# Effect Of Policy Rate On Market Deposit Rate Finance Essay

This paper examines the deposit side inter rate pass-through mechanism of interest rate channel for Monetary Policy Transmission Mechanism (MPTM) in Pakistan. The variables used in the research are T-Bill 6 month (TB6M) being the proxy of policy rate and the Weighted Average Deposit Rate (WADR) being the benchmark of return on deposits given to depositors by the banks. The data on variables of research is of five years, from 2005:07 to 2010:06, obtained from State Bank of Pakistan’s website. The widely accepted VAR model for explaining the MPTM is taken on the basis of literature and its flexibility in explaining the phenomena.

Results reveal that there is an insignificant effect of policy rate shock on deposit rates, showing the weak structure of interest rate channel on deposit side. VAR’s reveal that that the effect of a policy rate proxy, TB6M, on WADR in Pakistan is insignificant. It is comprehended that in short-term one percentage point change in TB6M brings only five basis points change in WADR, which takes a period of four months approximately. While the peak is achieved at approximately eight basis points taking almost nine months of the time period, after which WADR starts to fall very slowly and converge to its long-term level.

It is also comprehended that the biggest monetary policy shock observed during 2005:07 to 2010:06, brought a change of approximately eighteen basis points which peaks at nine months time period after the shock and after which WADR starts to fall very slowly and in order to converge to its long-term level.

## 1. INTRODUCTION

In any country there are three main economic policies, namely Trade Policy, Fiscal Policy and Monetary Policy. Trade policy refers to the changes in import and export taxes, subsidies and restrictions enhancement Net Exports portion of a country’s economic activity. Stanlake and Grant (2000) defines fiscal policy as the intended changes or influence in government (public) expenditure and income so as to achieve desired economic and social objectives, using government spending and taxation as the tools of fiscal policy. Parkin (2005) defines monetary policy as the changes in interest rate and quantity of money in the economy, which is prepared by the central bank of the country.

In Pakistan, mostly following monetary policy tools have been used by the State Bank of Pakistan (SBP):

Open Market Operations (OMO’s).

Statutory Liquidity Reserve Requirements (SLR).

Cash Reserve Requirements (CRR).

L/C Margins.

Discount Rate or Policy Rate Changes.

Export Refinancing Rate.

Moral Suasion.

Monetary Policy Transmission Mechanism (MPTM) refers to the passing of monetary policy decisions and the time it takes to effect the economy for maintaining and improving the financial health of the monetary system of economy. Al-Mashat and Billmeier (2007) briefly explains that MPTM includes following six channels:

Interest Rate Channel

Exchange Rate Channel

Asset Price Channel

Bank Lending Channel

Balance Sheet Channel

Expectations Channel

This paper is aimed at examining one half of the inter rate pass-through of interest rate channel of the MPTM, i.e. the deposits side of the interest rate pass-through mechanism, for understanding the shock of policy rate being reflected in the financial market deposit rate. However due to the time, data and scope limitation of the research project, results do not examine the whole interest rate channel and the robustness of the model is missing which requires inclusion of variables like Money Circulation benchmark Reserve Money (M2) etc.

To explain the phenomena vector autoregression (VAR) is used, as it is most widely used for explaining MPTM in literature as well as because of its flexibility in explaining the concept. The number of lags is selected by the Akaike Information Criterion, Schwarz Information Criterion and Hannan-Quinn Information Criterion as well as the subjective approach of correlogram, suggesting the use of first lag of VAR model.

Data taken for empirical analysis is for a period of five years, from 2005:07 to 2010:06 on T-Bill 6 month (TB6M), as a proxy for Policy Rate or Discount Rate, and Weighted Average Deposit Rate (WADR) for explaining the interest rate pass-through of MPTM in Pakistan. As the data of WADR is available on monthly basis only, so we have taken average of TB6M observations to bring in the homogeneity of data observations.

Monetary policy can be used to effectively promote growth, raise levels of employment, and stabilize price and currency levels in an economy. One of the main instruments for regulating the economy by any central bank has been discount rate or policy rate, which effects a change in financial market lending and deposit rates to increase or decrease the credit demand in the economy. In Pakistan, the policy rate is determined independently by the SBP Board of Governors, taking in to account the macroeconomic situation of the country.

Mishkin (1996) provides an overview of the monetary transmission mechanism channels, starting with the traditional interest rate channel that a contractionary monetary policy increases the discount rate, which increases the market short-term nominal interest rates, also increases the short-term real interest rates, which in turn increases the cost of capital, causing a decline in investment spending, thereby reducing aggregate demand and decreasing the output.

