TUSHAR MATHUR HAD

every reason to expect Bank of America's best deal on a mortgage. After all, the 31-year-old software developer kept a checking account and a credit card with the giant consumer bank. He had a ridiculously high credit score 785 and was willing to make a 20% down payment on a brand-new four-bedroom colonial in the Atlanta suburbs.

But a funny thing happened on the way to closing the deal on his first home. After reviewing his application, he says, a Bank of America representative offered him a rate of 6.5%, nowhere near as good as the 5.8% to 6.2% rates he had been quoted from other lenders. And the difference wasn't lunch money: Had he gone with Bank of America, his mortgage payments, over the life of the loan, would have cost an extra $21,000 more than the deal he ultimately accepted. "They were higher than anybody, and I don't know why," he says of his branch's offer.

For its part, Bank of America declines to discuss its offer to Mathur or the details on how it makes such decisions. But ask a financial-industry consultant and he'll tell you what may have been going on: something called price optimization. That's a fancy way of saying Mathur's rate was partially based on what the bank's computer thought he might be willing to pay. Sound familiar? Yes, banks have started to play the same elaborate and convoluted pricing game that airlines, hotels and a host of other industries have perfected, charging customers different prices for an identical product or service. Only what's at stake isn't a one-way ticket to Phoenix that might cost one customer $80 and another $800, but some of the largest financial transactions most folks will ever make: mortgages, car loans and home-equity lines of credit.

The move to price optimization, which most banks are still only testing, has been spurred by the mammoth challenges threatening the $6 trillion lending industry. Subprime lending losses contributed to a 45% drop in bank earnings last quarter, and mortgage-loan volume is expected to tumble 16% this year. Analysts say banks are looking to price optimization as a relatively quick and easy way to boost a sagging bottom line by as much as 5 to 10% in three to six months. "The return on investment is huge," says Terry Kuester, a banking consultant with Deloitte & Touche. "It's a huge opportunity for banks."

When airlines first moved to this sort of high-tech pricing in the '80s, consumers howled. Though optimization led to lower fares for some, fliers argued it smacked of price gouging, because fares go up when people need to travel the most, like during holidays and school breaks. But in the case of loans, customers will only be able to guess when the bank is swapping in higher rates. "I think it's terrible," says Merrick, N.Y., mortgage broker Robert Bram. "I don't believe anyone should be charged a higher rate just because they aren't as rate-conscious as the next guy."

THERE IS, OF COURSE, nothing new about banks offering different rates to different customers; lenders have long used sophisticated statistical models to set higher rates on risky loans made to customers with bad credit. But now, says TowerGroup consumer-lending analyst Bobbie Britting, banks are turning all that statistical firepower toward sniffing out profit-boosting opportunities. Most won't discuss or confirm the practice, but insiders say Wachovia and Washington Mutual are using the technology to set rates on home-equity loans, and Citibank is testing its own in-house version of the technology. Bank of America, meanwhile, has experimented with mortgage loans, and big auto lenders like Ford Motor Credit and AmeriCredit are using it to help price car loans.

One lender,

SunTrust

The result? A complicated pricing system that only a computer could love. Gone are the days when everyone with a 720 credit score gets offered the same rate on a $50,000 home-equity loan. Some banks are coming up with different rates for as many as 20,000 customer segments defined by variables like location, loan type, transaction history and banking habits. Prefer to apply at your local branch? A computer may decide you'll typically accept higher rates than those who apply online or by phone. Live in the Midwest or a rural outpost? The software may suggest you're likely to stomach higher rates than customers living in big coastal cities. While it's not surprising to learn that unsophisticated consumers with low credit scores often accept higher rates than they should, banks have also discovered that loyal customers are often more likely to accept a high rate. Tom Schwartz, vice president of profitability analytics at AmeriCredit, a $13 billion auto lender, says his company segments by geography but will also charge different interest rates depending on which of its three subsidiary companies AmeriCredit, Long Beach Acceptance Corp. and Bay View Acceptance Corp. the customer approaches first. And the technology's not just for new loan applicants, says SunTrust's Caron. If you have a home-equity line of credit you're not using, the bank might surprise you with a customized incentive offer at the very moment you're considering a kitchen renovation.

Mark Ferguson, an operations management professor at Georgia Tech and optimization proponent, says that when the airline and car-rental industries adopted the technology, it produced not only higher profits for companies but also lower prices for consumers. That's because optimization seeks to increase sales by getting the less price-sensitive customers to subsidize discounts for everyone else. When lenders use the strategy, 40% of their customers typically get lower rate offers, and rates stay the same for another 20%. Still, banks are well aware of the pain that the other 40% experience. Even though most "optimized" rate changes are small typically no more than half a point in either direction that half point can add an extra $70,000 on a $600,000 30-year mortgage. Indeed, Richard DeLotto, an analyst at the Gartner Group, recalls seeing bankers taking sudden bathroom breaks when the topic came up at conferences. "They don't even want to be associated with it," he says.

Lenders themselves say the strategy can be self-correcting: If they consistently make unreasonably high loan offers, they'll lose business to other banks, and the software will start suggesting lower prices. Indeed, "customers have the opportunity to not take the deal," adds AmeriCredit's Schwartz. "It's an open marketplace." The issue for consumers, of course, is trying to figure out which side of the price-optimization equation the software has relegated them to which may ultimately be the banks' secret weapon here. After all, says Kuester, the Deloitte & Touche consultant, "customers might never know they could've gotten a better offer."

Also See:

• The Two-Headed Pricing Prophet