Figure 3 and figure 4 give the evolution of expenditure shares in BEAC and BCEAO. S1, S2 and S3 are the weighted coefficient on currency, demand deposits and quasi money respectively. The expenditure shares are within 0 and 1 meaning that they have coherent economic values. However, the weighted coefficient on currency is the highest one. This confirms the fact that as an asset is closer to the pure money the expenditure share is close to one. The evolution of S3 confirms equally the theoretical predictions. In fact, the expenditure share is close to zero as we drive away from pure money. Figure 3 and 4 show that the weighted varies over time. In BEAC, we observe that after 2001Q1 the expenditure share on demand deposits becomes greater than the expenditure share on quasi money. The explanation behind this feature can be the increasing proportion of demand deposits in the total assets relatively to quasi money. Table 1 summarizes the descriptive statistics of simple sum monetary aggregates and Divisia monetary aggregates. The results show that the evolution of simple sum and Divisia monetary aggregates in BEAC and BCEAO is similar since the values are closed. This highlights the fact that there can be a little difference between the two types of monetary aggregates. Bank

The graphical evolution of the monetary aggregates confirms this point (see figures 5, 6, 7 and 8). Growth differences between the two types of monetary aggregates are plotted in figures 9, 10, 11 and 12. We observe that the Divisia monetary aggregates are more volatile than their simple sum counterparts in BEAC. The same remark can be made for BCEAO. The computation of standard deviation of the growth rate of each monetary aggregate gives some insight to this observation. For example, standard deviations for growth rate in BEAC are: 5.28 for tm1, 8.05 for tm1d, 3.82 for tm2 and 4.54 for tm2d. Overall, the evolution of the growth rates of monetary aggregates in BEAC shows some differences. In this sense, the correlation between growth rates tm1 and tm1d is 0.7749 and between tm2 and tm2d is 0.8618. Ishida and Nakamura reported a diverse result. The reason is that despite the financial liberalization, the financial assets in m2 are still less. In BCEAO, the correlations are respectively 0.9054 and 0.9820 between tm1 and tm1d, and tm2 and tm2d meaning there is no difference in the evolution of the growth rates. This is justified by the fact that within the period, the evolution of different monetary assets shows the same path.

*Figure 3: The evolution of expenditure shares in BEAC*

*Figure 4: The evolution of expenditure shares in BCEAO*

*Figure 5: Simple Sum M1 and Divisia M1, BEAC*

*Figure 6: Simple Sum M2 and Divisia M2, BEAC*

*Figure 7: Simple Sum M1 and Divisia M1, BCEAO*

*Figure 8: Simple Sum M2 and Divisia M2, BCEAO*

*Figure 9: Growth rate of simple sum (tm1) and Divisia (tm1d), BEAC*

*Figure 10: Growth rate of simple sum (tm2) and Divisia (tm2d), BEAC*

*Figure 11: Growth rate of simple sum (m1) and Divisia (m1d), BCEAO*

*Figure 12: Growth rate of simple sum (m2) and Divisia (m2d), BCEAO*

**Table 1: Summary statistic of monetary aggregates**

Central Bank | Variables | Mean | Standard deviation |

BEAC | M1 | 7.167888 | 0.455495 |

M1D | 6.237531 | 0.511884 | |

M2 | 7.588747 | 0.421116 | |

M2D | 6.471139 | 0.410794 | |

BCEAO | M1 | 7.845084 | 0.434587 |

M1D | 6.984210 | 0.459682 | |

M2 | 8.261289 | 0.401002 | |

M2D | 7.158885 | 0.413497 |

Figure 3 and figure 4 give the evolution of expenditure shares in BEAC and BCEAO. S1, S2 and S3 are the weighted coefficient on currency, demand deposits and quasi money respectively. The expenditure shares are within 0 and 1 meaning that they have coherent economic values. However, the weighted coefficient on currency is the highest one. This confirms the fact that as an asset is closer to the pure money the expenditure share is close to one. The evolution of S3 confirms equally the theoretical predictions. In fact, the expenditure share is close to zero as we drive away from pure money. Figure 3 and 4 show that the weighted varies over time. In BEAC, we observe that after 2001Q1