Analysis of Zhejiang Business Fluctuations with Dynamic Stochastic General Equilibrium Model - Innovative Management in Information and Production - page 81

Solution of the model can be described as

x t = D 1 x t − 1 + D 2 ε t (8)

f t = D 3 x t . (9)

4 Empirical Results

First we research each shock's numerical contribution to the volatility of output. In
the second column of Table 2 , it gives the standard deviation of each shock, which
explains the volatility of each shock. By the figures reported in the table, it is easy
to find that the biggest volatility of the shocks is technology shock, whose volatility
reaches 7.62 % and 1.44 times the standard deviation of real output. In Fig. 3 we
know that from 1978 to 2010, the volatility of simulated output caused only by
technology shock is 1.24 times the volatility of real output. For the technology shock
is pro-cyclical, it can be considered that technology shock is an important cause of
output fluctuations (Figs. 1 and 2 ).
In Table 4 , note all the statistics in the table are p value; the corresponding null
hypothesis is as follows: the variable in the row is not Granger cause of the variable
in the column. X * means significant in 10 % level, X ** means significant in 5 %
level, and X *** means significant in 1 % level. All the Granger causality tests
include order lag.
We can know from Table 2 that technology is lag variable, and it lags one time
of output. Investment and exogenous demand are all leading variable and they lead
one time of output. Labor is lag variable and it lags two times of output. Table 3 is
similar to Fig. 2 , so we don't explain it here.
Except the cycle behavior's effect to the volatility, we also use Granger causality
test to analyze the predicting relationship between each shock and output. All the
results are given in Table 4 .FromTable 4 we know that technology has a significant
ability to predict the impact on the actual output, which in turn strengthens our

Tabl e 2 The statistical properties of the various shocks