Lodhi et al. (2005) demonstrates an empirical investigation on monetary policy transmission mechanism of Pakistan using auto-regressive distributed lag econometric technique on T-Bill 6 month (TB6M) cut-off rate as benchmark of policy rate, weighted average lending rate (WALR) and weighted average deposit rate (WADR) determined by the financial markets forces of Pakistan, using data for a period of seven years from 1999:07 to 2006:06 on the variables of study. The found that 1% increase in TB6M tends to increase almost 20 bps in WALR in the short-term, whereas, in the long-term it takes about 5 months to completely transmit to the bank lending rate. WADR is considerably less affected, and a 1% increase in T-bills (6month) tends to increase WADR by 44 bps and takes 10.1 months in the long-term.

Agha et al (2005) believed that the following four channels will explain a significant part of the effects of a monetary policy shock on output and prices of Pakistan, namely Interest Rate Channel, Asset Price Channel, Exchange Rate Channel and Credit Channel. They used vector autoregressions (VAR) to examine the monetary transmission mechanism in Pakistan. The results indicated that monetary tightening leads first to a fall in domestic demand, primarily investment demand financed by bank lending, which translates into a gradual reduction in price pressures that eventually reduces the overall price level with a significant lag. In addition to the traditional interest rate channel, the results also pointed to a transmission mechanism in which banks play an important role.

Poddar et al. (2006) examines the monetary transmission in Jordan using vector autoregressive approach. They found that the real deposit rates are more responsive to the monetary policy change as compared to the real lending rates. They used auctioned real CD (3month) rate, a benchmark of monetary policy shock by the Central Bank of Jordan. For empirical findings they took into consideration a data on variables of research from 1995:12 to 2005:02. The result confirmed that a 1% change in real 3-month CD rate changes real deposit rate by 92 bps, while the real lending rate changes by 74 bps, showing a strong significant effect of policy rate on auctioned real CD (3month) rate.

Khawaja (2007) has examined the determinants of interest spread in Pakistan, and identified one of the reasons of the inelasticity of deposit supply to the banks being the absence of alternate options for savers in Pakistan. This explains that greater the inelasticity of the deposits the less compelled a bank would be to pass on the increase in T-bill rate, which is directly related to the policy rate. Current Accounts, Saving Accounts and other accounts are inelastic / interest insensitive, which constituted as much as 81% of the total industry deposits in 2005. Fixed deposits accounts, on the other hand, are elastic / interest sensitive, as they increase with an increase in T-bill rate.

Fuertes et al. (2008) examined interest rate transmission mechanism of UK deposit and credit products, to find the adjustment speed of various UK bank (retail) rates in the short run as well as in the long run. For their research, they have used conventional nonlinear and linear error correction model (ECM) over the period 1993:01 to 2005:06. They found that the adjustment speed varies over time proportionally to the size of policy rate change. They also found that the adjustment in the deposit rates tends to be faster during monetary expansion, while mortgage rates tends to be more rapid during monetary contraction.

## 1.2 Problem Statement:

It has been observed from the literature on monetary policy that most central banks have identified a positive significant relationship between policy rate and the market retail rates. While, in Pakistan there have been several structural changes by SBP to improve the monetary policy transmission mechanism, specially during last 5 years, because of heavy government borrowings, global financial turmoil and rising inflation.

## 1.3 Research Objective:

Main purpose is to identify whether there is a significant impact on the financial market weighted average deposit rate (WADR) in response to a policy rate shock in Pakistan.

## 1.4 Hypothesis:

H0 : Significant Impact of Policy Rate on the Market Deposit Rate does not exists in Pakistan.

## 2. METHODOLOGY

## 2.1 Choice of Variables:

We have taken the T-Bill 6 month (TB6M) as a benchmark of Policy Rate or Discount Rate, because the policy rate remains fixed for a certain time period which at times is for more than a year as well. Therefore, to improve the variability of the endogenous variable, we have taken TB6M as an endogenous variable in our research. Also we have been able to find several evidence of taking short-term interest rates as a representative of policy rate, which has received high eminence in recent published papers on MPTM.

Data used for TB6M is available on the of cut-off yield derived through Open Market Operations (OMO’s) conducted by SBP for absorbing the excess liquidity in the financial markets to control money circulation, also known as the REPO Rate in Pakistan.

