Key milestones and objectives of econometric models

INFORMATIZATION, SYSTEMATIC, FORECASTING, INFLATION

A specific feature of the activities of the economist is to work in a lack of information and incomplete baseline data. Analysis of this information requires special techniques that constitute one of the aspects of econometrics. The central problem of econometrics is to build an econometric model and determine the possibility of its use for the description, analysis, and forecasting real economic processes. Econometrics - a fast growing branch of science, which aims to convey quantitative measures of economic relations. Econometrics term was first introduced in 1910 as an accountant Tsempoy. the word "econometrics" consists of two words: "the economy" and "metric". The term emphasizes the specificity of science, i.e., quantification of the connections and relationships, which are disclosed and justified by economic theory. This science is the result of the interaction of three components: economic theory, mathematical methods, statistical methods. Later they were joined by the development of computer technology. Currently econometrics has a huge variety of models from large macroeconomic involving several hundred or thousands of small equations to equations designed to solve specific problems.

The dynamics of inflation largely determines the perception of the economic and political situation prevailing in the country, Russian and foreign expert organizations, including international rating agencies.[1]

Formation and development of econometric methods on the methods of computational statistics: - on the methods of simple and multiple correlation; -Selection of the trend and others. Components of time series; - By statistical estimation. Inflation model can be built on the basis of forming a dynamic function of aggregate demand and aggregate supply function dynamic. [2]

Need of causality correlation led to the creation of limit analysis - based on the study of the entire structure of a causal relationship between the variables, i.e., on the construction of the graph. His basic assumption is that the evaluation of standardized coefficients and recursive system of equations, which are called the coefficients of influence, calculated on the basis of simple correlation coefficients. When working with time series of different indicators in the study of the relationship between them was aware of the problems of false correlations, which arose under the influence factor lag, i.e. time shift. Much attention is paid to the problem of data econometrics, i.e. special way of working in the presence of data gaps, the impact of pooling the data, etc. Information may not be available for Dep. units together and be on the same level only, the information is not on the financial organizations, and by district. The results can be very different. Mathematical model is recalculated for each new simulation experiments in which values of the basic undefined variables are randomly selected based on random number generation. The results of simulation experiments are combined in the sample and analyzed using statistical methods in order to obtain the probability distribution of the resulting indicator and the calculation of basic gauges risk of the project. [3] On the problem of data is also the problem of selective sampling in microeconomics. Typical direction in this area: the labor market; identification of factors affecting the decision to sample the work; the economic incentives affect the decision on education.

In this case, the sample may not be random, but is limited to some specific situations and not all possibilities. The effect of self-selection occurs when an objective selection is replaced by "convenience sample". Econometric research includes decision seq. problems:

• 1.    The qualitative analysis of the relations of economic variables -dependent release of Yⁱ and dependent variables X^k. 2. The selection of data. 3. The specification model relationships between variables. 4. Evaluation of the model parameters. 5. The testing of hypotheses about the properties of the probability distribution for the random component: the hypothesis of the average; dispersions; covariance. 6. The introduction of dummy variables. 7. Identification of autocorrelation lags. 8. Identify the trend, cyclical and random components. 9. Balance check for heteroscedasticity (lack of standards for the distribution of the regression function). 10. The analysis of the structure of relationships and building a system of simultaneous equations. 11. The simulation system based on time series. 12. The problem and the identification and estimation of parameters.

The econometric model is based on the assumption that the dialectical circle of interrelated variables. With all striving for the best description of the connection priority is given to qualitative analysis.