Читать книгу Reductive-Investment Analysis - Fuad Akhundov - Страница 4

The Reductive-Investment Analysis, as a descriptive method, combines a set of theoretical techniques and models based on econometric theory, using mathematical and statistical tools (linear regression channel), providing quantitative expressions with visual perception with the further possibility of visual modelling of goals and prospects for further development of the ongoing process.

In economic researches the problems of identifying the factors that determine the dynamics of the economic process are often solved. Also, in order to reliably reflect the processes existing in the economy objectively, it is necessary to identify significant relationships and give them a quantitative assessment. This approach requires the disclosure of causal dependencies. The causal dependence means such a relationship between processes, when a change in one of them is a consequence of a change in another. The solutions of such matters most often use methods of correlation and regression analysis.

Economic data is usually presented in tabular form. The numerical data contained in the tables usually have explicit (known) or implicit (hidden) connections between them. The indicators that are obtained by direct counting methods, i.e., are calculated according to previously known formulas, are clearly connected. For example, percentages of plan fulfilment, levels, specific gravity, deviations in the sum, deviations in percentages, growth rates, accession rates, indices, etc. Connections of the second type (implicit) are not known in advance. However, it is necessary to be able to explain and predict complex phenomena to control them. For this reason, with the help of observations, specialists seek to reveal hidden dependencies and express them in the form of formulas, mathematically simulate phenomena or processes. One of such opportunities is provided by the correlation and regression analysis. Mathematical models are built and used for three generalized purposes: explanation, prediction and control. The presentation of economic and other data in spreadsheets or through the tools of trading platforms has become ordinary and widespread these days. Equipping electronic trading platforms with the means of correlation and regression analysis gives specialists in the financial field opportunity to transform well-founded probability-theoretic methods into everyday effective analytical tools. Using the tools of correlation and regression analysis, analysts measure the linear statistical dependence of the indicators by means of the correlation coefficient. This reveals the connection, different in strength and direction. Regression analysis is one of the main methods to identify implicit and covert connections between observational data in modern mathematical statistics.

Mastering the technique of using tools based on regression analysis, you can apply it as needed, gaining knowledge about hidden connections, improving analytical decision-making support and increasing their validity. Thus, the methodology of the Reductive-Investment Analysis is a research method closely related to the tools of correlation and regression analysis, which is based on the use of the Linear Regression Channel tool (Fig. 1) available in modern electronic trading platforms. The Linear Regression Channel is built based on the Linear Regression Trendline, an ordinary trend line built between two points on the price chart using the least squares method. This method calculates the Y=a+b*X trend line, minimizing the sum of squares of vertical deviations between the closing price value and the trend line value during a given time interval. The trading platform software calculates the values (a, b) and builds a Linear Regression line for any time interval. As a result, this line turns out to be the exact median line of the changing price.

The Linear Regression Channel, developed by Gilbert Raff in 1991, consists of two parallel lines equidistant up and down from the linear regression trend line. As a result, the Linear Regression Channel consists of three parts: the median line is the trend line; the upper and lower lines are the borders of the Linear Regression Channel. The distance between the borders of the channel and the median trend line is equal to the maximum closing price deviation from the median line of the Channel (Fig. 1).

A Reductive-Investment Analysis is a multidimensional method applied for the visualization of the interrelation between the values of quantitative variables. The basic idea of the concerned analysis lies in the fact that available dependencies among large number of initial observable variables give us opportunity to analyze the development of phenomena in time. The methods of the Reductive-Investment Analysis make possible not only to explore the data, but also to choose a method for their further in-depth analysis for examination of the statistical hypotheses and modelling of the further dynamics. In the current analysis, the price information about the examined phenomenon is shown in aggregated form by means of graphic tools. The main objective of the Reductive-Investment Analysis is the modelling of both the future development of the further financial market dynamics by means of descriptive tools, and the corresponding actions of the market participants meant to simplify the analysis procedures.