As recommended by the theory that exogenous variable to be the deposit rates, which is the return paid by the banks to their depositors on different types of deposits. We have taken Weighted Average Deposit Rate (WADR) for all banks in Pakistan, including zero mark-up deposits, which is published by the State Bank of Pakistan (SBP), on its website on monthly basis.

The formula for calculating WADR is given as below:

## Source: State Bank of Pakistan

## 2.2 Data:

Due to the limited time period for the completion of this research project, the data is taken for a period of five years, which took time in collection and adjustments, from 2005:07 to 2010:06 (obtained from SBP’s website). As literature suggests observations or the data for a period of eight to fifteen years to be taken in most cases for explaining the long-term effect of policy rate shock on market rates in explaining the interest rate pass-through of MPTM, so a period of five years is considered a bit low in explaining the concept. Still, the number of observations it is tested using Classical Linear Regression Model (CLRM), by the difference between R-squared and Adjusted R-Squared, which is quite low, showing that the number of observations (sixty in total) used do suffice the requirement1. Although for the estimation/ forecasting purpose higher number of observations is better.

## ________________________

1 See Appendix: Classical Linear Regression Model Summary for confirming the data of 2005:07 to 2010:10 being appropriate for this project.

## 2.2.1 Adjustments in Data:

The data on WADR is available on monthly basis, while for TB6M, the data is available on every auction or OMO conducted by the SBP. So in order to bring in homogeneity of data among variables of research, the data for TB6M is adjusted by taking simple average of TB6M cut-off yields in the month they were conducted, i.e. if there were three OMO’s in a month then the sum of yields is divided by three to calculate the average yield for that month.

## 2.3 Conceptual Framework:

It can be deduced that it creates an incentive for banks and financial intermediaries to invest more into almost no risk short-term government securities rather than lending to the medium to high risk private sector during a contractionary monetary policy which increase the short-term nominal and real interest rates, reducing the demand for investment spending, creating short-term liquidity shortage in the banking system, leading banks to increase quantum of deposits by offering higher offered deposit rate to the depositors, giving incentive for depositors to enhance banking deposits leading towards reduced money circulation in the economy for spending.

M ↓ DR ↑ TB ↑ ↓ I ↑ WADR ↑ BD ↓ MC, Ceteris paribus

where, M↓ indicates contractionary monetary policy, DR↑ indicates increase in discount rate, TB↑ represents increase in T-Bills rate, ↓I indicates decrease in investment demand, ↑WADR represents increase in weighted average deposit rate, ↑BD indicates increase in bank deposits and ↓MC represents decrease in money circulation in economy for spending.

It is evident to note that in VAR model, all variables are considered as endogenous variables of research.

## 2.4 Software Package:

As most widely used software package for financial econometrics is EViews, so this paper also uses EViews version 5.0 for model selection, selection for number of lags in model, empirical analysis and forecasting etc.

## 2.5 Model Selection:

Vector Autoregression (VAR) model is selected for the empirical analysis, which is selected on the basis of large portion of literature using VAR model. Also some distinct advantages are explained by Gujarati and Sangeetha (2007), Agha et al. (2005), Brooks (2008):

Recognizes explicitly the simultaneity between exogenous and endogenous variables.

VARs are more flexible than univariate AR models.

Focuses on reduced form of relationships between variables.

Explains each endogenous variable as by its lagged values and with the lagged values of other endogenous variables in the model.

Usually in VAR models, there are no exogenous variables – all are endogenous variables.

Sims (1980) developed his VAR model to explain that if the variables are of true simultaneity then they should be given equal treatment.

The order of the model is chosen based on the proven theories, which is mentioned/ derived in the following VAR model equation for the variables of the study:

y1t = β10 + β11 y1t-1 + α11 y2t-1 + u1t equation-(a)

where y1t is the WADR and y2t is the TB6M rate, i.e. the two variables of the bivariate VAR model, β11 and α11 are the parameters for slope of WADR and TB6M respectively, β10 is the coefficient of the model ant u1t being the error term.

The number of lags is selected by using information criterion method, namely using the Akaike Information Criterion, Schwarz Information Criterion and Hannan-Quinn Information Criterion, which comes out to be first lag order or length of VAR model.2

Another approach to the selection of number of lags in econometrics is the subjective correlogram or residuals approach to stationarity identification of time series for time series model. This approach also helps in the selection of econometric time series model as well, using autocorrelation function (ACF) and partial autocorrelation function (PACF) series and graphical representation.

## ________________________

2 See Appendix: Vector Autoregression (VAR) Lag Order / Length Criterion Summary of information Criterion for selection of appropriate number of lag for the model.

