Money Demand Function Revisited: the Role of Asset Prices in Pakistan

• Author: HINA SHAFIQ AND WASIM SHAHID MALIK

Abstract

Money demand function has always been an important constituent of the macroeconomic models and policy making. A stable money demand function is crucial for the efficient conduct of monetary policy which enables the policy makers in forecasting, policy analysis and choosing a nominal anchor. From 1980s onwards, questions were raised on the stability of money demand function and varying explanations were provided for the instability both internationally and in Pakistan, but the research in this area never remained conclusive. However, this study is the first attempt to prove that money demand function in Pakistan is mis-specified without incorporating asset prices in its estimation.

The study, therefore, contributes to the existing literature on the issue of stable money demand function in Pakistan by using asset price index to explain the money demand in a multivariate regression model and VECM for the time period 1981Q1-2017Q2. The study concludes that asset price index plays a significant role in explaining variation in money demand and the relationship is found positive. Moreover, results of the individual asset prices show that prices of two assets, housing and share prices, via positive wealth effect, lead to an increase in the demand for money balances in Pakistan. These results imply that previous studies finding instability of money demand function are mis-specified without asset prices

Keywords: Money demand function, Monetary policy, Asset prices

1. INTRODUCTION

   State Bank of Pakistan (SBP) used monetary aggregate as policy target for the conduct of monetary policy in Pakistan till 2008, whereby the narrow monetary aggregate M0 was used as instrument to target broad monetary aggregate such as M2 – used as a nominal anchor to achieve the objective of price stability and output growth. However, from 2009 onwards, SBP adopted interest rate targeting by introducing an interest rate corridor and using overnight money market repo rate as its operational target. This regime switch was due to the apparent unstable relationship between inflation and money supply and an unstable money demand function because of the structural changes in the economy and the financial sector of the country, fiscal pressure, financial innovation and increasing use of technology [Hanif, 2014; Omer and Saqib, 2009; Moinuddin, 2009; MPS-SBP, 2009].

   The available literature on interest rate versus monetary targeting in Pakistan presents differing views on the adoption of a particular strategy. Many studies found a positive and significant relationship between money supply and inflation in Pakistan [Hanif and Hayat, 2016; Farooq, Hassan, Shahid, 2015; Qayyum, 2006, Kemal, 2006] and found stable base and broad money velocities (Omer, 2010). Moreover, it is also suggested to consider money in the models of the Pakistan economy because all the monetary aggregates are strongly pro-cyclical and a few of them even serve as the leading indicators of economic activity and also cash based models along with money growth rule perform better in data matching than the cashless economy models with Taylor type rule (Ahmed et al, 2016). This line of research provides evidence for the use of monetary aggregates targeting for Pakistan to contain inflation but the excessive focus on the interest rate targeting has reduced the role of monetary targeting.

   However, it has also been proved that quantity theory of money is not applicable in Pakistan along with unstable income velocity of money, which casts doubt on the use of monetary aggregates targeting (Omer and Saqib, 2009).

   The monetary aggregate targeting also received criticism due to the upsurge in inflation in Pakistan along with the success stories of inflation targeting regimes followed by various countries. The stable growth in monetary aggregates led to instability in the price behavior due to the shocks to money demand and an unstable transmission mechanism. This led policy makers to resort to fine tune the aggregate demand through interest rate management (Felipe, 2009). The instability of the money demand function also led policy makers to prefer inflation targeting over monetary targeting. An inflation targeting monetary policy requires commitment to price stability, independent and transparent central bank, flexible exchange rate and fiscal discipline.

   This debate over inflation targeting led to the research in Pakistan that focused on the argument for and against the adoption of inflation targeting in Pakistan [see for instance, Moinuddin (2009), Felipe (2009), Akbari and Rankaduwa (2006), Khalid (2006), Khan and Schimmelpfennig (2006), Chaudhry and Choudhary (2006)].

