To reach a solution, treasury had to work closely with many interested parties, including the boards of directors and the credit, operations, finance, legal, tax, market risk, marketing and communications departments from both banks, Braithwaite explains. Ultimately, the final plan was two-pronged:

o Change sweep rules to divert deposit flows to Merrill Lynch Bank USA, where they could be used to fund a variety of retail and commercial loans and a broader range of investments than was permitted for B&T; and

o Sell liquid assets off the USA books–primarily residential mortgage loans and loans secured by securities (equivalent to margin loans)–to B&T, freeing USA to make more lucrative but relatively illiquid commercial loans.

To reduce the deposits flowing to B&T, treasury proposed changing the sweep rules so that the first $100,000 of retirement accounts would go to USA instead of B&T and so that deposits by single account holders in B&T were capped at $100,000. Once approved, the rule change diverted $11.4 billion of funds from B&T to USA, effectively cutting B&T's deposit liabilities in half between 2001 and the second quarter of 2004. Net deposits at B&T declined from $15.62 billion at yearend 2000 to $9.54 billion on June 30, 2005, while they rose at USA from $39.29 billion to $54.91 billion over that same period. Those deposits, when used to fund illiquid loans, potentially could generate $60 million in additional profit each year. That took care of the liabilities.

To shift assets for an optimal solution, B&T bought $3.5 billion of residential mortgage loans and loans secured by retail securities from USA in late 2003, freeing up capacity in USA to make more commercial loans and creating a pre-tax potential profit increase for B&T of $85 million annually. That took care of assets. "We could have sold the mortgages in the secondary market to free up capacity at USA, but we know those loans and liked them," Braithwaite says. "For B&T, it was a chance to exchange low-yielding investment securities for higher-yielding mortgages. It meant a pickup in net interest margin for both banks."

Financial results were impressive. The program helped USA's net interest income jump from $384 million in 2000 to $1.477 billion in 2004 and to $830 million for the first half of 2005. At the shrinking B&T, the infusion of assets from USA brought net interest income from $80 million in 2000 to $186 million in 2004.

Return on assets, perhaps the best indicator of bank profitability, more than doubled for the combined banks, from 0.7% in 2000 to 1.4% in 2004 and 1.7% for the first half of 2005. At B&T, the increase was an eye-popping jump from 0.5% in 2000, before the strategy was implemented, to 2.6% in the first half of 2005.

Best of all, the rationalization and redistribution cost almost nothing, aside from a slight acceleration of state taxes, Braithwaite notes.

SILVER AWARD WINNER

EMC Corp.

Cash-rich EMC Corp. faced a dilemma in mid-2003. If interest rates rose even a modest 20 basis points, its bonds would lose principal value equal to one quarter's net income and "put us at risk of posting a negative total return," says Bob Fanelli, investment manager of the $8.2 billion firm based in Hopkinton, Mass. Moving to cash to protect principal meant accepting a 1% return, well below the inflation rate. EMC needed an asset with low or negative correlation to its bonds. The answer turned out to be an excerpt from modern portfolio theory, which holds that aggregate risk is what counts, not the risk of individual assets.

So EMC bought riskier assets to reduce risk: It sold $600 million of short-term A-rated bonds from its $6 billion investment portfolio and reinvested it in BB-rated bank loans. The result: improved return and lower risk. "By combining different asset classes with low correlations, an investor will move closer to the aggregate portfolio that produces the highest expected return for a given level of risk as measured by standard deviation," explains Fanelli.

The loans' correlation with the bonds was a negative 0.10. The market for such loans had become suitably liquid–trading volume rose from $3.6 billion in 1991 to $87.4 billion in 2003. Investment firms offered mutual funds invested in bank loans. Moreover, bank loans were senior and secured. Even in default, the recovery rate was 100%, versus 41% for senior unsecured debt and 5% for junior subordinated debt.

