WORKING PAPER SERIES INTEREST RATES AND OUTPUT IN THE LONG-RUN NO. 434 / JANUARY by Yunus Aksoy and Miguel A. León-Ledesma - PDF

Description
WORKING PAPER SERIES NO. / JANUARY 00 INTEREST RATES AND OUTPUT IN THE LONG-RUN by Yunus Aksoy and Miguel A. León-Ledesma WORKING PAPER SERIES NO. / JANUARY 00 INTEREST RATES AND OUTPUT IN THE LONG-RUN

Please download to get full document.

View again

of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Public Notices

Publish on:

Views: 20 | Pages: 19

Extension: PDF | Download: 0

Share
Transcript
WORKING PAPER SERIES NO. / JANUARY 00 INTEREST RATES AND OUTPUT IN THE LONG-RUN by Yunus Aksoy and Miguel A. León-Ledesma WORKING PAPER SERIES NO. / JANUARY 00 INTEREST RATES AND OUTPUT IN THE LONG-RUN by Yunus Aksoy and Miguel A. León-Ledesma In 00 all publications will feature a motif taken from the 0 banknote. This paper can be downloaded without charge from or from the Social Science Research Network electronic library at Part of this paper was written while Aksoy was visiting the Research Department of the European Central Bank as research visiting fellow. He would like to express his gratitude to members of the research department, in particular to Frank Smets, for providing stimulating research environment and hospitality.we would like to thank to an anonymous referee,alan Carruth, Paul De Grauwe, Hans Dewachter, Sylvester Eijffinger, João Faria, Jim Griffin, Marco Lyrio, Robert Miller,Abhinay Muthoo, Hashem Pesaran, Ron Smith, Bas Werker and to participants in seminars at the European Central Bank, at the universities of Leuven, Essex,Tilburg, Nottingham and EEA-Madrid, Ecomod00 Paris and MMF00-London conferences and Kent Money/Macro workshop for their comments and suggestions. Naturally, we assume responsibility for all remaining errors. Corresponding author: School of Economics, Mathematics and Statistics, Birkbeck College, University of London, Malet Street London, WCE 7HX, United Kingdom; tel: (00) 7 07, fax: (00) 7, Department of Economics, Keynes College, University of Kent, Canterbury CT 7NP, United Kingdom; European Central Bank, 00 Address Kaiserstrasse 9 0 Frankfurt am Main, Germany Postal address Postfach Frankfurt am Main, Germany Telephone Internet Fax Telex ecb d All rights reserved. Reproduction for educational and noncommercial purposes is permitted provided that the source is acknowledged. The views expressed in this paper do not necessarily reflect those of the European Central Bank. The statement of purpose for the Working Paper Series is available from the website, ISSN -00 (print) ISSN 7-0 (online) CONTENTS Abstract Non-technical summary. Introduction. Data 9. Interest rates as monetary policy indicators 0. Univariate time series properties. Long term relationship tests: taking statistics seriously. Cointegration. Stability of cointegration relationships. Bounds tests: taking economic theory seriously 7. Are there omitted variables?. Conclusions References Appendix: data sources Tables and figures European Central Bank working paper series Abstract In this paper we argue that both statistics and economic theory-based evidence largely indicate the absence of long run relationships between the real output and the most relevant monetary indicator for the U.K. and the U.S, short term interest rates. These findings are not only a full sample result, but also valid in most of the sub-samples throughout the second half of the 0 th century and are robust to the inclusion of possible omitted real variables. JEL classification: E, E, E. Keywords: information value, long term relationship, cointegration, bounds tests. Non-technical summary In this paper, we focus on the relevant monetary policy business cycle indicator, short-term nominal interest rates, and their long-run relations with output. There are at least two reasons to study the matter. First, nominal short term rates are explicit, controllable, operating targets and more importantly useful indicator variables to explain business cycle fluctuations. We investigate implied long term equilibrium relationships between the operating target/indicator variable and real output. Second, given high degree of inflation inertia, nominal rates tend to track very closely ex-ante or ex-post real short term interest rates. In Woodford s (00, p.) words once one recognizes that many prices (and wages) are fairly sticky over short time intervals the arbitrariness of the path of nominal prices implies that the path of real activity and the associated path of equilibrium real interest rates are equally arbitrary. It is equally possible, from a logical standpoint, to imagine allowing the central bank to determine, by arbitrary fiat, the path of aggregate real activity, or the path of real interest rates. This clearly poses some challenging theoretical as well as empirical questions. If we believe that real interest rates proxy the price of capital, an exogenous change in the policy variable, nominal rates, should also have long lasting implications in the real output path via standard aggregate demand channels. Think for example sensitivity of consumption, housing