THE EFFECTS OF THE NATIONAL CREDIT ACT AND THE GLOBAL FINANCIAL CRISIS ON DOMESTIC CREDIT EXTENSION : EMPIRICAL EVIDENCE FROM SOUTH AFRICA

This paper examines the impact of the new National Credit Act (NCA) No. 34 of 2005 and the global financial crisis on credit extension provided by all monetary institutions in South Africa. The econometric approach is estimated by way of ordinary least squares while controlling for several macroeconomic factors. The findings indicate that there was a general increase in the consumer credit provision in the period subsequent to the full implementation of the Act. The promulgation of the Act increases credit card, bank overdrafts, other conventional loans and total credit to the private sector categories. The implementation of the Act fails to reverse this trend but exerts a negative influence on lease finance and the global financial crisis has significant negative effects on most of the credit provision categories. The paper seeks to investigate an under-researched area on the interrelatedness of credit provider regulation, financial crises and credit extension.


INTRODUCTION
Banking sector regulation seeks to ensure that private and collective interests are safeguarded, thereby promoting a sound and reputable domestic financial sector (Gully Hart, 2005).From the South African perspective, the process of credit provision has been under the spotlight.This is as a result of financial institutions' failure to uphold the principles of responsible lending, a situation that was deemed the prime cause of high levels of indebtedness.
While credit provision makes potential future income available for current spending, there have been concerns on the transparency issues in most consumer credit agreements facilitated by credit provision institutions.The resulting effect is devastating on the part of consumers, as they are faced with unanticipated interest costs and unforeseen contingencies.The new National Credit Act No. 34 of 2005 (NCA) was fully implemented on 1 June 2007 to prevent banks and other credit provision institutions from lending recklessly, a situation where the lender fails to conduct an affordability assessment, and to ensure that consumers of credit do not borrow more than they can afford.Furthermore, the NCA seeks to regulate the granting of credit through the National Credit Regulator (NCR), which facilitates the role of a national consumer tribunal, and as a debt counselling service.It was hoped that through adherence to the new regulation, consumers would be educated to make informed financial decisions.The NCA applies to credit cards, overdrafts, mortgages, instalment agreements, leases and micro loans.Following the promulgation of the NCA, there have been a number of activities that credit providers have engaged in to comply with the stipulated regulation.The various restrictions of the NCA and its other provisions may necessitate a change in the lending patterns of South African credit providers.
Another issue that warrants attention relates to the global financial crisis of 2008 which adversely affected the creditworthiness of borrowers and the appetite of financial institutions for advancing credit.The effects of the global financial crisis on bank lending behaviour have not been adequately documented.Having said this, the aim of this paper is to establish whether the NCA and the subsequent global financial crisis have impacted significantly on credit provision by South African monetary institutions.This paper documents evidence of an increase in credit extension for the period prior to the full implementation of the Act for the "other loans and advances" category, which includes credit cards, bank overdrafts and other conventional loans.The total credit to the private sector category is also positive and statistically significant.The implementation of the NCA fails to reverse this trend, with the exception of lease finance, which involves expensive capital intensive equipment.The global financial crisis has had significant negative effects on all categories of credit, with the exception of mortgage finance, which is negative, but statistically insignificant.
The rest of this paper is organised as follows: section two provides a theoretical background of the study.Section three deals with the research methodology by providing an overview of the data and variables used and developing the estimation model.Section four reports the descriptive and empirical results, and section five concludes the paper.

New reforms under the new National Credit Act
According to the NCA No. 34 of 2005, credit providers are required to provide a quotation to credit consumers specifying the full detail of the cost to the consumer.The quotation will be valid for five days, giving ample time to the consumer to receive quotations from competitors should they wish to do so.The consumer is required to provide full and correct information regarding his or her debt position, failing which the provider of credit can sue for failure to do so.The opposite also applies if the provider fails to provide correct information regarding the quotation.Previously, providers of credit would be satisfied if the loan repayments did not exceed 30% of the consumer's income.
With the new NCA, a lot more credit assessment procedures are required.The Act now prohibits cold calling by credit providers and the automatic extension of credit limits to consumers.The Act further seeks to protect consumers from misleading advertisements.The inclusion of phrases like "guaranteed loans" and "no credit checks" are dealt away with.The Act provides for the agreement to be included in the local language the consumer is conversant with.The Act further seeks to root out discriminatory assessments of the borrower within the assessment criteria by giving reasons for refusal of the credit within the risk management criteria of the credit provider.