   A stable money demand function is crucial for the efficient conduct of an appropriate monetary policy which enables the policy makers in forecasting and policy analysis. The accurate forecast of money demands helps in determining the optimum growth rate of money supply which is important in controlling inflation rate in the economy and also because money supply can have predictable effects on real economic variables. Moreover, a stable money demand function is important for monetary and fiscal policies to have predictable effects on the macroeconomic variables.

   The existing literature has identified various sources of instability in money demand function such as structural changes in the economy, financial innovation and deregulation, unexpected changes in the income velocity of money, changes in the relationship between money stock and other fundamental macroeconomic variables such as income and interest rate, discrepancies between money supply and money demand etc. Many studies have worked on the estimation of money demand function and its determinants in Pakistan such as Rehman and Afzal (2003), Qayyum (2005), Khan and Sajid, (2005), and Hye et al., (2009). The studies specifically focusing on the stability of money demand function in Pakistan proved its stability [Sarwar (2010); Omer (2010)], or concluded otherwise (Moinuddin, 2009).

   This study contributes to the existing literature in Pakistan on the issue of stability of money demand function by explaining it with the inclusion of asset prices in the money demand function as the existing models of the money demand are mis-specified in this respect. The inclusion of asset prices solves the problem of unstable money demand function and improves the stability and economic properties of the function as well (Borio, Kennedy and Prowse, 1994). This aspect of including asset prices in the money demand function can be understood from the Keynesian liquidity preference theory in which money is also demanded to invest in interest bearing bonds to earn profit through speculation. However, interest rate is not a true indicator of assets such as stock prices or real estate as it is an indicator for bank deposits and T-bill rates. Therefore, the need is to bring asset prices directly into the money demand function to prove its stability properties.

   With this context, the objective of this study is to participate in the debate on (un)stable money demand function by estimating the money demand function for Pakistan with the inclusion of asset price indices. The present study also aims to analyze the effect of separate asset prices such as house prices, stock prices and exchange rate on the demand for money. The empirical evidence we tried to find is important for Pakistan economy where rent-seeking activities have higher weight in overall economic activity compared to that of entrepreneurial activity.

   To accomplish this objectives, we have used multivariate regression analysis and vector error correction model for the estimation of the role that asset price indices play in explaining the money demand function for the time period 1981-Q1 to 2017-Q2. It has been found that the asset prices play a significant role in explaining the money demand function in Pakistan and the relationship is found positive. Moreover, an increase in the housing and share prices via positive wealth effect leads to an increase in the demand for money balances in Pakistan.

   Remainder of the study proceeds as follows: In section 2 we discuss the review of existing literature on the topic. Section 3 discusses the econometric methodology used for the estimations and in section 4 the results of the study are discussed. Section 5 concludes the study.

2. LITERATURE REVIEW

   THEORIES OF MONEY DEMAND

   There are five major approaches to theorize money demand function. First, the classical economists, who never explicitly formulated a demand for money theory, stressed on ‘transactions velocity of circulation of money’ in the Fisher’s (1911) Quantity theory of money and equation of exchange i.e. MV= PT. The equation of exchange states that the total amount of money in circulation in the economy equals the value of transactions undertaken in the economy. Money performs the medium of exchange function only and facilitates the exchange of goods and services. Moreover, transactions demand for money is determined by the full employment level of income. The strict version of the quantity theory has certain interrelated propositions. First is the exact proportionate relationship between the money stock and price level for which the stability of the money demand functions is imperative.

   The second proposition is that the changes in money supply cause changes in the price level, that is, the causation runs from money (M) to price level (P). Third proposition is of the neutrality of money, which implies that changes in the money supply have no effect on real variables such as output and employment. Fourth proposition states that money is the only determinant of the general price level in the economy and the price level is not affected by the non-monetary factors. Fifth, the money supply is exogenous and it is not determined by the changes in demand for money (Humphrey, 1974).

   Second, Keynes (1936) formulated the...