The results of the testing of this hypothesis shall be reviewed as an evidence favoring the assumption that loss-making is responsible for the generation of payment arrears (Model 3). Accumulation of unpaid finished products results from ineffective demand and the disparities in the structure of aggregate demand and supply. Insufficient financing may be reviewed as a reduction of prices of end products that being equivalent to loss-making. The problem of insufficient financing of state expenditures is outside the framework of short term cash gaps and can not be eliminated by applying methods creating incentives for bank crediting.

Ownership Structure

The effectiveness of enterprise operations may to some extent be determined by the form of ownership. It is well known that as a rule state-owned property is managed less effectively than private property. The principal – agent problem may to a certain extent account for this.

Therefore, the presence of a relationship between the ownership structure and effectiveness, the ownership structure and payment arrears may indicate the principal – agent problem.

It shall be noted that in case of Russia the effectiveness of enterprises may be related to privatization. On the one hand, privatization might primarily concern the most effective enterprises. On the other hand, it is well known that enterprises were often deliberately made bankrupt to facilitate their privatization.

State protectionism may account for the fact that state-owned enterprises show a greater propensity to generate payment arrears. Enterprises belonging to the military and industrial complex, social sphere and otherwise related to the production of public goods depend on the financing of state procurement and therefore may be more inclined to non-payments. This problem is closely related to the ineffectiveness arising due to lack of budgetary financing.

Our previous research provided evidence that there is a significant statistical relationship between shares of state-owned and loss-making enterprises. There was also obtained the statistical evidence that there exists a relationship between non-payments and the ownership structure. Regions with higher shares of public sector (share of enterprises, share of industrial output) accumulated more payment arrears. However, in spite of a clears statistical relationship the percentage of explained dispersion is relatively small. Proceeding from this fact it may be asserted that non-payments are characteristic not only of public sector enterprises. This conclusion agrees with the results obtained by Alfandari and Schaffer (1996), who found out no significant relationship between payment arrears and types of ownership at the micro-level.

Periods of Instability

The economic theory (Hicks, 1939, Keynes, 1936) describes the following effect. Economic agents seek to sharply increase their liquid assets in periods, where there are registered a steady process of economic operations and a sharp increase in uncertainty and risk factors. Apparently, this factor may play an important role in transition economies, where liberalization of prices, privatization, and other reforms sharply increase the uncertainty of further economic development, what may be reflected, for instance, by accumulation of indebtedness. In the periods characterized by sharp economic and even political shifts (for instance, the Presidential elections in Russia held in 1996 might have had a considerable impact on the further economic development) entrepreneurs will apparently rush to accumulate liquid resources (in some cases transferring them to off-shore companies) and therefore generating their current non-payments.

Due to the above reasons, in the course of the empirical analysis of overdue indebtedness it makes sense to single out periods characterized by such an increase in uncertainty. By introducing dummies related to these periods in the dynamic model in the course of our preceding research we could support this hypothesis in econometric terms that being an evidence that uncertainty factors are significant in the process of payment arrears accumulation. It seems that these factors shall be attributed to the premeditated causes of non-payments (Model 1), since they are determined by the framework of rational strategies pursued by economic agents seeking to retain capitals under uncertainty. However, there is another possible mechanism transmitting the impact of these factors on payment arrears, for instance via a rise in interest rates and contraction of bank crediting in the periods of instability (Model 2).

Offset Transactions

The avalanche of mutual payment arrears, including indebtedness to the budget (budgetary payment arrears) resulted in the emergence of non-traditional tax collection methods, i.e. monetary and non-monetary budgetary offsets. These operations have a number of negative aspects, which include such major factors as creation of incentives to further accumulate indebtedness and insufficient transparency of these operations (for details on offset practices in the Russian Federation and their impact on economic effectiveness see Annex 4).

Therefore, the moral hazard problem most often refers to offsets. Firms may accumulate non-payments irrespectively of their financial standing in anticipation of offsets in order to derive additional gains (Model 1). The relationship between non-payments and offsets, tax amnesties was studied in Ivanova, Wyplosz (1999), and in the course of the comparison between dynamics of budgetary payment arrears and federal offsets there was found out a certain evidence that offset transactions have an impact on further growth in non-payments to the budget (offset transactions were followed by increases in budgetary payment arrears).

It shall be noted that the problem concerning the prime cause of the emergence of offsets (the government – enterprises conflict (Model 1), third factors (for instance, loss-making (Model 3), or payment arrears on the part of counteragents (macro-model)) remains unsolved. On the one hand, a cartel (collusion) of enterprises having a certain political power may force the government to yield and conduct an offset. On the other hand, offsets may emerge due to the absence of alternatives both for the government and firms. In the former case offsets (alongside with budgetary payment arrears) result from the moral hazard problem (Model 1), in the latter they represent the factors of non-voluntary nature responsible for insolvency of enterprises (Models 2, 3, macro-model).

SECTION 2. EMPIRICAL TESTING OF HYPOTHESES, CONSTRUCTION OF ECONOMETRIC MODELS12

This section focuses on a number of econometric tests of relationships between payment arrears and certain factors indicated in the theoretical section of this paper in order to find out the level of the non-payment problem in the Russian economy.