demand or manufacturing investment decisions to short term rates. Alternatively, by adjusting the nominal interest rate, the policymaker may simply be accommodating changes in the structural conditions in the economy taking into account the term structure of inflation expectations. In that case, even under inflation inertia, equilibrium long term real rates may be disconnected from short term real rates formulated in markets short and long term inflation expectations. In that case we do not expect to see a long term relationship between nominal and real short term interest rates and real output. Otherwise, we do. We are explicitly interested on the information content of policy indicators in explaining long term equilibrium real output. Therefore, issues raised by earlier work related to structural models and Lucas critique is not of direct relevance. Our results show that both statistics and economic theory based evidence largely rejects the existence of long term relationships between the relevant policy indicators and real output. The absence of long run relationships between short-term interest rates and real output is not only a full sample result, but also valid in most of the subsamples in the post Second World War period and are robust to the inclusion of possible omitted real variables. One can also interpret these findings as evidence of some support for the long-term monetary policy neutrality hypothesis. . Introduction Relationship between monetary policy and real output and inflation has always been the core focus of monetary policy research. First of all, policymakers, financial analysts and researchers are interested to know how effective monetary policy to stabilize business cycle fluctuations is. There is by now a huge volume of theoretical and empirical literature in this area of research. Secondly, the very same people would like to know how monetary policy decisions to stabilize business cycle fluctuations affect long term equilibrium real output, inflation and other fundamentals. In this paper we are interested in investigating empirically the second question. We analyze long-term relationships between monetary policy indicators that are able to explain U.S. and the U.K business cycle fluctuations and real output. In other words, we will make use of statistical information that comes from business cycle research to analyze long term equilibrium relationships between monetary policy indicators and real output. In order to stabilize business cycle fluctuations most central banks employ operating targets in the form of short term interest rates or monetary aggregates. Since the seminal work of Poole (970) it is well known that, in a frictionless certainty equivalent economy, money supply and interest rate policies to stabilize business cycle fluctuations would be identical. In this view money demand volatility is the sole criteria to judge on the operating target. Furthermore, even if monetary aggregates or short term interest rates are not used as operating targets these can be used as indicator variables if these contain useful information to explain business cycle fluctuations. However, in an economic environment with large uncertainties and real and nominal rigidities this instrument equivalence tends to disappear. More worryingly, recent empirical research provides statistical evidence that after around 9 the relationship between the changes in the U.S. monetary aggregates and macroeconomic fundamentals collapsed and therefore the information content of monetary aggregates simply vanished. To date, monetary aggregates can be characterized at best as weak indicators for real output whereas there is substantial evidence that short term interest rates are useful monetary indicator variables in explaining real output in the U.K. and U.S. in all subsamples available from the second half of the 0 th century. Currently, in most developed economies short-term nominal interest rates are employed as operating targets with the aim to stabilize business cycle fluctuations. Accordingly business cycle research has shifted its focus from monetary aggregates to nominal interest rates in analyzing monetary policy effectiveness. See e.g. detailed survey in Walsh (00). See for example Bernanke and Blinder (99), Friedman and Kuttner (99, 99), Estrella and Mishkin (997), Friedman (99), Stock and Watson (999, 00). In the literature several explanations are provided to understand as to why U.S. money demand became very unstable. Among others, these are innovations related to mortgage activity, (Board of Governors of Federal Reserve System, 00) foreign holdings of U.S. Dollars (Aksoy and Piskorski (00), the spread of sweep accounts (Anderson and Rasche (00), and increased use of plastic cards. See also Duca and VanHoose (00) for an excellent survey on the recent empirical developments on money demand. On the other hand, long term monetary neutrality is the key building block of mainstream business cycle research. Business cycle analysis relies on the assumption that monetary innovations can not have long term implications on the equilibrium level