strength. As shown in Lemma 2, the reserve channel reduces both aggregate liquidity risk
and expected liquidity needs. This occurs because the banking system has more reserves in
aggregate through the higher reserve holdings of the merged banks. As stated in Proposition
4, however, the asymmetry channel always increases expected aggregate liquidity needs,
whereas it reduces aggregate liquidity risk only if the relative cost of refinancing is sufficiently
low.
Thus, when the two channels interact, the merger reduces aggregate liquidity risk for a
larger range of parameter values than in Proposition 4, where only the asymmetry channel
is active. Similarly, it increases aggregate liquidity risk in a larger range of parameter values
than in Lemma 2, where only the reserve channel is present.
As for the expected aggregate liquidity needs, the reserve channel dominates when the
asymmetry induced by the merger is sufficiently small. Thus, there is a range of values of the
relative cost of refinancing for which the merger reduces expected aggregate liquidity needs.
The larger the asymmetry in banks’ balance sheets, the larger is this range of parameters
in which the merger increases expected aggregate liquidity needs. Taken together, the
results in Proposition 6 suggest that central banks have to be more attentive to the liquidity
fluctuations of the interbank market and intervene more often after a consolidation process
that leads to higher aggregate liquidity risk and higher expected liquidity needs.
How relevant are these different scenarios for changes in aggregate liquidity? One way to
proceed is to associate the level of the relative cost of refinancing with different countries or
financial systems. For example, in industrial countries with relatively sizable and developed
financial systems one would expect this cost to be rather low. In contrast, in developing or
emerging countries with less developed financial systems this cost may be quite high. Then,
Proposition 5 suggests that in the latter group of countries bank consolidation may lead
to a deterioration of aggregate liquidity. Proposition 6 indicates instead that the impact
of mergers on aggregate liquidity in industrial countries crucially depends on whether the
reserve or the asymmetry channel dominates. The asymmetry channel may dominate when
consolidation takes the form of mergers between large banks leading to a ‘polarization’ of
the banking system. Table 1 suggests that something like this seems to have happened in a
number of industrial countries during the 1990s. For example, in Belgium, Canada, France,
the Netherlands, and Sweden, consolidation enlarged substantially the share of the largest
24