“Determining and predicting correlation of macroeconomic indicators on credit risk caused by overdue credit”

The banking system guarantees the economic strength of the country. Its sustainabil­ ity is due to the sustainability of the credit portfolio. Therefore, scientific research on banking risks is always relevant. Basel recommendations and central bank regulations provide risk minimization in case of default of borrower by creating risk reserve, but the high range of macroeconomic factors creates a basis for creating credit risk. The model, which determines the risk factors, may be structurally the same, but the quality of the influence of factors is different in various countries. The influence of macroeco­ nomic factors is particularly evident in developing countries. The impact of economic factors in different countries is high in GDP of these countries. The article focuses on determining the influence of macroeconomic factors on credit risk of systematic banks in Georgia. The coefficients of individual macroeconomic indicators are calculated by using Pearson’s correlation. The credit risk ratio is taken from the bank’s overdue cred­ its and credit portfolio ratio. Based on the correlation coefficients obtained, the ex­ pected risk of shock changes is calculated. Asie Tsintsadze (Georgia), Lela Oniani (Georgia), Tamar Ghoghoberidze (Georgia) BUSINESS PERSPECTIVES LLC “СPС “Business Perspectives” Hryhorii Skovoroda lane, 10, Sumy, 40022, Ukraine www.businessperspectives.org Determining and predicting correlation of macroeconomic indicators on credit risk caused by overdue credit Received on: 6th of July, 2018 Accepted on: 31st of August, 2018


INTRODUCTION
Credit risk is a major threat to banking activities that cannot be com pletely eliminated. Like the production cycle, Credit risk management is a difficult and lengthy process and it includes full coverage of the credit amount and interest income accrued from the client's applica tion. A variety of approaches to the improvement of credit risk man agement models and mechanisms in scientific circles are examined. For many years, financial institutions were facing financial stability problems. Among the reasons, the first was the credit risk, which scien tists think was caused by weak credit standards and low quality of their protection. The borrower has been receiving credit in accordance with the financial condition of the loan in the process of taking the loan, but the borrower's condition deteriorated due to changes in the borrower's business, economic policies and other changes, resulted in a credit risk. There is no separate credit risk measurement tool, as credit risk is not caused itself, but by the influence of operational and market risks. Aver (2008), in his work "An Empirical Analysis of Credit Risk Factors of Slovenian Banking System", investigates the influence of macro economic factors on credit risk. Factors operating on credit risk are united into two factors, the first factor is the interest rate, and the second factor in cludes all other macroeconomic factors. Based on the analysis obtained during the research, the im portance of the first factor is high and the influence of other macroeconomic factors in the other factor is secondary, then the banking system's credit risk is higher than average. On the other hand, the low impact of both factors shows that the risk in credit portfolio is lower than average (Aver, 2008). The established ratio is a significant indicator of credit risk research, but findings can be changed as a re sult of the current situation changes in the country. Musau et al. (2017), in the work "Financial Inclusion, GDP and Credit Risk of Commercial Banks in Kenya", discuss credit risk as a combina tion of the total loans for overdue loans and eval uate three factors: access to bank resources, avail ability of credits and the level of use. Based on the theoretical analysis and the financial intermediar ies, the conclusion is that the credit risk affects the quality of financial engagement and partial GDP. The Kenyan banking sector's 2007-2005 regressive analysis confirms that access to banking resources reduces credit risk and stabilizes the banking sys tem and the level of accessing to consumer credits increases credit risk and threatens the stability of the banking sector. Bank deposit is an important part of money supplied by the bank. The growth of deposits is directly related to GDP growth. The outcome of the author will be changed for the banking system with other data because the level of financial engagement is the growth rate of GDP in the country. GDP growth is correlated with the level of employment.

METHODS
Jakubik, in the work "Macroeconomic Credit Risk Model", has proposed a credit risk assessment model that defines the share of outstanding loans in the credit portfolio. The approach of the author allows the influence of macroeconomic shocks to predict the influence of macroeconomic variables in the long run. The model is based on the normal distribution law and the accuracy is high when the variables in the risk of change are within the norm. Basel's recommendations and regulations of central banks ensure the risk of minimization in case of defaults of borrowers, but with the high range of macroeconomic factors lead to a credit risk formation. The model, which determines the risk factors, may be structurally the same, but the quality of the influence of factors is different in dif ferent countries. The influence of macroeconomic factors is particularly evident in developing coun tries. The article focuses on determining the im pact of macroeconomic factors on the credit risk of Georgia's System Bank. By using Pearson's cor relation to increase accuracy, the coefficients of in dividual macroeconomic indicators are calculated. The credit risk ratio is taken from the bank's over due credits and credit portfolio ratio.

