with the fundamentals in levels, the first differences as well as first differences with up to four
period lags and four period leads were also considered. Here, for lack of space only the
coefficients of the fundamentals in levels are reported. The t-statistics are adjusted by a factor, as
recommended by Hamilton (1994, pp. 610) and Hayashi (2000, pp. 656 - 658).
2 The residuals are calculated as the differences between the observed values and the fitted
values (given by the fundamentals in levels). The critical values are from Philips and Ouliaris
(1990, Table IIc).
All of the coefficient estimates (except that of OPEN) in the final specifications are
statistically significant at 5% levels, and have the expected signs. The OPEN variable is
statistically significant at the 10% level for both countries. The TREND variable essentially
accounts for the effects of productivity growth. To the extent productivity growth takes place in
the non-traded goods sector, the real exchange rate is expected to fall. However, if productivity
increases in the traded goods sector instead, then the demand for labor in this sector increases,
wages rise, and as a result, the price of non-traded goods increases and the real exchange rate
rises (a positive Balassa-Samuelson effect). For both the countries, TREND is found to be
significant with a positive sign, thus indicating productivity growth in the traded goods sector.
With the transition economies opening to the world markets, the domestic producers of traded
goods face increased competition on the world market. The result confirms the suggestion by
Halpern and Wyplosz (1997), inter alia, that changes in demand and productivity conditions
contributed to the real appreciations of the exchange rates in the transition economies. The
coefficient for the terms of trade, TOT, has a negative sign for both countries. This indicates that
in these countries the substitution effects are larger than the income effect of a change in the
terms of trade. Capital flows, FLOW, is found to be statistically significant for both countries
with a positive sign. This is important since Brada (1998) ) and Drabek and Brada (1998) point
out that all the transition economies liberalized capital accounts, and this resulted in massive
26