banks was $686,000 in 2008, down from $1,071,000 in 2007. The median at 3,338 banks and savings banks that reported some insurance activity--44 percent of the total number of banks--was $1,160,000, the company said.

It found the largest discrepancy was in the smallest banks. Median net income at 5,243 banks with assets less than $250 million was $469,000 in 2008. (That is, the middle ranking bank in this asset-size group reported $469,000 in profits.) Among the 2,136 with some insurance activity, however, the median was $825,000--76 percent higher.

The median ratio of noninterest income to total bank revenues among all 7,563 banks was 14 percent. At banks with some insurance activity, this closely watched ratio was 17 percent. Both ratios were unchanged from 2007.

The Bank Insurance Market Research Group, online at www.singerpubs.com, provides market research and investment sales data to the bank and insurance industries. Data is based on in-depth surveys of depository and insurance entities augmented by analysis of government data.