RESULTS
The study of financial data of the Bank found that 7998% (2011-2017) of the bank's credit portfolio was credited to individuals.
The financial crisis caused by the unbalanced economic situation of the country (2008-2009) has caused military conflict to reduce the sol vency of individuals, reflecting the loss of jobs. The work deals with a correlation of credit risk due to overdue credits, based on the system "Bank of Georgia", with the macroeconomic fac tors and the coefficient obtained by the quality of influence as a result of growth factors in fu ture credit risk.
Pearson's correlation coefficient is used to de termine the risk of credit risk for macroeco nomic indicators. Pearson's coefficient de scribes the linear dependence between the vari ables, but in this case, it does not have a strictly linear dependency, but the advantage of the use of Pearson's correlation can be explained by the fact that the impact of the individual factor on credit risk can be determined. At the same time, it should be noted that the unemployment rate has not been realized and inflation is rising due to emerging factors outside the country, and their regressive multifactor model could not be realized. The use of Pearson's correlation meth od requires the following conditions: linear cor relation between variables, the subordination of the normal distribution of economic processes.
The authors guarantee that the change in the parameters is subject to the normal distribu tion law. The schedule of individual macroeco nomic indicators and credit risk connection (see Figure 1 in the Appendix) is designed to deter mine linear dependence.
The results of Pearson's correlation coefficient, the criterion of critical criteria are reflected in Table 1.
R x1y = 0.43 (unemployment) -under the inter val of the value of Pearson's correlation coef ficient, the correlation between the credit risk and the unemployment rate in the country during 2007-2017 is lower than average (see Figure 2 in the Appendix). The criterion of the personnel is determined to indentified the cor relation between the importance of the correla tion and was compared to the critical criterion (see Table 1). tr 1 = 0.9, tr Critical = 0.60, when p ≤ 0.05 and tr Critical = 0.74, when p ≤ 0.01, i.e. tr 1 = 0.9 > t Critical . As one can see, in spite of low cor relation, unemployment is statistically signif icant for the credit risk assessment caused by overdue.
R x2y = 0.62 (GDP ratio tr Critical = 0.60, when p ≤ 0.05 and tr Critic = 0.74, when p ≤ 0.01. Critical criterion terms, tr 2 > t Criticism , which implies that correlation is correct. In both cases, 9599% re liability is correlation with the tempo) is higher than average. Correlation criterion tr = 0.89, in this case t Critical = 0.60, when p ≤ 0.05 and t Critic = 0.74, when p ≤ 0.01. Thus, tr > t Critical and cred it risk. The GDP's influence level is true with 9599%.
R x3y = 0.08 (correlation with inflation rate) is weak. Correlation criteria tr 3 = 0,24, t Critical = 0.60, and t Critical = 0.74. The values confirm that the connection between inflation and credit risks is not accidental, but the weakest correla tion of the most sensitive macroeconomic indi cators is the question of how inflation is at risk of inflation weakening influence.
According to the authors, the National Bank of Georgia instantly launches its monetary lever age and changes the shortterm interest rate for the price stability, which will be reflected in the economy through the transfer mechanism, in particular through the credit transfer channelcredit card.
The increased rate of policy will be transferred to the money market, which will be reflected on a shortterm interest rate, and then affects the longterm interest rate. In these conditions, credit availability is difficult, low credit plans are reduced, companies are trying to get higher rev enues in highrisk investment projects, increased interest rates raise the credit service coefficient, thus the bank tightens the terms of the loan is suance and reduces the lending of the economy, which also leads to mitigating credit risk.
Although the formation of monetary policy is based on the target inflation rate, Georgia has the small and open economy and has no less im pact on the shaky processes developed in part ner countries, as a result of inflation forecasted by target inflation.
The correlation coefficients are calculated accord ing to the credit risk of most of the impact of the gross domestic product of the unemployment rate and the inflation rate, but in Table 1, the annual da ta comparison shows that the higher credit risk cor responds to the case of the gross domestic product at a high rate, in other cases, the lowest indicators I.
Delayed loans in the banking system by December 2016 are 36.6 mln. Increased by GEL Of these, 41.5% are denominated in the national curren cy, indicating a high level of dollarization. The growth rate of inflation is the lowest (1.8%) in 2016 from 2007-2017. It is also evident that periodically the impact of inflation is weak on credit risk and credit risk rises because overdue loans are caused by other factors, in particular, increased unem ployment, exchange rate, poverty level, inequality of incomes, wrong assessment of risks and more. As for the calculated correlation index (0.08), this is explained by the fact that the monetary policy of the National Bank of Georgia was planned based on the wrong forecasted indicator and, on the oth er hand, it became a shocking process.
The values of correlation can raise the equation of the linear function between variables, which will be able to determine the impact of the shock pro cesses based on the research rate.
where Y -the credit risk created by the overdue loan; X -macroeconomic indicator; β 1 -Pearson's correlation coefficient. Assuming that GDP reduction is expected to be 7%, the correlation between the correlation and the reliability of the credibility will be the credit risks caused by overdue loans: Thus, credit risk will be reduced by 3.33%. Is this a given situation for a commercial bank and what factors cause credit risk reduction? The GDP cut in itself means a weakening of economic activity, so the bank cannot afford credit and the risk of credit risk reduction (see Figure 3 in the Appendix).
Inflation increases despite the weak correlation, the cause of which is explained above, the cred it risk should be high. Suppose that the stresstest scenario was designed to increase the risk of rising inflation by 15%.
As one can see, inflation grows by 15%, it is ex pected to increase credit risk by 9.36%. The actions aimed at the shocks are "stirring" to the National Bank of Georgia and individual commercial banks and enable them to react to the expected events.
As the calculations show, there is an average corre lation between the levels of unemployment: In the case of unemployment increase by 8%, cred it risk will increase by about 3.94%. This is also natural, because the unemployment rate reduces the income of the population and therefore the bankruptcy of the loans. Correlation with unem ployment is not a valid indicator (see Figure 4 in the Appendix).