The Academy of Economic Studies Bucharest Doctoral School of Banking and Finance

The Analysis of the Monetary Policy Stance In Romania Using Monetary Conditions Index (MCI). The Case of

Managed Floating Under MCI Targeting

The Academy of Economic Studies The Academy of Economic Studies BucharestBucharest

Doctoral School of Banking and FinanceDoctoral School of Banking and Finance

DISSERTATION PAPER

MSc. Student: Cristian SabouSupervisor: Prof. Moisa Altar

Bucharest, July 10, 2001Bucharest, July 10, 2001

• Objectives

•Models of the MCI in a small open economy

•Econometrical estimation and parameter determination•Conclusions

Contents

•Concepts rewiew

•A model of MCI in a managed floating fx regime

The characteristics of monetary policy stance in Romania

The influence of the FX managed float regime of Romaniaover the real monetary policy stance in the economy

The variables used by the NBR in conducting its monetary policy

The monetary policy of the NBR: rules vs. discretionarity

The operational targets of the NBR: the interest rate, the exchange rate of ROL and the monetary base

Can the Monetary Conditions Index be calculated for Romania?

The information provided by the MCI for the monetary policy in

Romania

Derivation of the MCI in an Open Economy

A simple model of an open economy

111 )( tttttt eey

11 ttttt yery

111 ttt re

)()1( 1 ttttt ebaywrewMCI

From this model it is derived (Ball,1997) the MCI:

Through minimization of the loss function, Gerlach and Smetz obtain:

22

21 )()~( T

tSttt yL

)(/)()(/1 111T

tTS

tDt

optt EMCI

tT

tttoptt ybaryfr ~)(ˆ)~,(

tT

tttoptt ybaICMyfMCI ~)(ˆ)~,(

The case of managed floating in determiningthe MCI

dtttt ery 210

dtttttt eeEry 31210

ttt eeE 1

dttttt eery 3210

dttttt eeerry )ˆ()ˆ( ~

321

)()( 03201 eeerrMCI ttt

ttopt

t rICMi

)(1

1 *

**)(1

1ttt

optt rICMs

A monetary policy framework for small open economies using MCI

The empiric testing of the Monetary Conditions Index (MCI) for a managed floating currency regime

)()( 03201 eeerrMCI ttt

We will estimate the structural parameters taking into account the IS curve in the same model

dttttt eery 3210

We use the industrial production as a proxy variable for the output

Name Description

L_CS Fx nominal rate for USD/ROL in log

L_CS_REAL Fx real rate for USD/ROL in log

DOB_REALA The real interest rate for corporates

L_PROD_FX Industrial production index with fixed base

ianuarie 1997 in logaritm

L_USD_FX Real fx rate of USD/ROL with fixed base as of

1/97 in log

L_BSK_FX Real fx rate of BASKET /ROL with fixed base

as of 1/97 in log

Variables

L_CS L_CS_REAL DOB_REALA L_PROD_FX L_USD_FX L_BSK_FXconstantăşi trend

Constantă Constantăconstantă şi

trendconstantă

constantă

Lag 1 1 1 1 1 1

t statistic -3,1072 -5,4095 -4,4941 -2,9820 -4,6777 -6,48461% -4,1458 -3,5778 -3,5778 -4,1678 -3,5778 -3,57785% -3,4987 -2,9256 -2,9256 -3,5088 -2,9256 -2,9256

ADF test

The only nonstationary series are L_CS and L_PROD_FX

For the order of integration, we determine the first difference:

ΔL_CS ΔL_PROD_FXConstant

ăConstantă

Lag 1 1

t statistic -11,7830 -6,64101% -3,5653 -3,58145% -2,9202 -2,9271

We find that the L_CS and L_PROD_F are first order integrated, so the OLS or VAR are not appropiate. I will use a partial adjustment model.

Structural parameter estimation taking into account the USD/ROL fx rate

L_PROD_FX = –0.0167 + 0.7880·L_PROD_FX(-1) + 0.1030·L_USD_FX(-3) +

(-0.572) (8.670) (1,205)

0.4477·ΔL_USD_FX –0.621·DOB_REALA+ -0.106·DM9803 -

(1,194) (-1,081) (1,837)

0.123·SEAS(12)

(-4,131)

The residual graph:

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

97:07 98:01 98:07 99:01 99:07 00:01 00:07

According to the Jarque - Bera test, the errors are normally distributed

0

2

4

6

8

10

12

14

-0.10 -0.05 0.00 0.05 0.10

Series: ResidualsSample 1997:03 2000:12Observations 46

Mean -1.61E-17Median 0.004532Maximum 0.122594Minimum -0.110723Std. Dev. 0.051092Skewness -0.112420Kurtosis 3.285504

Jarque-Bera 0.253126Probability 0.881119

R-squared 0.778199 Mean dependent var -0.235604Adjusted R-squared 0.744076 S.D. dependent var 0.108486S.E. of regression 0.054882 Akaike info criterion -2.828000Sum squared resid 0.117469 Schwarz criterion -2.549728Log likelihood 72.04399 F-statistic 22.80559Durbin-Watson stat 2.244708 Prob(F-statistic) 0.000000

Other tests:

So, in this case we have obtained:

1030.04477.0,621.0 321 si

The structural parameter estimation taking into account the Basket / ROL real FX rate

L_PROD_FX = –0.003 + 0.8045·L_PROD_FX(-1) + 0.1060·L_BSK_FX(-3) +

(-0.101) (8.920) (1,189)

0.617·ΔL_BSK_FX –0.6740·DOB_REALA+ -0.112·DM9803 -

(1,579) (-1,178) (1,941)

0.1259·SEAS(12)

(-4,237)

The residual graph looks like:

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

97:07 98:01 98:07 99:01 99:07 00:01 00:07

Jarque - Bera test shows that the errors are normally distributed

0

4

8

12

16

-0.10 -0.05 0.00 0.05 0.10

Series: ResidualsSample 1997:04 2000:12Observations 45

Mean -2.37E-17Median 0.001646Maximum 0.121494Minimum -0.116521Std. Dev. 0.050540Skewness -0.144529Kurtosis 3.432204

Jarque-Bera 0.506915Probability 0.776113

Other tests are:

R-squared 0.755546 Mean dependent var -0.241414Adjusted R-squared 0.716948 S.D. dependent var 0.102220S.E. of regression 0.054384 Akaike info criterion -2.843464Sum squared resid 0.112389 Schwarz criterion -2.562428Log likelihood 70.97795 F-statistic 19.57472Durbin-Watson stat 2.254033 Prob(F-statistic) 0.000000

So, in this case we have obtained:

1060.06170.0,6740.0 321 si

EVOLUTION OF THE MCI, WITH BASIS PERIOD AS OF JANUARY 1997

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

TIMP

ICM

MCI_USD MCI_BASKET

Other factors that influence the choice of the foreign exchange rate regime in Romania

•the manner in which inflationary expectations are formed

•the currency structure of the external andd internal debt

•the stage of the process of the integration in U.E.

Conclusions

• the stance of the monetary policy was much more restrictive in the period of 1997 - 2000, than NBR aimed at

• if a managed floating regime is not taking into consideration the currency of the main trading partners, this would cause high output variability

•in the near future, Romania will have to redirect its fx regime towards the euro (starting with the targeting the EUR/ROL fx rate within still a managed floating currency regime)