Rationale for bank credit regulation
As the South African economy is growing, there has been rising concern about an increasing number of vulnerable consumers who have little protection from falling into the trap of indebtedness.For example, Braunsberger, Lucas and Roach (2004) found that consumers generally are not very knowledgeable about credit cards.Their report recommends that the issuers of debt should start to educate consumers about the product they are selling.In addition, market saturation has led financial institutions to target non-traditional customers such as university students and low-income groups in order to incorporate the masses into the banking mainstream.These consumers have very little experience in evaluating complex and competing product offerings and often have no credit history (Salomon, 1998).There remains a potential risk for banks to exploit this market.
The 1968 Truth in Lending Act of the United States, the 1974 Consumer Credit Act of the United Kingdom and legislation in many other countries advocate greater transparency regarding the cost information inherent in consumer credit products.For example, the Swedish Consumer Credit Act of 1977, which was subsequently revised in 1993, requires that a lender providing credit should include two cost indicators -that is, the annual percentage rate (APR) and the total credit charge (Yard, 2004).Indeed, the cost of the loan calculated in monetary terms is a substantial element of the repayment by the consumer.It is therefore imperative for the credit providers to accurately disclose in a transparent manner the cost implications to consumers so that consumers can make informed decisions regarding various competing products.The 2004 amended Consumer Credit Act (Advertisements) of the United Kingdom further supports this by stipulating the use of plain and understandable language so as not to confuse the consumer.The NCA has adopted this approach by requiring credit providers to disclose information in the language the consumer is comfortable with.
Ironfield-Smith, Keasey, Summers, Duxbury and Hudson (2005) conducted a survey to test consumer attitudes to credit limits.The survey revealed that 41% of the respondents disagreed that lenders follow a scientific approach in determining credit limits.45% disagreed that lenders will only lend money if they believe the person can repay the debt, whereas 61% agreed that lenders will lend anyone money whatever the person's circumstances because they want to make profit.A key question was also asked to determine the reason why respondents fall into debt predicaments that they cannot manage.51.5% cited irresponsible lending as a main contributing factor.In a similar study, Ralston and Wright (2003) found that credit union managers in Australia fail to impose stringent measures on high risk borrowers.This gives a clear indication of the perceptions that consumers have regarding the attitude of lenders towards protecting the consumer.Prior to the introduction of the NCA, the lending situation in South Africa somewhat corresponded to these views as credit limits were extended automatically.In the months leading up to the implementation of the NCA, economists believed the high debt levels despite recent hikes in the interest rates were as a result of banks advancing as much debt as they could before the Act was implemented (Stovin-Bradford & Shevel, 2007).As shown in FIGURE 1, the household debt to disposable income ratio was increasing steadily in the period running up to the implementation of the Act.These high levels of indebtedness suggest that households in South Africa are living beyond their means, and government needs to take swift action to support consumers in making informed decisions about their credit preferences.

Possible effects of the new National Credit Act
The implementation of the NCA could have spearheaded an expansionary effect on certain sectors of the economy, while a contraction of other sectors cannot be ruled out.For example, under the new provisions of the NCA, consumers with no car allowance are no longer required to secure their vehicle with a deposit.As a result, more people who previously could not afford a deposit to apply for credit -especially those from low-income groups and recent graduatescan now have access to vehicle finance.Furthermore, institutions that are not registered as credit providers are no longer allowed to provide credit with interest repayments to their employees.This could have had an expansionary effect on the credit extension of banks and other credit providers.By contrast, the stringent measures required on the part of the banks to examine the creditworthiness of the consumer before any funds can be paid out are a matter of concern.On the balance of probabilities, there could have been a slowdown on credit provision in the short run, as the industry came to terms with the NCA.As the NCA applies to the whole banking industry, the playing field has been levelled, thereby ensuring fair competition through one unifying Act.Given the rise in interest rates in the period preceding the financial crisis, the long run impact of the NCA is uncertain.
As shown in FIGURE 2, there is some prima facie evidence that the pre-financial crisis rise in interest rates failed to slow down the provision of credit.However, it is now evident that the residential property sector has been dampened by the NCA.The residential property barometer released by First National Bank (FNB) has revealed that activity in the residential property market for the fourth quarter of 2007 was at its lowest point ever since the survey was instituted in 2003 (First National Bank, 2009).The research further found that there was a decrease in the number of first time property buyers, from 32% in 2005, to 16% in July 2007.Then again, a number of factors need to be considered: the credit crunch that existed in several countries in the early 1990s presented a unique environment to test for the effects of regulation on bank lending (Shrieves & Dahl, 1995).It was quite evident that the resulting credit restriction was not a result of a stand-alone factor, but due to several other reasons that could not be ruled out.
These factors include tightening capital regulations, substantially depleting bank capital and changing risk attitudes (Vihriala, 1996).While the NCA implementation was intended to address the risk attitude problem on the part of both the lenders and the borrowers, this may not be a determining "heavyweight" factor that could slow down the lending patterns.In this regard Pazarbasioglu (1996) suggests other reasons to conservative lending, citing deteriorating asset quality and reduced profits.Although these arguments may not be unique to South Africa, the impact of the Act on a stand-alone basis remains to be seen.
In the early 1990s, banks in the United States also curtailed their lending.Sharpe (1995) attributes this to loss of bank capital, strict bank regulatory standards and heightened market scrutiny of capital.It appears therefore that while regulatory requirements are an essential element to reduced spending, one cannot rule out the significance of other factors on the impact of credit provision.For example, a 1992 survey of senior bank officials by the ABA Banking Journal revealed that regulatory pressure was, in part, a factor that contributed to a restriction in bank lending (American Bankers Association, 1992).Shrieves and Dahl (1995) note that the pressure was apparent in credit provision for highly levered transactions, including real estate finance.The former Federal Reserve Board Chairman, Alan Greenspan, attributed regulatory overreaction and excessive regulation to the credit contraction of 1992.
It further appears that capital adequacy issues remain the prime element as a possible factor in hampering credit provision.These results have also been confirmed by Peek and Rosengren (1995).

