MODELING THE DETERMINANTS OF THE DEMAND FOR MONEYIN RWANDA

Abstract

This study models the demand for money in Rwanda and its determinants for the period 1996
to 2013 using the autoregressive distributed lag (ARDL) framework. It investigates the
relationship between monetary aggregates M2 and M3, and real final consumption expenditure,
real gross capital formation and real exports, inflation rate, interest rate, exchange rate and
financial innovation. The study uses disaggregated components of GDP as scale variables to
separate their effects on money demand. Furthermore, the study investigates the stability of the
money demand function in Rwanda using the CUSUM and CUSUM Squared tests. The
calculated F-statistic for the M2 model is 4.835 which is above the upper bounds of the critical
values of the bounds test at all levels of significance while the F-statistic for the M3 model is
3.885 and above the upper bounds at 5% significance level. These empirical results prove the
existence of a cointegrated relationship between M2 (M3) and the explanatory variables real
final consumption expenditure, inflation, exchange rate and financial innovation. The study
finds that the long-run elasticities of real final consumption expenditure, inflation, financial
innovation and the exchange rate are positive. The coefficient of 1.028 for final consumption
expenditure is highly elastic and shows that final consumption expenditure is a key driver of
the demand for M2. On the other hand, an exchange rate long-run elasticity of 1.068 indicates
that a shock of Rwanda’s exchange rate has a more than proportionate impact on the demand
for M3. A positive exchange rate coefficient shows the dominance of the wealth effect over the
substitution effect in situations of exchange rate depreciation. A positive long-run semi- elasticity of the inflation rate proves the importance of the transactions motive to economic
agents’ portfolio decisions. Financial innovation has not yet offered Rwandans alternative
financial assets as indicated by its positive influence on money demand. The results of the
CUSUM and CUSUMSQ tests reveal that the M2 and M3 money demand functions are stable
between 1996 and 2013. Thus, M2 is still an appropriate target for monetary policy in Rwanda
while M3 can also be targeted with equivalent effectiveness. Complimentary fiscal policies that
focus on final consumption expenditure are needed since final consumption expenditure is a
key driver of money demand in Rwanda.