Intraday Predictability of Overnight Interest Rates

Abstract

Lee (2001) found that the overnight Eurodollar rate in London and the effective Fed funds rate exhibit similar calendar day effects although the absolute magnitudes are slightly less. The excess return on overnight Eurodollars over Fed funds is predictable based on the lagged overnight Eurodollar rate, the lagged Fed funds rate and calendar day dummies. Explanations for the smaller calendar day effects on the overnight Eurodollar rate at 11:00 am GMT than on the effective Fed funds rate, and the predictable excess return include the difference between market-specific conventions in the two markets and the time difference in measuring the two overnight interest rates. This paper investigates the relationship between the Fed funds rate at 11:30 am EST, the effective Fed funds rate and the overnight Eurodollar rate in London. It is found that the predictable excess return is caused by both the difference between the market structures and by the measuring time difference.

Overnight interbank loan markets

“…In particular, trading in Eurodollars and repo, as well as direct trades of federal funds, tends to be concentrated during morning hours (U.S. Eastern time), while trading in brokered fed funds is heaviest toward the end of the business day. However, Lee (2003b) finds that time of day trading patterns cannot account for the differences between Eurodollar and brokered fed funds rates, and therefore argues that differences in bid-ask spreads, line-limits, or transaction costs must account for the more muted response of Eurodollar rates to the Federal Reserve's reserve accounting conventions in her sample. Our findings are not inconsistent with the notion that differences between the markets account for their different patterns of calendar effects.…”

Money Market Integration

“…Currently, most derivative products, such as options and futures traded in Chicago, are still priced off London Eurodollar rates, but the New York market likely represents the main venue for borrowing of Eurodollars by U.S. banks. 4 Given the effective substitutability of federal funds and Eurodollar deposits for U.S. banks, the evidence uncovered by Cyree, Griffiths, and Winters (2003), Lee (2003aLee ( , 2003b, and Demiralp, Preslopsky, and Whitesell (2004) that yields on these instruments show sizable and predictable gaps is surprising. Some of this evidence can be explained by its inclusion of data from the pre-1990 Eurodollar liberalization period.…”

Money Market Integration

“…Given the effective substitutability of federal funds and Eurodollar deposits for U.S. banks, the evidence uncovered by Cyree, Griffiths, and Winters (2003), Lee (2003a, 2003b), and Demiralp, Preslopsky, and Whitesell (2004) that yields on these instruments show sizable and predictable gaps is surprising. Some of this evidence can be explained by its inclusion of data from the pre‐1990 Eurodollar liberalization period.…”

“…Some previous evidence of federal funds—Eurodollar segmentation, comparing London mid‐day Eurodollar rates with New York mid‐day or daily‐averaged funds rates, may also reflect a time‐aggregation bias: the London–New York time lag causes Eurodollar rates to lead federal funds rates, possibly causing correlated (i.e., conditionally predictable) yield spreads 6 . Using intra‐day Eurodollar quotes, however, Lee (2003b) partly addresses issues of time synchronization and still finds evidence of federal funds–Eurodollar segmentation.…”

“…From the viewpoint of the former, it is important to re‐evaluate recent evidence of segmentation between the markets for federal funds and Eurodollars (Cyree, Griffiths, and Winters 2003, Lee 2003a, 2003b, Demiralp, Preslopsky, and Whitesell 2004) for at least two reasons. First, such evidence is surprising, given the similar regulatory treatment of federal funds and Eurodollars since 1990.…”