Large Retail Time Deposits and U.S. Treasury Securities (1986-95): Evidence of a Segmenting Market (WP 97-12)

Abstract

Although investors face an ever-growing menu of securities with divergent cash flows and risk patterns, some continue to place funds in fixed-rate, fixed-maturity, default-risk-free instruments, primarily U.S. Treasury securities and FDIC-insured time deposits. Given free, easy access to noncompetitive bids in the primary Treasury market by means of the Treasury Direct program, investors could plausibly view Treasury securities and large retail time deposits (up to $100,000) as very close substitutes, and demand equal pricing. We test this hypothesis using weekly series of Treasury and large retail time deposit yields for three maturities over a 10-year period (1986-95). The data reveal a pattern of equilibrium pricing between 1986 and 1990. However, the data also show that large retail time deposits were routinely underpriced relative to Treasuries from mid-1990 onward. We conclude that the change in the pricing of large retail time deposits reflects increasing market segmentation. That is, the flight from insured balances into money market funds and other uninsured investments has left banks with a group of investors who are uncommonly insensitive to interest rates or, equivalently, have the highest switching costs. In addition, we find that increases in and persistence of the negative spread beginning in 1990 had no measurable effect on changes in aggregate time deposit balances and that average large time deposit yields exhibit stickiness relative to those of U.S. Treasuries. This suggests that banks are able to extract available surplus from the depositor segment of the risk-free debt investors. This surplus helps to make up for the loss of the banks' more rate-sensitive investors.

Authors

James H. Gilkeson and Gary E. Porter