Practical Application Of Econometrics In Economic Research

introduction

Modern economic theory, both on the micro and macro level, increasingly complex economic processes have led to the need to create and improve the special methods of study and analysis. Thus, widespread use of modeling and quantitative analysis. Based on the latest stand out and formed one of the areas of economic research - Econometrics.

Econometrics - a science that gives a quantitative expression of the relationship of economic phenomena and processes. This science is the result of the interaction and integration of three components: economic theory, statistical and economic methods. The objective of this paper is to examine the econometrics as a science, as a whole, that is, the consideration of its object, principles, goals

The econometric studies use different types of models. But there are three main classes of models that are used in econometrics: time series models, regression models (one equation) and the system of simultaneous equations.

Econometrics is part of a complex of disciplines "Economic-mathematical methods." Det is the subject of quantified relationships and dependencies of economic phenomena and processes, economy laws.

3. Basic Principles of Econometrics

To demonstrate the basic principles of econometrics, consider an example from the insurance business (car insurance). Here the main problem is due to the complex nature depending on the size of the insurance premium on many variables, some of which can not be taken into account. So, it is obvious that the annual run of the car - it is an important factor, but to use it as an evaluation difficult. A practical solution is to define a number of easily observable factors - capacity machines, age (the owner of the insurance policy and car), geographical position, wear, each of which has some connection with the real risk, in turn, determines the actual size of the insurance premium. For example, assume that there are five such factors and each of them is measured on five levels. This results in 55 = 3215 individual classification requirements. If the insured 100,000 cars in each class will be an average of 32 cars. Since the probability of an insurance claim of 10% a year, the data in each discharge would be subjected to too large fluctuations due to random sampling error and it would be difficult to assess the true connection between what is happening in different ranks. Moreover, to engage in such a large number of individual groups would be difficult and