Monetary Transmission : The Federal Funds Rate and the London Interbank Offered Rate ( LIBOR )

This paper examines the effectiveness of a monetary transmission mechanism from the federal funds rate to the London Interbank Offered Rate (LIBOR). In particular, the paper employs a co-integration and vector error correction models to examine the degree and the direction of pass-through from the federal funds rate to the LIBOR. Two sub-periods are selected, 1987:02-1994:01 and 1994:02-2002:05, in order to examine this relationship. Results indicate a significant co-integration relationship between the federal funds rate and the LIBOR for the first and second periods. However, in the second period, the two variables adjust differently to a deviation from equilibrium. JEL classifications: E00, E3, E4, E5, E6, F3, F4, G1, G2, N1, N2


Introduction
Monetary policy impacts the economy through a number of transmission channels.In recent years, the Federal Reserve has increasingly focused on the federal funds rate (FFR) -the interest rate on overnight loans of bank reserves from one bank to another -as the primary instrument of monetary policy.Thus, a ~ 2 ~ large amount of literature has emerged examining the relationship between the federal funds rate and market interest rates (see, Calem &Mester, 1995 andScholnick, 1999).This paper extends the research of Atesoglu (2003;2005), Payne (2006) and Nishiyama (2007) by examining the relationship between the Federal Funds Rate (FFR) and the London Interbank Offered Rate (LIBOR).The LIBOR is a benchmark rate that leading international banks charge each other for shortterm loans (Kiff, 2012).As Mishkin (2015) states: "One of the key transmission channels of monetary policy is the exchange rate.A tightening of monetary policy, for example, raises U.S. interest rates relative to those abroad, thereby inducing upward pressure on the foreign exchange value of the dollar.An appreciation of the dollar, in turn, restrains exports (because the price of U.S. goods rises when measured in foreign currencies) and stimulates imports (because imports become cheaper in dollar terms).The resulting decrease in net exports implies a reduction in aggregate demand.In addition, an appreciation of the dollar that leads to a decline in import prices also helps restrain overall U.S. inflation…" However, if foreign interest rate, the LIBOR in this paper, moves quickly in tandem with the domestic interest rate, the FFR in this paper, then this foreign channel is weak.That is, if foreign interest rates respond in the same direction and magnitude as the domestic one, then the effects of the international channel would be minor.This paper employs the co-integration and vector error correction models in order to empirically examine the relationship between the federal funds rate and the LIBOR (see Engle & Granger, 1987;Heffernan, 1997).Following Atesoglu (2003;2005) two sub-periods are selected for examining this relationship.The first period spans from 1987:02 until 1994:01 and the second from 1994:02 to 2002:05.These periods are chosen because they represent two distinct policy regimes exercised by the Fed.Since the purpose of this paper is to investigate a case study of a specific change in monetary policy that occurred in February 1994, it is appropriate to concentrate only on data around this particular change.This paper quantifies the direction and degree of pass-through between the policy rates (the FFR) to foreign lending rates (the LIBOR).This paper studies both the short and long term effects of this relationship.
Additionally, this paper contributes to the ongoing debate between Post Keynesian economists about whether the structuralist or horizontalist perspectives offer the best explanation for the endogeneity of the money supply.As Payne (2006) and Palley (1991) explain, the structuralist hypothesis maintains that loan supply is influenced by the Federal Reserve's policies and is positively associated with bank lending determined by both loan demand and supply.On the other hand, the horizontalist approach asserts that the Federal Reserve determines the short-term cost of funds with banks placing a markup over the federal funds rate in the provision of loans.As a result, the loan supply schedule is horizontal with loan demand determining the amount of bank lending.
The remainder of the paper is organized as follows.Section II reviews the literature.Section III presents the data, methodology, and results.Section IV offers concluding remarks.