The numbers showed that a 10% allocation to bank loans could be expected to increase return by two basis points and reduce risk by 21 basis points. Management was impressed and revised the investment policy, which previously had banned securities rated below A-.

The change in investment strategy was a clear winner. "For the 21-month period ended March 31, 2005, the actual cumulative total returns were: 6.45% for our bank loan portfolios, 1.98% for our short-term bond portfolios and 2.42% for three-month T-bills," Fanelli reports. "In dollar terms, the excess return for the $600 million bank loan strategy amounted to $26.82 million over our short-term bond portfolios and $24.18 million over cash. That level of excess return is probably not sustainable, but we think 13 basis points of excess return is reasonable over a long time frame."

SILVER AWARD WINNER

University of Pittsburgh Medical Center

What do you do when you need to build a system for small consumer payments that must stretch across more than 20 service providers with 17 different billing systems? And you have to do it with no new hardware or software at the point of payment, no new hires for the project and a budget under $360,000, and you have to accommodate significant variations among the service providers. If you're the University of Pittsburgh Medical Center (UPMC), you do it with U-Pay, home-built software. And you do it in just six months.

Years ago, when patient payments were less than 5% of revenue, UPMC, like most health-care systems, didn't push consumer collections. But as average co-pays keep rising, something had to change. "Given that UPMC has over 140,000 physician office visits and 13,000 hospital outpatient visits per month, the sums at stake were clearly too large to ignore," says Linda Zang, director of treasury operations.

The software that UPMC wrote is able to enter cash, check or credit card payments; authenticate credit cards within three seconds; print receipts; show balances on-screen; void transactions; e-mail reports to supervisors daily or monthly; and post payments electronically to the downstream billing systems.

Zang and her team expected U-Pay to process 10,000 transactions a month but by March 2005, 88,000 payments worth $6 million were being collected through the system, which is now used at 382 worksites, with 99% uptime, she reports. With this robust usage, U-Pay cut annual bad-debt writeoffs by $250,000 and saved another $100,000 a year by consolidating credit card processing. Hospital point-of-sale collections are up more than 50% since U-Pay was introduced, and patient complaints have dropped now that payments are posted quickly. "For the first time, payment dollars and transaction volume are visible regardless of which billing system is used and where a user is located," Zang says.

BRONZE AWARD WINNER

Honeywell International Inc.

Honeywell International Inc. knows how to build effective cash management–in this case global cash pooling–on a technological foundation. Action had to follow knowledge. Large pockets of international cash were effectively hidden until treasury opened a Web site in 2002 and gradually got business units to report all cash there. Then it automated the process with a treasury workstation (Quantum) in 2003 and 2004.

It added up, at the end of 2002, to $2 billion, which was held by 233 legal entities in 1,400 accounts at 145 banks in 46 currencies in over 60 countries, much of it earning low interest or no interest, reports Jim Colby, assistant treasurer. Very little of it was in the United States. As cash grew to $4.1 billion by first quarter 2005, the rewards for managing it well became even greater.

So when Honeywell's treasury set out to tighten global liquidity management, it could see where the cash was. Success meant getting idle cash out of bank current accounts into cash pools and then into productive investments. An interest income budget was set, based on forecasted cash balances, one-month interbank deposit rates and FX rates in each country. Performance was measured against that budget. How well cash was invested figured in incentive compensation reviews.

It worked. Actively managed cash grew from 68% in 2002 to 95% today, a big reason why interest income almost doubled, to $130 million, between 2002 and 2004. At least $11 million of the $62 million increase came from the global investments project, says Colby. Compliance with investment policy was 100%.

An additional value of tighter management of international cash became evident in December 2004 when $28 billion Honeywell had to finance a $2.4 billion U.K acquisition. "We were readily able to identify which cash we could use, project the impact on future interest income and present senior management and the rating agencies with a funding plan, confident of our ability to mobilize the cash on short notice," Colby says.