## 3. RESULTS

The empirical results of the VAR model defined by the equation-(a) are summarized in the below table:

Vector Autoregression Estimates

Sample (adjusted): 2005M08 2010M06

Included observations: 59 after adjustments

Standard errors in ( ) & t-statistics in [ ]

WADR

TB6M

WADR(-1)

0.880517

0.208116

(0.05304)

(0.14068)

[ 16.6015]

[ 1.47940]

TB6M(-1)

0.079624

0.807391

(0.04286)

(0.11367)

[ 1.85788]

[ 7.10276]

C

-0.208524

1.098633

(0.21668)

(0.57470)

[-0.96237]

[ 1.91165]

R-squared

0.988815

0.950099

Adj. R-squared

0.988416

0.948317

Sum sq. resids

1.468949

10.33387

S.E. equation

0.161961

0.429574

F-statistic

2475.426

533.1125

Log likelihood

25.22585

-32.32511

Akaike AIC

-0.753419

1.197461

Schwarz SC

-0.647781

1.303099

Mean dependent

4.627627

10.33409

S.D. dependent

1.504796

1.889572

Determinant resid covariance (dof adj.)

0.003056

Determinant resid covariance

0.002754

Log likelihood

6.463595

Akaike information criterion

-0.015715

Schwarz criterion

0.195560

## Source: Authors’ Calculation

It can be comprehended from the above table that one percent point increase in TB6M brings almost eight basis points change in WADR, which is showing the insignificance for existence of interest rate pass-through on market deposit rates in Pakistan’s MPTM. Also it is evident to note that due to the strong dependence of WADR value on its own first lag, is explaining approximately eighty eight basis points change in the current WADR value.

The model of VAR is considered to be a good fit as the Adjusted R-squared (adjusted coefficient of determination) value is almost ninety percent, showing the model is explaining almost all the variation in WADR.

The forecasted equation model, based on our data and VAR model equation comparison and inclusion, is as below:

y1t = β10 + β11 y1t-1 + α11 y2t-1 + u1t equation-(a)

y1t = - 0.208524 + 0.880517 y1t-1 + 0.079624 y2t-1 3

## Source: Authors’ Calculation

where y1t is the WADR and y2t is the TB6M rate, i.e. the two variables of the bivariate VAR model, β11 and α11 are the parameters for slope of WADR and TB6M respectively, β10 is the coefficient of the model ant u1t being the error term.

Also, it is worth while identifying that the selection of first lag in the model is confirmed by the Akaike information criterion (AIC) and Schwarz criterion (SC), which are having quite minimal values, i.e. – 0.0157 and 0.1955 respectively.

In order to analyze the time period for passage of monetary policy shock on WADR in Pakistan, below is the impulse response of WADR to Cholesky (degree of freedom adjusted) one standard deviation TB6M innovation:

## _______________________

3 DISCLAIMER: Researchers hold no responsibility for usage and accuracy of forecast or estimation using presented model, what so ever.

## Source: Authors’ Calculation

Impulse response shows that in short-term one percentage point change in TB6M brings only five basis points change in WADR, which takes a period of four months approximately, while the peak is achieved at approximately eight basis points taking almost nine months time period, after which WADR starts to fall very slowly and converge to its long-term level.

It may also be comprehended that the biggest monetary policy shock observed in the data through TB6M during 2005:07 to 2010:06 period, brought a change of approximately eighteen basis points which peaks at nine months time period after the shock and after which WADR starts to fall very slowly and in order to converge to its long-term level.

Variance Decompositions in our VAR model present similar views as of Impulse response, therefore, are not discussed separately. 4

## _______________________

4 See Appendix: Variance Decomposition Percent WADR variance due to TB6M.

## 4. DISCUSSION

As it has been empirically proved that there is an insignificant effect of policy rate shock on market deposit rate in Pakistan, although vast the literature suggests the strong presence of classical interest rate channel in most of the economies. Also this was empirically investigated by Lodhi et al. (2005) that during the time period of 1999:07 to 2006:06, TB6M tends to change WADR by forty four basis points, taking 10.1 months in the long-term. This reveals that the interest rate channel has further lost its significance over deposit side in Pakistan.

Few of the reasons for this further loss of significance on deposit side of interest rate channel for MPTM in Pakistan seems:

SBP’s stance change for promoting economic growth to curbing inflation during the analyzed time period should have brought theoretically a higher significance of policy rate shock, which has not been the case.

