| dc.description.abstract | Long-run or desired demand variable for money is not directly observable. By using the stock adjustment hypothesis , the long-run demand for money model may be expressed as the short-run demand function for money. The short-run demand for money model is essentialy nonlinear in the parameters an autoregressive in nature because the presence of stochastic explanatory variable. Therefore this research is conducted to know the estimation problem of such model, because the classical least-squares may not be directly applicable to the model. Based on the formulation of the problem above, the purpose of this research are as follows: 1. To determine whether the structural analysis that aims to measure and understand the magnitude of quantitative relations in economic variables statistically significant in the model and in accordance with the theoretical expectations. 2. To determine whether there is a violation of classical assumptions multikolinearitas in short-run money demand models. 3. To determine whether the presence of lag dependent variable of money demand on the right short-run money demand equation will result autocorrelation. 4. To determine whether the error (residual) regression model estimation normal distribution. By using quarterly data ranging from 2000 to 2014 at constant prices 2000, having analyzed and evaluated several conclusions can be drawn as follows. 1. Structural analysis that aims to measure and understand the magnitude of quantitative relations in economic variables in the short-run money demand models producing the correct relationship corresponding theoretical expectations and significant. 2. From the analysis and evaluation of the value of the coefficient of determination R2, the correlation matrix of the independent variables, the Pearson correlation coefficient, variance inflation factor (VIF) and the value of tolerance (TOL), partial regression, it can be concluded that the violation of the assumptions of classical multikolinearitas in the model demand for short-run money can be ignored. 3. With the presence of variables lagMt in the short-run money demand that have stochastic nature on the right short-run money demand equation, after analyzing the results of the DW test, Durbin-h test, test Langrange Multiplier (LM test) or test Breusch-Godfrey (BG test), and the test Run is not located autocorrelation in money demand models. 4. By using the test charts, statistical tests such Zskewness and Zkurtosis and nonparameter test that is the Kolmogorov-Smirnov test can be stated that the error (residual) estimation is a normal distribution. | en_US |