Literature Review
Monetary policy impacts the economy through a variety of monetary transmission mechanisms.The federal funds rate is the principle focus of the literature on monetary transmission since it remains the main tool through which the Federal Reserve conducts monetary policy.There exists extensive literature that studies the relationship between the federal funds rate and other financial market instruments.~ 3 ~ investigates the relationship between the FFR and long-term interest rates.Hofmann and Mizen (2004) use evidence from retail interest rates of individual financial institutions to explore the monetary transmission mechanism.In a sequence of papers, Sander and Kleimeier (2004;2006) and de Bondt (2005) study the transmission mechanism in euro-zone retail banking.Payne (2006) and Payne and Waters (2008) investigate the response of the conventional mortgage rate to the federal funds rate.
With the exception of Payne (2006), the literature uses two sub-periods, 1987:02-1994:01 and 1994:02-2002:05.These sub-periods are chosen by Atesoglu op.cit.Because in February 1994, the Fed changed the way it conducted monetary policy by explicitly announcing its target for the FFR.This change marked an important shift in the way monetary policy was implemented.Thus, this paper uses the same sub-periods.The second sub-period begins in February, 1994, when, as stated above, the Fed began making its target for the FFR public.Before February 1994, the target for FFR was less transparent.The literature regarding monetary transmission maintains that changes in federal funds rate targeting policy at the beginning of the second period increased the pace with which changes in the FFR are transmitted to other interest rates.The change improved the effectiveness of monetary policy.Payne (2006) uses the same starting date of 1987:02 but extends the time range to 2005:06.Atesoglu (2003;2005) and Payne (2006) employ Johansen co-integration and vector error correction modeling methods to ascertain the relationship between the FFR and various domestic interest rates.They examine two principle measures.The first is the co-integration coefficient that measures the degree of pass-through between the FFR and the second rate.The second measure is the error correction term, which reveals the lagged residual of the estimated co-integration equation and thus indicates the directional causality between the two interest rates (see, Newey & West, 1994).Complete pass-through occurs if the co-integration coefficient is equal to unity, while incomplete pass-through occurs if the cointegration coefficient is less than one.Both Atesoglu (2003) and Payne (2006) find a high degree of passthrough from the FFR onto other interest rates.
The literature on monetary transmission concerns itself mainly with the ongoing debate between Post Keynesian economists proposing diverging explanations to the endogeneity of the money supply, e.g., Rosen (2002), Atesoglu op.cit.(2003), and Payne (2006).The principle approaches are the structuralist and the horizontalist.Conclusions that reveal an asymmetric error correction model indicate unidirectional causality in monetary transmission and support the horizontalist view.Atesoglu (2003;2005) and Payne (2006) reveal a stable long-term relationship and unidirectional causality between the FFR and other interest rates for the second sub-period.Atesoglu (2003) examines the relationship between the federal funds rate and the Prime Rate during the two periods: 1987:02-1994:01 and 1994:02-2002:05.The results from both periods demonstrate a positive co-integration relation indicating a pass-through from the federal funds rate to the Prime Rate.However, Atesoglu (2003) points to several differences between the two periods.The results from the first period reveal a bi-directional causality between the FFR and the Prime Rate while the second period displays a unidirectional causality that runs from the FFR to the Prime Rate.Moreover, the margin between the federal funds rate and the Prime Rate is more volatile in the first period than the second one.Furthermore, the data for the second period reveal that an upward shift occurs in the markup of the Prime Rate over the FFR.Atesoglu (2005) explores the relationship between the FFR and long-term interest rates, including the yields on both AAA corporate bond and the 30-year Treasury bond.The findings reveal a co-integrated relationship between the FFR and long-term interest rate.Atesoglu (2005) draws a division between the effects of the long-term interest rates in the short-term versus the long-term, a distinction that he does not make in his earlier paper (Atesoglu, 2003).Atesoglu (2005) demonstrates that in the short run the FFR does not have much effect on long-term interest rates, i.e., the first 12 months.In contrast with Atesoglu (2003), the peak effect from changes in the FFR occurs after about 30 months.The conclusion indicates that while there is an effect on long-term interest rates in the long run, the impact is smaller in the short run.Moreover, Atesoglu (2005) shows that despite moving together in the long run, the spreads between ~ 4 ~ AAA bond yield and the FFR, as well as the spread between 30-year Treasury bonds and the FFR, are quite volatile compared to the spread between the Prime Rate and the FFR found in (Atesoglu, 2003).Payne (2006) studies the relationship between the federal funds rate and the fixed mortgage rate.He finds co-integration relationships between both rates and an incomplete, unidirectional pass-through.Compared to Atesoglu (2003Atesoglu ( , 2005)), the co-integration coefficient is significantly different from unity for the relationship between the FFR and the fixed mortgage rate than for the relationship between the FFR and the Prime Rate or the long-term interest rate.Similar to Atesoglu (2005), Payne (2006) analyzes the relationship between the FFR and the fixed mortgage rate distinguishing between short and long-term effects.The findings by Payne (2006) dovetail Atesoglu (2005) by showing that the error correction model indicates a less significant relationship between the mortgage rate and federal funds rate in the short run, with the mortgage rate displaying a larger adjustment in the long run.
Researchers also study the relationship between the LIBOR and monetary policy.Fuertes and Heffernan (2009) find a sluggish adjustment of loan and deposit rates to changes in the LIBOR for the United Kingdom (UK).Ahmad, Aziz and Rummun (2013) study the relationship between the LIBOR and four other UK retail interest rates.The authors find evidence of incomplete pass-through in the short run but fairly complete pass-through in the long run.
The current paper is the first study to investigate the pass-through of the Federal Reserve's policy changes to non-domestic interest rates (see Mishkin, 2009).The LIBOR in this paper represents the nondomestic foreign interest rate.

The Data, Methodology and Results
This paper investigates the relationship between the FFR and the LIBOR before and after a major change in the disclosure by the Fed of its target for the FFR.Beginning in February 1994, the Fed clearly announced its target for the FFR, while prior to that date, the target was opaque.  1 shows that the two interest rates closely move together.

Unit Root Tests
Table 1A reports the Augmented Dickey-Fuller (ADF) tests for the FFR and the LIBOR for the two periods (see Dickey & Fuller, 1979).When estimated in the level of the variables, the unit-root tests reject the assumption of not having a unit root, implying that the relationships among the various variables, if analyzed in the level, are spurious.

Granger-Causality Tests
Table 2 presents the results of the Granger-Causality tests using the first differences of the variables, since the levels of the variables are non-stationary.The table shows that, for the two periods examined, the direction of causality is from FFR to LIBOR.

Figure 1 :
The FFR and LIBOR The data consist of monthly time series of two interest rates, the federal funds rate and the LIBOR.The data for the federal funds rate and the LIBOR are obtained from the Federal Reserve Economic Data (FRED), a database maintained by the Federal Reserve Bank of St. Louis.The entire period under study spans from February, 1987 until May, 2002.Figure