All empirical tests presented in this paper use regional data. In contradistinction to our previous tests, a number of which also making use of regional data, this presentation seeks to combine the data in the framework of a single dynamic model and to test new hypotheses. This approach will permit to substantially expand the sampling (up to 2 thousand observations), improve the quality of evaluation, and to ensure greater confidence in derived conclusions, as well as considerably expand the types of tests, since it combines both dynamic and panel data. The use of regional statistics allows to determine general regularities of development of the analyzed processes at the regional level (over time), to detect their differences and obtain additional information about the process.

The paper elaborates the methodology allowing to evaluate amounts of offset transactions at the regional level and studies this indicator.

The empirical testing is aimed to build regional dynamic econometric models of non-payments and offset transactions allowing to evaluate the aggregate impact of the analyzed factors.

Dynamic Regional Model of Non-Payments

In the theoretical section of this paper there were reviewed three models explaining the emergence of payment arrears at the micro-economic level, analyzed different factors affecting or related to non-payments, which could help to detect the origins of the indebtedness described by the models. These relationships are studied below.

The increment in the overdue creditor indebtedness relatively to the total volume of industrial output (over the respective period). This indicator characterizes the dynamics of aggregate payment arrears in the industrial sector13 (excluding the arrears of payments due to banks). For details concerning the choice of this indicator see Lugovoi, Semenov (2000).

Indicators of Availability of Credits

The hypothesis that bank crediting is a significant factor for the non-payment problem shall be tested first. As it was mentioned above, the testing of this hypothesis presents certain difficulties due to the fact that banks credited the real sector at a rather low level over the period under observation. A certain increase in the variation of this indicator may be achieved in case we switch to the regional data.

Interest rates may present another factor reflecting the availability of credits. Interest rates on and amounts of bank loans to legal persons are the most important characteristics of the performance of the banking sector (with regard to the real sector). It shall be noted that in contradistinction to the amount of crediting interest rates do not account for the rationing of credits.

To test this hypothesis the following model shall be evaluated:

, (2.1)

where

is the increment in the overdue creditor indebtedness of industrial enterprises in the i-th region over the period () relatively to the average annual industrial output (the share of non-payments accumulated by enterprises in this region in the output unit value, the regional variable, quarterly data);

is the rate of increase in the nominal interest rate on granted credits (the macroeconomic variable);

is the real interest rate on granted credits (the macroeconomic variable);

is the amount of granted credits relatively to the volume of industrial output (the regional variable, annual data);

,, are the stochastic components;

are coefficients, parameters of the regression equation.

The real interest rate in the model reflects the real value of credit resources. The higher is the real interest rate, the less possibilities enterprises have (and the less are inclined) to resort to bank credits and the higher is probability of payment arrears.

Sharp increases in the nominal interest rates reflect the contraction of liquidity, increasing uncertainty, inflationary expectations, what may, according to the hypothesis, facilitate growth in payment arrears.

In contradistinction to interest rates the amount of granted credits is a regional variable representing the activity of banks in regions. From the substantive point of view it is more appropriate to use the dynamics of this indicator (increments), which characterize changes in the level of banks’ activity in the analyzed model, since the used explained variable is also expressed in increments. However, the available statistics do not allow to find out the difference, since the methodology of calculation was changed several times over the period of observation, what renders the data poorly comparable across time. Therefore, the model used actual amounts of granted credits, which nevertheless permit to conduct a cross-regional comparison with regard to the banking activity. According to the hypothesis, there is expected a negative coefficient, since increases in crediting weakens the demand for trade credits, decreases payment arrears.

For the results of the evaluation of coefficients of model (2.1) see Table 2.

Table 2. eq03Results of the evaluation of model (2.1), OLS, I/1995-IV/2000.

The variables in the model have high statistical significance, however the assumption about the normalcy of the model residuals. According to the results of the Jarque-Bera test, the hypothesis about the normality of the residuals is discarded (see Fig. 8, Annex 2). The removal of outliers does not considerably change the situation (see Fig. 9, Annex 2). Although evaluations remain unbiased, the distribution of evaluations does not correspond to the normal, what prevents conclusions about their statistical significance basing on t-statistics.

The non-normality of residuals may result from the specifics of the pooled regressions, i.e. the probable heteroskedasticity of the analyzed value (both over time and across regions). Even in case the assumption about the normality of residuals proves to be true for each period of time (quarter) and each region, the mix of samples with different variances will result in a non-normal distribution (with kurtosis and “heavy tails”).

Proceeding from the plot of model residuals (Fig. 10, Annex 2) it may be suggested that there exists heteroskedasticity. This fact is also supported by the results of the White's Heteroskedasticity Test according to which the hypothesis about the absence of heteroskedasticity is rejected (see Table 18, Annex 2).

The quarter-based evaluation of standard deviations of the modeled value () supports the hypothesis about the heteroskedasticity of non-payments across time periods. The variance of increments in non-payments across regions varied from period to period. Since this form of heteroskedasticity may be considered as known (it may be evaluated), we evaluate model (2.1) applying the weighted least square method (WLS) using the inverse value of evaluated quarterly standard errors (Table 19, Annex 2). For the results of the model WLS evaluation see Table 3.

Table 3. Results of the evaluation of model (2.1), WLS, I/1995-IV/2000.

Материалы этого сайта размещены для ознакомления, все права принадлежат их авторам.
Если Вы не согласны с тем, что Ваш материал размещён на этом сайте, пожалуйста, напишите нам, мы в течении 1-2 рабочих дней удалим его.