Due to the increase in Minimum Capital Requirements (MCR) for the banks has brought several mergers and acquisitions within the banking sector, which might have reduced the competition between banks to attract higher quantum of deposits.

It is evident that there can be a reason of inelasticity of deposits in Pakistan’s banking system, which is shown in below table:

## Source: Table-2 from Khawaja (2007)

## 5. CONCLUSION

## 5.1 Hypothesis:

## Accept H0

Based on the above empirical findings, there is no evidence to reject the null hypothesis; hence it is concluded hat there is an insignificant impact on WADR due to a policy rate (TB6M) shock in Pakistan. This shows the weak effect and presence of interest rate channel of monetary policy transmission mechanism.

## 5.2 Limitations and the Road Ahead:

There are several limitations of this research project, which should be taken into account to further enhance the empirical investigation as well as better suggest the policy implications:

Interest rate pass-through mechanism of MPTM is partially analyzed.

Data for a time period of five years is considered low as compared to similar empirical researches.

Robustness of VAR model is missing to improve the quality of empirical analysis, for which Reserve Money (M2) and some more pertinent variable(s) are recommended to be included in the model.

A structural shock by the State Bank of Pakistan in BPRD Circular No. 07 of 2008, dated 30th May, to place a bottom ceiling rate of five percent to be minimum offered to all savings & PLS saving deposits offered by banks is not adjusted.

Data on WADR is also available on Public Banks and Private Banks as well, which could provide detailed investigation of interest rate pass-through mechanism.

## REFERENCES

Agha I. A. et al. (2005) ‘Transmission Mechanism of Monetary Policy in Pakistan’, State Bank of Pakistan. SBP-Research Bulletin, Volume 1, No.1.

Al-Mashat R and Billmeier A. (2007) ‘The Monetary Transmission Mechanism in Egypt’, International Monetary Fund Working Paper WP/07/285 (December).

Bernanke B. S. and Blinder (1992) ‘The Federal Fund Rate and the Channels of Monetary Transmission’, American Economic Association.

Brooks C. (2008) ‘Introductory Econometrics for Finance’, 2nd Edition, p. 291-292, Cambridge University Press, Cambridge CB2 8RU.

Cook S. (2008) ‘Econometric analysis of interest rate pass-through’, Applied Financial Economics Letters, 4:4, p. 249-251.

Cheng K. C. (2006) ‘A VAR Analysis of Kenya’s Monetary Policy Transmission Mechanism: How Does the Central Bank’s REPO Rate Affect the Economy?’, International Monetary Fund Working Paper WP/06/300 (December).

Fuertes A-M, Heffernan S and Kalotychou E. (2007) ‘Forecasting Foreign Direct Investment Inflow in Jordan: Univariate ARIMA Model’ (ed.), In RES-000-22-0862 project ‘Interest Rate Pass-Through’ funded by ESRC (May).

Gujarati, D. N. and Sangeetha (2007) ‘Basic Econometrics’, 4th Edition, p. 857-858, Tata McGraw Hill Publishing Company Limited.

Horváth B. and Maino R. (2006) ‘Monetary Transmission Mechanisms in Belarus’, International Monetary Fund Working Paper WP/06/246 (November).

Homann B. and Mizen P. (2004) ‘Interest Rate Pas-Through and Monetary Transmission: Evidence from Individual Financial Institutions’ Retail Rates’, Economica, New Series, Vol. 71, No. 281 (February).

Khawaja I. (2007) ‘Determinants of Interest Spread in Pakistan’, Pakistan Institute of Development Economics, Islamabad. PIDE Working Paper 2007:22.

Lodhi M. A. K. et al. (2005) ‘Performance of the Banking Sector’, Pakistan Financial Sector Assessment, Chap. 3: p. 55, State Bank of Pakistan, Research Department.

Mishkin F. S. (1996) ‘The Channels of Monetary Transmission: Lessons for Monetary Policy’, National Bureau of Economic Research Working Paper 5464 (February).

Parkin M. (2005) ‘Economics’, 6th Edition, p. 483, Addison Wesley Publishing Company, New York.

Poddar T., Sab R. and Khachatryan H. (2006) ‘The Monetary Transmission Mechanism in Jordan’, International Monetary Fund Working Paper WP/06/48 (February).

Sims C. A. (1980) ‘Macroeconomics and Reality’, Economica, Vol. 48 (1), p. 1-48.

Stanlake G. and Grant S. (2000) ‘Introductory Economics’. 7th Edition, p.296, Pearson Education Limited, Essex.

### Request Removal

If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal:

Request the removal of this essay