# Monetary Conditions Index for Kenya: Empirical Results

The augmented Dickey Fuller (ADF), and Kwiatkowski, Phillips, Schmidt and Shin (KPSS) are used to test for the presence of unit roots. Results in the following table show the rejection of the null hypothesis at first difference meaning that the series are all I. Having established that the variables were integrated of the same order, a cointegration test was necessary to determine if the variables had a long run steady state using the Engel Granger Test. Equation was first estimated and then the residuals generated tested for unit roots. The unit roots tests done by KPSS showed that the variables were cointegrated since the computed test statistic was 0.182 compared to the critical values of 0.463 at 5% and 0.739 at 1%.

The weight of e (or 0.9766) divided by the weight of r (or 0.0234), or vice versa, yields the MCI ratio. In this case, the MCI was 41.735 which was the equivalent of 0.024 when expressed as the weight of the interest rate (r) over the exchange rate (e). This result therefore suggests that a one percent point rise in the interest rate has the same effect on GDP as a 0.024 percent rise in the exchange rate (depreciation). Conversely, a percentage point depreciation has the same effect on the MCI as the 41.735 percent point increase in the real interest rate, all other factors held constant.

The interest rate has a lower weight compared to the exchange rate meaning that a percentage point change in interest rate has a lesser effect on output than a one percent change in the exchange rate. This finding is consistent with Peng‘s finding that the estimates for small open economies produce smaller ratios of the real interest rates compared to the exchange rates.

Whether or not the estimated MCI fits the monetary policies of the CBK in the period can be seen through a visual inspection of the Figure 3 to see whether CBK responded to prevailing economic times in the pattern described by the MCI. This is shown by plotting the 91-day Treasury bill rate, the MCI and the real exchange rate on the same axis. The MCI tracks the changes in the interest rate quite effectively as well as the exchange rate.

The MCI tracks the interest rates really well. Despite Kenya maintaining a floating exchange rate, the MCI shows relative stability in the period covered. Since a rise in the MCI should reflect a tightening in monetary conditions and vice versa, Figure 3 shows that the central bank has pursued a tightening of monetary policy stance since a trend line generated on the MCI generates a negative slope.

**Table 1: Unit Root Tests**

ADF | KPSS | |||

Level | First difference | Level | First difference | |

^{y} |
0.12 | -8.51 | 0.84 | 0.08 |

r | -3.20 | -4.23 | 0.22 | 0.08 |

e | -2.35 | -5.67 | 0.12 | 0.14 |

At 1% | -2.62 | -2.62 | 0.74 | 0.74 |

At 5% | -1. 95 | -1.95 | 0.46 | 0.46 |

*Figure 3: Computed MCI and Contributions from Its Components*