Expanded Summary

The article empirically tests R. Kopke’s investment cash flows model according to Georgian economy. During empirical realization of the model problems of lack of information or non-existent data are resolved by different economic methods (using Substitutive variables, estimation data of other econometric model and etc.).

Empirical model’s dependent variable is Georgia’s total real investment as factor variables are Georgia’s total real capital, companies’ turnover and their lagged values. Because of lack of empirical information about Georgia’s total capital, indicator values are estimated according L. M. Koyck’s  private modification of investment accelerator model. For the same reason in the model factor variable of companies’ total internal cash flow are replaced with substitutive variable, companies’ total turnover. Factor variables’ lags length are not defined in the investment cash flow model. To overcome mentioned problem factor variable’s lag length was gradually increased, which revealed that statistically optimal length was equal to 0.

Empirical model’s residual members are distributed normally. Durbin-Watson test is not recommended for checking autocorrelation in residual member, because model assumption is violated: model does not contain a constant [Ananiashvili, 2014: 10-13]. Breusch-Godfrey’s Lagrange multiplier (LM) test is recommended for checking serial correlation in residual member. According to the null hypothesis of LM test there is no autocorrelation in residual member, whereas alternative hypothesis states that there is serial correlation in residual member [Breusch, 1978: 334-355; Godfrey, 1978: 1293-1301]. Breusch-Godfrey test results show that null hypothesis is statistically significant, which states that there is no serial correlation in residual member.

While checking heteroscedasticity in residual member with White null hypothesis is accepted with 99% confidence interval, which means existence of homoscedatsticity. Accordingly, variance of residual member is constant.

It is shown that the coefficients of Georgia’s total capital and constant member are not statistically significant as for other coefficients H1 hypothesis of T-test is accepted for statistical significance. Therefore, non-significant factors are excluded from the model. After stated changes empirical model has following form:

 ΔI_t = 0,252ΔB_t + 0,109ΔB_t-1 + U_t 

Also, coefficients of determination and corrected determination have values - 0.53 and 0.52 accordingly, which shows that model variation is explained by model variables but theoretical model does not include statistical significant factor or factor group. To increase determination coefficients it’s recommended to re veal variables and insert them into model. Also, F=17.15, Fcr=5.39, F> Fcr shows that model is valid [Ananiashvili, 2010: 109-129].

The empirical research confirms positive correlation between amount of total investment and companies’ current and lagged turnover. Current period’s turnover has a relatively greater effect than it’s lagged value, which is logical consequence. In the current period during companies turnover growth decision of internal cash flow investment is made. Delayed process of investments’ realization lead to relationship between current period’s investment with previous period’s companies’ turnover, because investment process has gradually character nor instant. Accordingly investment process includes not only one time period, but also next periods too. In the model factor variable’s lagged value reflects stated process.

Empirical cash flow investment model does not include constant and total capital as factor variables unlike theoretical model.

According values of determination coefficients Empirical investment model does not include statistically significant variable or group of variables. Therefore, it is desirable to continue research in this direction to find answer question non-existed variables are internal or external funding source.

Keywords: Total investments, Investment cash flow model, total turnover of companies.

JEL Codes: P 20, P 31, P 40, P 52