Since the enactment of Gramm-Leach-Bliley, much has been written on the subject of banking companies' interest or lack of interest in acquiring insurers - life insurers, that is. For most banks, the thought of acquiring a property/casualty company appears to be a nonstarter.
Few observers have found compelling reasons for banking companies to lead life insurers to the altar. In fact, most have focused on the now-familiar barriers to such unions. The litany of deal breakers includes insurers' relatively lower returns on equity and on assets; their complex, labor-intensive, and costly back offices; their inefficient, high-cost, and independent distribution channels; their less creative use of technology to enhance customer value and streamline internal processes; and the fact that their profits are increasingly pinched by rampant product commoditization. 
These operational challenges are reasonably well understood after years of industry soul searching, and admittedly they will reduce the attractiveness of insurers to banks. In the short run, as the life insurance industry continues to address these shortcomings, potential acquirers will have to assess the long-term strategic value of a life-company deal within the existing operational context.
Some near-term improvements are certainly possible, even likely. But given the general lack of functional synergies between banks and insurers, do not expect bank-insurer marriages to produce any dazzling near-term operational improvements within the acquired insurer, such as fundamental changes in business models or the use of technology.
What about the numbers, though? Do operating returns, earnings growth, and shareholder returns really suggest that most deals will be bad?
First, let's look at operating returns. It is true that recent returns on equity at the top life insurance companies, which are at the high point of the last 20 years, are about 14% or 15%. Recent bank ROEs, also now at a cyclical peak, are routinely 18% to 20% or more. Thus without fundamental structural change, little doubt exists that a banking company's acquisition of a typical insurer will dilute its ROE, even without a significant pricing premium.
But how important is that to the stock market's valuation of the combined entity? Numerous studies show a link between ROE and price-to-book multiples, so an acquirer should, at a minimum, expect downward market pressure as one result of an insurance deal. The combination's long-term strategic value would need to overcome this pressure. But the markets have never been good at balancing long-term value against short-term performance, so at least moderate price discounting should be expected after a bank-insurer deal.
Growth in earnings per share is a different matter, however, as are the shareholder returns achieved by insurers and banks. Our recent analysis of the top domestic insurers and banks (excluding the global commercial banking companies) shows that in the past five years insurers' growth in earnings per share equaled, or even slightly exceeded, banks'. Similarly, S&P stock pricing data show that the insurance composite stock index has outperformed the bank composite stock index for both three- and 10-year periods through Dec. 31, 1999. (See table.)
In today's market, high ROEs are no guarantee of strong stock market valuations, as banking companies are finding out to their chagrin. This reality, along with financial measures like earnings-per-share growth and shareholder returns, suggests that buying an insurance company might not be such a bad idea after all.
So despite the conventional wisdom, it looks as though bankers will not be able to dismiss the notion of buying an insurer quite so readily. Instead, they will have to decide based on the relative merits of the strategic business options available, just as they would in considering any other type of deal. All in all, not a bad way to begin a marriage.
Mr.Stein is the national director of financial services at Ernst&Young LLP in New York.
