28.02.2015 Views

Exchange Rate Economics: Theories and Evidence

Exchange Rate Economics: Theories and Evidence

Exchange Rate Economics: Theories and Evidence

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Real exchange rate determination 215<br />

<strong>and</strong> C 0 is an n × n matrix which describes the contemporaneous correlations<br />

among the disturbances:<br />

x t = C 0 e t + C 1 e t−1 + C 2 e t−2 +··· (8.33)<br />

Identifying an estimated reduced form VAR system,such as (8.31),becomes a<br />

matter of choosing a unique value for C 0 . Since C 0 contains n × n elements,<br />

identification requires imposing n 2 restrictions. The structural disturbances <strong>and</strong><br />

the reduced form residuals are related as e t = C0 −1 ε t,<strong>and</strong> so the choice of C 0 also<br />

implies the choice of the covariance matrix:<br />

= C0 −1 −1′<br />

C0 ,<br />

where is the variance covariance matrix of structural disturbances. Maximum<br />

likelihood estimates of <strong>and</strong> C 0 can be obtained through sample estimates of .<br />

As we have said,identification requires choosing n 2 elements of C 0 . Almost all<br />

approaches to identifying VAR models start with restricting the n(n +1)/2 parameters<br />

of the covariance matrix . The first n of these usually come from normalising<br />

the n diagonal elements to be unity,while the remaining n(n − 1)/2 restrictions<br />

come from assuming that the structural shocks are mutually uncorrelated<br />

or orthogonal. Taken together,these restrictions on C 0 imply:<br />

C0 −1 −1′<br />

C0 = I ,<br />

where I is the identity matrix. This leaves a further n(n −1)/2 restrictions on C 0 to<br />

fully identify the model. The approach adopted in all of the papers in the SVAR<br />

literature is to assume that C 0 is equal to C(1),the long-run coefficient matrix on<br />

the structural shocks,<strong>and</strong> to impose the remaining restrictions by assuming that<br />

this is lower triangular. This means that the long-run coefficient matrix has a Wold<br />

representation <strong>and</strong> the particular ordering of variables is achieved by appealing to<br />

economic theory.<br />

Latrapes (1992) was the first to apply an SVAR decomposition to real exchange<br />

rate behaviour. In particular,he estimated bivariate VAR models consisting of<br />

real <strong>and</strong> nominal exchange for five US dollar bilateral rates over the period March<br />

1973–December 1989. Using the identification methods described earlier,he<br />

extracted two shocks – a real <strong>and</strong> nominal – <strong>and</strong> he restricted the nominal shock to<br />

have a zero long-run impact on the level of the real exchange rate. A set of variance<br />

decompositions showed that real shocks were the predominant source of both real<br />

<strong>and</strong> nominal exchange rate behaviour for the sample,<strong>and</strong> he interprets this as<br />

evidence favouring the Stockman–Lucas view of exchange rate determination (see<br />

Chapter 4). However,as Lastrapes recognises,his results may be a reflection of<br />

the two dimensional nature of his system: with multiple real <strong>and</strong> nominal shocks<br />

the results could turn out to be different. Clarida <strong>and</strong> Gali (1994) use the SVAR<br />

approach to decompose real exchange rate behaviour into three shocks <strong>and</strong> since<br />

there work has become something of a benchmark in this literature,we consider<br />

it in some detail here.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!