Inventing the Short Term Fund

A New Approach to Cash Management 30 Years Ago 

by George F. Keane, Commonfund President Emeritus

When Commonfund received its initial grant from the Ford Foundation in 1971, $500,000 of the total $2.6 million was earmarked to do research into ways in which colleges might improve their financial management. It was understood from the research and discussions of the previous four years that the initial Commonfund program would be a pooled common stock investment fund with professional management. The need for such a program was made clear in the Ford Foundation-sponsored study “Managing Educational Endowments,” published toward the end of 1968, and the Ford grant was intended primarily to pay the start-up costs of the common stock investment program.  With that well underway, in the latter part of 1972 we began to consider how to get started on the research program.

“We commissioned a 1973 research study of cash management practices at a selected group of 55 colleges and universities of various sizes…”

One of the things that had struck me during the start-up period was the fact that of the more than 2,000 colleges and universities then in existence, only about 600 had endowment funds of meaningful size. While these institutions would certainly benefit from better management of their funds, that would not help the majority of institutions.

It also occurred to me that all educational institutions had to manage cash, and that the nature of their operations was such that they received cash inflows from tuition in two or three large payments per year, then spent these funds gradually over the course of the year. Meanwhile, they had funds that could be invested to generate income. Our Research Committee, headed by economist Roger Murray of Columbia University, approved a project to examine how colleges managed their operating cash and whether a pooled approach might generate increased income from this resource.

“The findings of this study were fascinating. …a pooled fund would have considerable stability and would be very sizable.”

We commissioned a Cambridge, Massachusetts-based research group, Technology Management Inc., to develop the basic information. I had previously worked with Colin Scarborough of that firm on a 1967 project to improve financial planning for colleges and knew that the firm had a good understanding of college operations. Colin designed a research study of cash management practices at a selected group of 55 colleges and universities of various sizes, geographically diverse, including both public and private institutions.

The findings of this study were fascinating. We discovered that the amount of cash managed by these institutions was much larger than we had anticipated. The inclusion of publicly supported colleges and universities raised the total considerably – to the point that for higher education as a whole, cash management approached endowment management as an opportunity for increasing investment returns. We also learned that institutions received and spent funds at differing times, meaning that a pooled fund would have considerably greater stability in terms of size than we would have imagined. Moreover, cash management had not been given much attention, and there were no institutions that thought of actively managing their cash investments. In fact, while most institutions were reasonably diligent about keeping cash invested, many smaller colleges did not and, on average, only about 80 percent of their cash was actually earning interest.

“Through diligent effort, we found some interesting examples of active cash management that might provide improved returns.”

The second part of the study was to search for and identify cash management investment strategies that might improve investment returns. This was not an easy task because while there were hundreds of professional firms engaged in active management of common stock portfolios, there were very few that were focused on the management of cash investments. The concept of a money market mutual fund had been introduced by Harry Brown and Bruce Bent when they created the Reserve Fund in 1971, but that did not seem to be a promising model for what we had in mind. While it enabled small investors to participate in large denominated CDs and commercial paper, the mutual fund format offered little opportunity for total return management, and administrative and regulatory expenses cut into returns.

Through diligent effort, Colin did find some interesting examples of active cash management that might offer more promise. He studied the investment management of the traveler check float held by American Express, as well as the cash management capabilities of several large commercial banks, such as First National of Chicago, that were active in corporate cash management for customers. The three most interesting, however, were: 1) an active cash management program developed by Bank of America, serving as agent for the state of Alaska for the investment of the $900 million the state had received for the sale of the North Slope oil rights in 1969; 2) a pooled cash management program that had been developed by the Federal Home Loan Banks; and 3) a pooled fund that had been established for federal credit unions, managed by Brown Brothers, Harriman.

“ ‘We tested’ about 100 various smoothing formulas using computer based simulation techniques. We came up with the one that is still used today for the Short Term Fund.”

The credit union fund ultimately became the model we chose to pursue. It had been established as a bank common trust fund under the supervision of the office of the Comptroller of the Currency, the Federal agency that regulates banking in the United States. The comptroller's office had approved a plan for the credit unions that permitted a smoothing of investment gains and losses over time, which facilitated active, total return management. The SEC, which regulated most of the investment industry, was firmly committed to daily mark-to-market accounting, which would inhibit active cash management.