of real output in the long run. To date, empirical consensus is in favor of the long-term neutrality of monetary aggregates on key real economic variables, such as GDP and industrial production. In this paper, we focus on the relevant monetary policy business cycle indicator, short-term nominal interest rates, and their long-run relations with output. There are at least two reasons to study the matter. First, nominal short term rates are explicit, controllable, operating targets and more importantly useful indicator variables to explain business cycle fluctuations as argued above. We investigate implied long term equilibrium relationships between the operating target/indicator variable and real output. Second, given high degree of inflation inertia, nominal rates tend to track very closely ex-ante or ex-post real short term interest rates. In Woodford s (00, p.) words once one recognizes that many prices (and wages) are fairly sticky over short time intervals the arbitrariness of the path of nominal prices implies that the path of real activity and the associated path of equilibrium real interest rates are equally arbitrary. It is equally possible, from a logical standpoint, to imagine allowing the central bank to determine, by arbitrary fiat, the path of aggregate real activity, or the path of real interest rates. This clearly poses some challenging theoretical as well as empirical questions. If we believe that real interest rates proxy the price of capital, an exogenous change in the policy variable, nominal rates, should also have long lasting implications in the real output path via standard aggregate demand channels. Think for example sensitivity of consumption, housing demand or manufacturing investment decisions to short term rates. Alternatively, by adjusting the nominal interest rate, the policymaker may simply be accommodating changes in the structural conditions in the economy taking into account the term structure of inflation expectations. In that case, even under inflation inertia, equilibrium long term real rates may be disconnected from short term real rates formulated in markets short and long term inflation expectations. In that case we do not expect to see a long term relationship between nominal and real short term interest rates and real output. Otherwise, we do. In any case it is important to know the implied long term equilibrium relations between the changes in the policy rate or monetary indicator variables that can explain business cycle fluctuations and the fundamentals. In this paper, we are interested in the issue in an empirical sense. We are not aware of a study that systematically analyses long-term statistical relationships between real output and monetary indicators explicitly focusing on nominal short-term interest rates. Research by Bernanke and Mihov (99) probably See among others Bae and Ratti (000), Bernanke and Mihov (99), Boschen and Mills (99), Boschen and Otrok (99), Bullard (999), Fisher and Seater (99), Geweke (9), King and Watson (997), Serletis and Koustas (99), Weber (99) for testing neutrality of monetary aggregates in a structural framework. Note that in most of the empirical literature, hypothesis of superneutrality of monetary aggregates in general can be rejected. It is also worth to note that in most of this empirical research an analysis of long-term monetary neutrality is interpreted to be monetary aggregates neutrality instead of the more general concept of monetary policy neutrality. Throughout the paper, we also conduct tests based on U.S. ex-ante real interest rates. Our results confirm close co-movement of these two rates, being therefore statistically indistinguishable. 7 is the only exception to the literature in that it recognizes a causal role for nominal interest rates in the provision of liquidity into the economy and its implications in the long run. In their structural model, they find little evidence for rejecting either the liquidity effect or long term monetary neutrality. We are explicitly interested on the information content of policy indicators in explaining long term equilibrium real output. The information value approach for business cycle analysis as introduced by Sims (97, 90) allows us to address the issue on whether there is some reliable long run relationship between real output and potential instruments, such as interest rates. It is important to stress that the information value approach, as a first test of statistical connection between certain variables, is immune to questions of causality, exogeneity or controllability of potential instruments. In other words, as long as long-term swings in the policy indicator contain information about long term movements in income beyond what is already contained in movements in income itself, monetary policy can potentially exploit this regardless of whether the information it contains reflects true causation, reverse causation based on anticipations, or mutual causation by some independent but unobserved influence. Therefore, issues raised by earlier work related to structural models and Lucas critique is not of direct