Cashin, P., L. Céspedes, and R. Sahay, 2002, “Keynes, Cocoa, and Copper: In Search of Commodity Currencies,” IMF Working Paper 02/223 (Washington: International Monetary Fund).
Corden, W., 1984, “Booming Economic Sector and Dutch Disease Economics: Survey and Consolidation,” Oxford Economic Papers, Vol. 36, pp. 359–80.
Hinkle, L., and P. Montiel, 1999, Exchange Rate Misalignment: Concepts and Measurement for Developing Countries (New York: Oxford University Press).
Koranchelian, T., 2005, “The Equilibrium Real Exchange Rate in a Commodity Exporting Country: Algeria’s Experience,” IMF Working Paper 05/135 (Washington: International Monetary Fund).
MacDonald, R., and L. Ricci, 2003, “Estimation of the Equilibrium Real Exchange Rate for South Africa,” IMF Working Paper 03/44 (Washington: International Monetary Fund).
Paruolo, P., 1997, “Asymptotic Inference on the Moving Average Impact Matrix in Cointegrated I(1) VAR Systems,” Econometric Theory, Vol. 13, pp. 79–118.
Reinhart, C., and K. Rogoff, 2002, “The Modern History of Exchange Rate Arrangements: A Reinterpretation,” NBER Working Paper Series 8963 (Cambridge, Massachusetts: National Bureau of Economic Research).
Samiei, H., 1998, “Determinants of the Real Exchange Rate in Norway: Does Policy Matter?” Norway—Selected Issues (Washington: International Monetary Fund).
I wish to acknowledge the comments and suggestions received from Adrienne Cheasty and Rodolphe Blavy. The usual disclaimer applies.
The PPP hypothesis implies a stationary REER data series where any deviation from the equilibrium real exchange rate is short-lived.
See Hinkle and Montiel (1999) for a survey of the challenges in assessing a country’s equilibrium real exchange rate.
MacDonald and Ricci (2003; South Africa) and Koranchelian (2005; Algeria) apply similar techniques to those used in this paper, but do not examine the effects of parallel market rates on the equilibrium real exchange rates.
This paper focuses only on oil prices although a country’s oil wealth may have additional permanent income effects. The difficulty with a broader definition is that the concept of oil sector wealth is debatable. Oil wealth may be seen as unrelated to its location (i.e., as foreign exchange reserves or unexploited oil reserves), or it may be argued that greater uncertainty surrounds oil reserves. Those who argue that the effects on the economy can be limited by saving foreign exchange generated from oil exports implicitly believe that wealth location—above or below ground—is irrelevant. But this assumes zero Ricardian equivalence behavior among economic actors. In reality, the extent to which Ricardian factors play a role is likely to depend partly on past fiscal prudence and on whether oil sector developments are temporary or permanent.
For instance, a deterioration in the fiscal balance is unlikely to be matched by a one-to-one improvement in private savings and, given that net domestic demand increases, appreciation pressures will emerge. This depends, however, on the import content of the spending. Similarly, high real interest rates could attract resource inflows, thus strengthening the value of the currency. But higher rates could also discourage domestic demand, slowing down the rate of increase in the price of nontradables and depreciating the exchange rate.
An alternative would have been to use quarterly data. However, although quarterly data would provide more degrees of freedom, they are not readily available for some of the time series used. Moreover, they do not necessarily add much new information, as the primary interest is to examine long-term (rather than seasonal) factors that affect the equilibrium rate.
Several data sources were used, including the IMF’s International Financial Statistics and World Economic Outlook, the World Development Indicators prepared by the World Bank, and the Penn tables for the PPP-based GDP per capita data.
Both PPP-based GDP per capita, which averaged US$5,600 in 1950–54 and US$5,300 in 2000–04, and the share of trade with the U.S., Venezuela’s largest trading partner, have remained largely unchanged. In addition, the oil share in total exports (about 85-90 percent) is quite stable, as is also the share of the oil sector in total GDP—currently the sector accounts for 20 percent of GDP, compared to about 25 percent in the early 1950s.
The largest trade weights in the IFS database (accounting for 88 percent of Venezuela’s total trade) correspond to: United States, 37 percent; Germany, 10 percent; Italy, 8 percent; Japan, 8 percent; France, 6 percent; United Kingdom, 5 percent; Brazil, 4 percent; Canada, 4 percent; Spain, 3 percent; and Belgium, 3 percent.
Venezuela’s oil sells for 90 percent of the U.K. Brent spot price; this share has been stable.
A large share of oil resources is spent outside the central government budget. Besides its own spending, the state oil company finances some social programs directly.
Growth per capita (measured in PPP terms) peaked in the 1950s, but has been declining ever since. Specifically, per capita growth rates since the 1970s have been negative, declining at an annual average of about (i) 2 percent in the 1970-89 period; (ii) 1 percent during the 1990s; and (iii) ¼ percent in 2000-04. A similar pattern is observed among other trading partners and is not atypical of commodity-dependent countries—Koranchelian, and MacDonald and Ricci, identify similar declines in Algeria and South Africa, respectively.
The usual tests for VEC models are conducted. The null hypothesis on the normality of the residuals (Jarque–Bera) is rejected at the 1 percent level owing to excess kurtosis—that is, the VEC results are not affected (Paruolo, 1997). The test for the lag structure supports the use of three lags; specifically, adding another lag results in the rejection of joint significance tests.
MacDonald and Ricci find that a 1 percentage point increase in real interest rates is associated with an appreciation of South Africa’s REER of about 4 percent.
The cointegrating equation provides a means to assess the speed of adjustment; namely, the time in years, t, necessary to reduce the initial deviation from equilibrium can be calculated by estimating (1–γ) t =(1−δ), where γ is the estimated cointegrating coefficient and δ is the share of the catch-up being targeted (e.g., 0.5 for a half-life reduction).
Figure 3 also shows that in certain periods the differences between the official and the equilibrium rate are more marked. To eliminate the possible effects of outliers on the estimation results, a dummy variable can be added that takes a value of one whenever the equilibrium exchange rate deviates from the official rate by two or more standard deviations. Three years during the mid-1980s meet this criteria. The estimation of the VEC model including this dummy (not shown) leads only to minor changes in the estimated coefficients, suggesting that the model is not driven by the effects of data outliers.