Page 131 - Contributed Paper Session (CPS) - Volume 5

CPS1144 Adeniji Nike Abosede et al.
error variance decomposition. Most of the variation is due to its shock. We
consider a VMA (vector moving average) representation of eq(7) at aggregate
level which is given as: Where, yt= total export, zt naira dollaer exchange rate,
and e1t,e2t are the error terms which are composite to the structural innovation
from the restricted vecto auto regression.

Where and

c
c
where B = cofactor of B, (B ) = Transpose

T
Further, the study of dynamic relationship at disaggregate level between
(non – oil, oil export) and naira dollar exchange rate is forwarded by repeating
eq(3) to eq(7), and the procedure is carried in in the say way

Now

where =1-b21b12

, ,s 2
 s are white noise, thus e (0, i)

Where
is time dependent, and same as for Var (e2t)

eq(7) is estimated with ordinary least square regression model since the right
hand side consists of predetermined variables and the error are due to white
noise. When fitting of the dynamic model is done, we proceed ahead and
determine how much of a change in a variable is due to its own shock and due
to shock of other variables at both aggregate and disaggregate level. This has
been the interest in this research studies, and this is achieved by the forecast
error variance decomposition. Most of the variation is due to its shock. We
consider a VMA (vector moving average) representation of eq(7) at aggregate
level which is given as:

120 | I S I W S C 2 0 1 9