We studied the smoothing formula used by the credit union fund, and concluded that it had the disadvantage of holding back too much of the earnings by spreading gains and losses over a five-year period. Steve Francis of Fischer Francis Trees & Watts and I worked with the staff of Technology Management to test about 100 various formulas using computer based simulation techniques. We came up with the one that is still used today for the Short Term Fund. Each month the actual return on the fund is compared to the interest rate on 90-day Treasury bills. If the fund’s total return is greater than the Treasury bill rate – which usually happens during periods of stable or falling interest rates – then the Treasury bill rate plus half of the excess return is credited to participants that month and the remainder is credited to a reserve account. If the total return is less than the Treasury bill rate – which can happen in a rising interest rate environment – then the fund pays the Treasury bill rate and charges the difference against the reserve account.

In developing this formula – which may sound like something worked out on the back of an envelope – we also considered not only the historical price volatility of short duration fixed income securities, but also looked at how such a formula would work if that volatility were to double in the future. This was fortunate, because over the next 10 years, that actually happened. The Short Term Fund weathered that unexpected storm, while the credit union fund formula proved inadequate to meet the challenge. Finally, we studied various limits for the reserve account and decided on a 2 percent limit, so as to hold back only the minimum needed to smooth the credited earnings.  Over the past 30 years we have had several interest rate cycles, with short term interest rates ranging from more than 18% in 1981 to less than 2% during the past year. The Short Term Fund has worked well in both rising and falling interest rate environments, thanks in part to the effectiveness of this smoothing formula.

During 1973, we interviewed potential investment managers for the proposed new fund. We were most favorably impressed with the performance record of the team at Brown Brothers, Harriman in managing the credit union fund, but that year the key members of that team – Dick Fisher, Steve Francis, Jim Trees and John Watts – had departed to form their own firm. Our other finalist was the Bank of America, based on its record of managing the Alaska fund. Choosing between one of the largest banks and a start-up investment firm was a tough choice, but we made the right decision and hired Fisher, Francis, Trees and Watts, a firm that has had consistently good results for nearly 30 years now, and still manages a major portion of the fund. Brown Brothers did not lose out completely, since the firm was also hired as a manager years later when the fund grew much bigger. 

We also interviewed a number of banks to serve as the Trustee for the fund, and selected a mid-size bank, Philadelphia National, to fill this role. Although the bank has gone through a number of name changes over the years as the consolidation of the banking industry proceeded, the team that was assembled 30 years ago had remarkable stability and continuity over the years, and has done a first rate job.

“Over the past 30 years, this program has enabled its participating educational institutions to earn hundreds of millions in additional income to support their important mission of educating America’s young people.”

The final step was to get approval from the office of the Comptroller of the Currency for our novel approach to managing cash for educational institutions. This took a number of meetings and discussions, but with the important help of attorney Roy Haberkern of the Milbank, Tweed, Hadley and McCloy law firm, we finally succeeded.

Meanwhile, the original common stock fund had been going through very difficult times because of the 45 percent decline in stock prices that had occurred from the end of 1972 through September of 1974. Some members of our Board of Trustees began to wonder whether Commonfund could survive this terribly discouraging time, and questioned whether this was the right time to be starting a new investment fund. The Board, however, did approve the recommendation by the Research Committee, which I strongly supported, that we proceed to start this important new program. So, at the very bottom of the worst stock market decline since the 1930s, we launched the Short Term Fund in September of 1974. It took a number of years and a considerable effort to grow this into the major program that it has become, but the innovative structure and the choices of managers and trustees proved over time to be right. It is great to be able to look back over many years to those pioneering days and remember what an exciting and challenging time it was.

This program over the past 30 years has enabled its participating educational institutions to earn hundreds of millions in additional income to support their important mission of educating America’s young people. That was the purpose for which the Commonfund Group was established, and this program has been a major component of its success. More importantly, the Short Term Fund and its related cash management activities have grown to become a very large program serving 1,500 educational institutions. It has the clear potential to continue growing and to expand its contribution to higher education for many years to come.

Refer this article to a friend