مصیبت منابع

After a kind invitation of Sadegh to contribute in this weblog, I thought to present a summary of main finding of our recent study about the macroeconomy effects of oil price shocks on the Iranian economy. The first version of this study, which has been revised especially in the methodology section, delivered in the 11th IIES Int. conference (2006) in Tehran and can be downloaded from this link:

http://iies.org/Data/BulletinANDMagazine/faslnamenum10-05.pdf


1- one of the contribution of this study is identifying the positive and negative oil price fluctuations following Hamilton ( 1983) and Mork (1989). By separating this increasing and decreasing trends of oil price movements , we can examine much more accurate the effects of asymmetric oil price shocks on the other variables in the system.
2- The empirical model is an unrestricted VAR model. In such a model all variables in the system are endogenous, except dummies and intercept. This approach is useful especially at the time of lack of strong theoretical reasons for behaving one variable as exogenous and independent of the others. In fact, by treating all variables as endogounce in the system, we target the interconnection and dynamic of variables within the system. Variables are non-stationary in the levels but stationary after differencing. They are also co-integrated. In this situation some researchers apply VECM and the others VAR in levels or in differences (or Structural VAR in differences). We followed Hamilton (1994) and used variables in the levels. The case of losing useful information while there are cointegration vectors in the system argued by Sims (1980), and Doan (1992),too.
3- In VAR modelling, the coefficients of variables do not have a practical benefit, but we can apply the 2 main tools by estimating this VAR model. One is Impulse response functions and the other is Variance decomposition analysis. The former tool enable us to examine the response of the other variables in the system to one standard shock ( or innovation) into oil price variable. By using IRFs tool we can reply to question such as: A) what is the response of real effective exchange rate to a shock in positive (or negative) oil price changes? How large is this reply? Is it significant response? How long this response takes? And so on. One of the important things to report is siginificancy of responses by establishing impulse error bands. Unfortunately, some researchers like Ali Sarzaeem (See : http://www.iies.org/Data/BulletinANDMagazine/Faslnameh12F2.pdf ) do not report the error bands for their responses. It is like reporting coefficients in an OLS regression without mentioning the t statistics!
4- The variables which we have used in this study are: real effective exchange rate, real industrial GDP per capital, real governmental consumption, real imports, inflation, and real oil price ( positive and negative definitions) . There are dummy variables covering Iraq-Kuwait war(1990), financial crisis of SAsia(1998). 11 Sep(2001),and Iraq war (2003). The time period is 1988:1-2004:4.
5- RESULTS

• Oil price increases (decreases) have a significant positive (negative) impact on industrial production.
• Oil price shocks do not have significant effect on the real governmental consumption. The outcome is somehow unexpected. However, we can justify this issue by taking into account the estabilishemnt of Oil Stabilization fund since 2000 in Iran. This fund was successful , according to our empirical results, to offset unexpected shocks in oil prices . The non-significant response of governmental expenditures, then, can be explained by OSF. However, Ali Sarzaeem has found opposite results in his mentioned paper.
• The response of inflation to positive AND negative oil price fluctuations is positive and significant. That means Iranians will lose their power of purchase regardless of increasing or decreasing oil price shocks. Increasing oil prices which improves terms of trade in Iran would imply an increase in oil export revenues. This leads to an increase in spending on all products , which increase domestic prices relative to foreign prices. On the other side, negative oil price shocks worsen the budget deficits of Iran which force the government to borrow from central bank to offset the negative shocks. That may also increase the money supply and brings inflationary consequences..
• Interestingly, we did not find a significant response of real effective exchange rate after one standard innovation into positive oil prices. One of the Dutch disease symptoms is increase of REER by increasing oil prices. That would be a sign of appreciation of Rial against US dollar and loss of completion in international markets. However, it is not so in our study. In spit of increasing REER in midterm after a shock in positive oil prices, this response is not statistically significant. Contrary, the one SD shock to negative oil prices ( decreasing oil prices) improve the competitiveness of the Iranian economy by significant reduction of REER for the total period of simulation .
• I hope that I was able to explain some of finding of this study here.