30.06.2013 Views

TESTING INTERNATIONAL PRICE TRANSMISSION UNDER ...

TESTING INTERNATIONAL PRICE TRANSMISSION UNDER ...

TESTING INTERNATIONAL PRICE TRANSMISSION UNDER ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

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

METHODOLOGY DATA USED MAJOR FINDINGS<br />

DYNAMIC REGRESSION MODELS BASED ON A POINT LOCATION MODEL (GC TESTS, IRFS, FEV)<br />

Gupta and Mueller, 1982<br />

They use GC tests (Haugh test to<br />

assess the dependence-independence<br />

between series; Sims test to ascertain<br />

the direction of causality) to see<br />

whether markets are perfectly priceefficient.<br />

Differently from the use of<br />

correlation coefficients, which only<br />

report the association between prices,<br />

this methodology allows to test if<br />

markets are independent,<br />

interdependent, or if lead-lag relations<br />

exist.<br />

Myers et al., 1990<br />

VAR methods are used to analyze the<br />

contribution of supply, demand and<br />

policy shocks to unpredictable<br />

fluctuations in the market for<br />

Australian wool. They develop a<br />

structural VAR in which the<br />

restrictions (which define<br />

contemporaneous interactions among<br />

variables) do not impose a recursive<br />

ordering of the variables. They<br />

perform IRF and FEV.<br />

Vollrath and Hallahan, 2006<br />

They investigate market integration<br />

for meat and livestock in the US and<br />

Canada, making use of three models:<br />

“streamlined” (prices expressed in a<br />

common currency) and “detailed”<br />

LOP models (to quantify both foreign<br />

price and exchange rate transmission),<br />

VAR models (IRF). They consider<br />

trade policy barriers and exchange rate<br />

changes, and evaluate the CUSTA<br />

(Canadian US Free Trade Agreement,<br />

since 1989) and its extension, the<br />

NAFTA (North American Free Trade<br />

Agreement, since 1994)<br />

Weekly prices (differentiated<br />

logs) for slaughter hogs in three<br />

German markets.<br />

Week 1:1977 - week 50:1980<br />

Quarterly prices for Australian<br />

wool, quantity purchased by<br />

private traders, total quantity of<br />

wool sold (in logs).<br />

1971:1 to 1988:2.<br />

US and Canada wholesale<br />

monthly and weekly prices for<br />

slaughter (steers, hogs, whole<br />

chicken), two cuts of beef and<br />

two pork products (in logs);<br />

US/Canada exchange rate.<br />

1988 - 2002<br />

DYNAMIC REGRESSION MODELS BASED ON A POINT LOCATION MODEL: COINTEGRATION<br />

Ardeni, 1989<br />

For the first time, cointegration is<br />

proposed to analyze the LOP, arguing<br />

that previous tests are flawed for<br />

disregarding the time series properties<br />

of the data.<br />

Quarterly export prices (in logs)<br />

for wheat, wool, beef, sugar,<br />

tea, tin and zinc for Australia,<br />

Canada, the UK, the US. Data<br />

are adjusted for exchange rates.<br />

January 1957-January 1986<br />

(from 79 to 117 obs., depending<br />

on the commodity).<br />

Annexes<br />

They find that markets are<br />

price-efficient, since the tests<br />

show that they are<br />

interdependent.<br />

Demand shocks are the<br />

dominant source of wool<br />

market fluctuations but<br />

intervention has blunted their<br />

effects.<br />

The market of whole chicken<br />

is segmented; hog and pork<br />

product markets are more<br />

integrated than steer and beef<br />

product markets. Changes in<br />

the Canadian-US exchange<br />

rate inhibits cross border<br />

integration.<br />

The LOP holds only in a<br />

small number of cases (but is<br />

nonetheless valid for US-<br />

Australian wheat, US-<br />

Canadian wheat); deviations<br />

from the pattern are<br />

permanent.<br />

127

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

Saved successfully!

Ooh no, something went wrong!