چکیده:
Foreign investments play a significant role in the economy of developing countries. In this article, an attempt is made to classify the member countries of the Organization of the Islamic Countries (OIC) based on the factors influencing their inward foreign direct investments. At first the relevant economic, social and political factors were introduced. Then by using a Numerical Taxonomy Method the OIC member countries were classified. Based on empirical studies, we have selected some of the most important factors (compound & single) influencing foreign investments for the ranking purpose. These factors are: Economic freedom index, the per Capita Number of Tourists, Regionalism index, Trade openness index, Market Size & Economic Stability index. Based on the results, Malaysia is the state with the highest chance of attracting foreign direct investments. Because of her high economic growth rate & the implementation of trade openness policies in recent years, this country has been able to gain the top position visa-vies other OIC member statesAs for the countries at the bottom of the ranking such as Libya & Uzbekistan the economic freedom index, as put out by the Heritage Foundation, shows that these countries have had the worst status when it comes to the compound economic freedom index among all the Islamic countries in the past years. Inappropriate policies of the public sectors in controlling financial markets, the goods and services markets, the labor market & the foreign sector have had a negative impact on foreign investments inflows. The results also indicate that among 46 OIC countries Iran is ranked in the 44th place.
خلاصه ماشینی:
"As Meese and Rogoff (1983a), Chinn and Meese (1995), Boughton (1987), McDonald and Taylor (1994), Berkowits and Giorgianni (2001), Mark and Sul (2001), Rapach and Wohar (2001) show, the structural models could not outperform the random walk model in forecasting exchange rates.
But, if the underlying data generating process of an economic series is chaotic, using traditional linear and nonlinear time series models to estimate and forecast the series would result in misleading outcomes.
According to ADF test results, we cannot reject the null hypothesis of unit root at the 1 and 5 percent significance levels, suggesting that the exchange rate series is not stationary and needs to be differentiated.
(1996) investigate the possibility of a low dimensional chaotic attractor in hourly returns of spot exchange rate in British Pound, Deutschmark, the Swiss France and the Japanese Yen. They find that correlation dimension estimates do not converge to a stable value, but the largest lyapunov exponents appear to be positive.
Table 4: The ANN test results for daily exchange rate and its returns (1991-2005) Models Exchnage rate: ARMA(15,25) residuals 31.
Thus, since the underlying data generating process of the exchange rate and its returns are nonlinear, but unknown, it seems that the flexible nonlinear models such as the Artificial Neural Networks (ANN) can be used for forecasting the series.
In the following section, we will develop an ANN model to forecast the nonlinear exchange rate series and compare the results with linear and nonlinear models.
We then apply the Lyapunov exponents and correlation dimension tests to detect chaos in daily exchange rate market of Iran."