Audit and economical analysis

Table 31

Elasticity coefficient and beta-coefficients

Change number	x1	x2	x3	x4	x5
Elasticity coefficient	0.374	0.308	0.318	-0,080	0,061
Beta-Coefficient	0.335	0.275	0.213	-0,118	0,133

 

Similarity can also be compared to the elasticity coefficient. It is calculated by the following formula:

                                                                      (18)

Elasticity coefficient indicates that the function will change by average percentage when the argument changes to 1%.

- the yield data yield increased by 1% to 0,374% and the accumulation efficiency increased by 1% - by 0,308%.

Methods of assessing correlation analysis results . To make sure that the association equation is reliable and that it can be used in practice, it is necessary to evaluate the reliability of the association indices. To do this, the Fisher criterion ( F. - response, Darbin-Watson's criterion ( DW ), average approximate approximation ( ? ), correlation coefficients ( R ) and determinants ( D ).

Fisher criterion ( F. - attitude) is calculated as follows here                                                                                              

where,  Y_xi- the individual value of the calculated decimal point;

Y_x - average value of the calculated decimal point;

Y_i - actual individual value of the resulting indicator;

m - the number of parameters of the equation of reference taking into account the free part of the equation;

n - number of observations (selection size).

Comparison of the actual magnitude of the F- immune system is based on the reliability of the connectivity. In our example, the F- immunobat is 95.67 at the fifth step. F javdal F is calculated in terms of value in tables. The probability range is p = 0.05 and the number of degrees of freedom ( m - 1) = 6 - 1 = 5, ( n - m ) = (40 - 6) = 34 in 2.49. If F > F _jad , profitability and examined factors that are not in contact hypothesis was rejected.

In order to improve the methodology of correlation analysis, the regression model should also be evaluated by Darbin-Watson ( DW ) criterion . This is used to find out the existence of auto correlation among the factors studied. Special tables Watch v s number of factors, and the results obtained at the expense of their levels, based on the minimum and maximum limits.

If the measure of the levels of d < DW <4 - D on the border, then the regression model between the o'rganilayotgar factors may come to the conclusion that there is no avtokorrelyatsiya. If the autocorrels exist, the bonding equation is considered to be unsatisfactory.

In our example, DW = 1.96, sudden (critical) points d ₁ , d ₂ is the number of observations, n = 40, the number of contact variables of the equation m = 5 and given the significant level a = 0.05 and 1.23, respectively 1,786.

The modomic DW is between its initial and final boundaries (1,786 <1,96 <2,214), indicating that there is no autocorrelation. This is the highest quality approved model.

The average error ( ? ) Of the approximation is used to estimate the accuracy of the link equation :       

The less the average error of less than the approximate approximation (the experimental) from the theoretical line of regression (from the equation). This indicates the correctness of the bond equation. In our example it is equal to 0.0364 or 3.64%. Given that economic calculations may contain up to 5-8% errors, this equation can be judged to be a well-defined link. It is possible to predict the degree of profitability of this equation with such a small error.

Through correlation coefficients and the size of determinants, it is possible to discuss more fully the link equation. In our case, the last step is R = 0.92 and D = 0.85. Thus, the profitability variability is dependent on the change in the factors under study, with 85% and 15% of the factors under consideration. Thus, this equation can be practically applied.

Methods of application of correlation analysis in practice . The regression equation investigated on all parameters can be used as follows:

- to assess the results of economic activity;

- calculation of the influence of factors on the increase of the results;

- calculation of inventories to increase the level of the indicators being studied;

- planning and forecasting its quantity.

When evaluating the results of the economic activity of the enterprise, the actual value of the resulting indicator is compared to the theoretical (calculated) indicators determined on the basis of a large number of regression equations. In our example (see Table ₁ ), the amount of material in enterprise No ₁ ( x ₁ ) - 2.4 UZS, savings unit ( x ₂ ) - 80 tiyin , labor productivity ( x ₃ ) - 8,000 sums, circulation term ( x ₄ ) - 25 days, specific weight of the highest quality product ( x ₅ ) - 25%. The profitability of this enterprise is calculated as follows

Yx0,493,652,40,09801,0280,122250,05225 22,86%

It is increasing by 0.36%. It is possible to conclude that the company is badly profitable.

The effect of each factor on the outcome (outflow) is calculated as follows:

Yxi= bi*xi                                                                                  (21)

where b _i - contact rate of regression equation;

xi - change of factor index in reporting period.

Hence, according to the indicators of each factor, the plan was not fulfilled (table), the profitability rate did not reach 2.09%.