relevance. However, since an assessment of the long term relationships very much depends on the stationarity properties of the variables, we will carefully address the order of integration of variables. Although standard univariate analysis cannot reject the nonstationarity of most short-term real or nominal interest rate series, one cannot take this result at face value. Economic intuition suggests that short-term nominal interest rates should be rather stationary. In order to address this uncomfortable statistical feature of short term interest rates we proceed in two steps. In the first step, we take simple statistical evidence seriously. We test the univariate and bivariate properties of the short-term nominal interest rates and real output. We provide a series of cointegration tests based on univariate statistical properties of short term interest rates. Cointegration tests based on Johansen s maximum likelihood procedure impose minimal auxiliary assumptions to account for long term relationships. However, here we interpret our results with caution due to tensions between economic theory and the univariate statistical features of short term nominal interest rates. In the second step, we take the critique from economic theory seriously and implement the Pesaran et al. (00) bounds tests. These bounds tests for long run level relationships do not require non-stationarity of short-term interest rates and, therefore, are economic theory consistent. Once we establish whether or not there is long-run information content about output in the interest rate series, we address the issue of whether these findings are the result of the omission of important variables that explain the long-run evolution of output. If we believe that long-run real output is determined by a set of real variables including technology, energy prices, factor inputs, etc., then the omission of these For the rational expectations critique, see for example Sargent (97), Sargent and Wallace (97), Lucas (99) and King and Watson (997). See also Lin (00) for a recent survey of the issue. For a discussion of the information variable approach see for example Friedman and Kuttner (99). For recent evidence on the debate of interest rate stationarity see, for instance, Wu and Zhang (99) and Wu and Chen (00). variables could lead us to finding no common trend between output and interest rates just because these real variables are needed to complete the long-run stationary combination. Our results show that both statistics and economic theory based evidence largely rejects the existence of long term relationships between the relevant policy indicators and real output. The absence of long run relationships between short-term interest rates and real output is not only a full sample result, but also valid in most of the subsamples in the post Second World War period and are robust to the inclusion of possible omitted real variables. One can interpret these findings as evidence of support for the long-term policy neutrality hypothesis. The paper is organized as follows. In Section, we present the data. Section discusses the choice of monetary indicator. Section presents univariate time series properties of the variables before conducting long-term tests. In Section we conduct long-term tests based on statistical evidence. We present cointegration results with a particular emphasis on sub-sample stability. In Section we implement economic theory consistent bounds tests with particular emphasis on sub-sample stability. Section 7 addresses the omitted variables problem. Section concludes.. Data The annual data for the U.K. covers the period We will study real output represented by real GNP. This data was obtained from the study of Hendry (00) [http://www.nuff.ox.ac.uk/users/hendry/]. This study stops in 99 and hence, from this year onwards we update the data using OECD s Main Economic Indicators and IMF s International Financial Statistics database (IFS). We use the Treasury Bill rate as the short term interest rate measure and 0-years Government Bond yield as long term interest rate as reported by Hendry (00). In the case of the U.S. data on output and the Treasury Bill Rate is obtained from the U.S. Federal Reserve. Treasury Bill Rates have missing observations during the end of the 90s and beginning of WWII, so we could only start in 9. We also use two long-term interest rates such as the 0-year Government Bond Rate and Moody s AAA Yield Index starting from 99. As a cross check of our annual data results we also carried out our tests using quarterly data from 90: to 00:. In this case we used as short term rates the Treasury Bill rate for both UK and US and also the Federal Funds Rate for the US. This quarterly data comes from IFS, OECD and the statistics provided by the U.S. Federal Reserve Board (FRB). We report the quarterly data results whenever they yielded substantially different results from the annual data. Figure plots the annual data on the Treasury Bill rate and the log of GDP for the US and UK. The main feature that arises from both plots is the
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks