multiplying with the global growth rates. Hence, the level of the variable v is:
T
indexv,τ = 100 * ∏(1 + gagg,t)
t=2
This method is applied to all variables except for the MSCI World, which already
represents shares on a global level. Moreover, for the interest rate variable, aggre-
gation is performed directly without calculating growth rates.
Regarding the monetary aggregate which plays a central role in our analysis this
method lowers the bias resulting from different national definitions of broad money
which obviously exist. Building a simple sum of national monetary aggregates - a
method frequently applied in the related literature - would underrepresent countries
with narrower definitions of the monetary aggregate and vice versa. A second prob-
lem that is avoided is the ”dollar bias” resulting from converting national monetary
aggregates with actual exchange rates into USD and building a simple sum to obtain
global money. In this case the fall of the dollar contributes to an overestimation of
global monetary growth.7
To illustrate the development of global liquidity since 1984, Figure 1 shows global
monetary aggregates in absolute and relative terms. For nominal and real money,
a simple regression on an intercept and a linear time trend is performed. Both
series are above their time trend since about 2001 when the rapid downturn in stock
markets caused households and investors to increase the share of safe assets like
money in their portfolios. Monetary growth remained strong afterwards, which can
be seen in the persistent growth of the ratio of nominal money to nominal GDP, a
measure commonly used as an indicator of excess liquidity.8 As this series is equal
to the inverse of the income velocity of money, it seems obvious that global velocity
is not trend-stationary, a phenomenon which has appeared on a country level as well
and has contributed to the instability of national money demand equations. Overall,
the series confirm our prior that global liquidity is indeed at a high level and that
the term excess liquidity ought to be justified.
Figure 2 shows the whole array of our global time series. The price level series
7See Commerzbank Economic & Commodity Research (2007), p. 3.
8See inter alia Belke et al. (2004) or Ruffer and Stracca (2006).
12