Data and variables
Monthly time series data for the period 1 January 2005 to 30 September 2010 was obtained from the South African Reserve Bank.To model the effects of the implication of the Act on the various categories of domestic credit, the credit extension variables are regressed on a set of control and dummy variables.The dependent variables are the monthly change in the logarithm of Instalment Sales Credit, Leasing Finance, Mortgage Advances, Other Loans and Advances, Total Credit to the Private Sector (TCPS) and Total Domestic Credit (TDC).
The control variables include monthly real interest rates, inflation, real effective exchange rate and real Gross Domestic Product (GDP) spanning the period 1 January 2005 to 30 September 2010.These macroeconomic variables are included because they are likely to affect the extension of credit by South African monetary institutions.The REAL INTEREST variable is represented by the prime overdraft rate adjusted for inflation.The INFLATION variable is represented by the monthly Consumer Price Index figures.The REAL EXCHANGE variable is captured by the logarithm of the real effective exchange rate of the weighted average of 15 currencies.The REAL GDP variable is calculated as the logarithm of real GDP at constant prices, with January 2005 as the base year.The GDP figures are divided by three in order to obtain monthly estimates.
Since the NCA was implemented in phases, the methodology adopted was to identify the lending patterns in two phases.The first phase is the period between the promulgation of the Act and its full implementation on 1 June 2007.This first phase period extends from 1 January 2005 to 31 May 2007.During this period, the main administrative provisions such as the establishment of the National Credit Regulator (NCR) and registration of credit providers on the NCR database were implemented.Data during this period is also analysed to establish trends and compare these with the pre-implementation period.The second phase is the most important, since this paper aims to establish whether the NCA and the global financial crisis have impacted significantly on credit provision by South African monetary institutions.This second phase period is from 1 June 2007 onwards.
To capture the effects of the implication of the Act, two dummy variables are used.NCA1 takes on the value of one for the phase one period between the promulgation and the implementation of the Act, and zero otherwise.This dummy is intended to capture the announcement effect of the Act.NCA2 takes on the value of one for the second phase period after the implementation of the Act, and zero otherwise.This variable captures the effect of the implementation of the Act.To capture the effects of the global financial crisis, a dummy variable (CRISIS) that takes on the value of one is used for the period beginning September 2008 to December 2009, and zero otherwise.

Econometric approach
The proposed model is estimated using the following general specification: The Heteroscedasticity Consistent Covariance Matrix (HCCM) proposed by White (1980) is commonly used to control for heteroscedasticity, and relies on the asymptotic version known as the HC0.However, Cribari-Neto (2004) argues that HC0 may be biased in finite samples.A variant of the HC0 estimator known as HC3 is utilised by Long and Ervin (2000), who argue that it is more suitable than other related estimators for smaller samples.Based on these observations, the HC3 estimator is utilised to control for heteroscedasticity.The robust standard errors which relax either or both of the assumptions of independent and identically distributed standard errors are estimated and used in the analysis of results.
Autocorrelation is tested by using the Durbin Watson d statistic.Serial correlation is controlled by running the Prais-Winsten AR(1) regression, which transforms the model into serially uncorrelated classical disturbances.This procedure was performed for the MORTGAGE, HOUSEHOLD, OTHER, TCPS and TDC regressions.The other regressions showed no evidence of serial correlation.

Descriptive results
It can be seen from FIGURE 2 that there is a general increase in the credit extension figures for all categories of debt.This increase is observed for the period leading up to the global financial crisis of 2008.A levelling off is observed around the period of October 2008, one month after the collapse of Lehman Brothers.Overall, it appears that the implementation of the NCA does not seem to cause a structural shift in the credit extension time series.However, it is important to observe the impact of the NCA while controlling for other factors such as real interest rates, real GDP, real effective exchange rate, and the inflation rate.

Macroeconomic effects
The regression results for the empirical model are reported in TABLE 3. The REAL INTEREST variable is statistically significant for the mortgage and household categories of credit.An increase in the real interest rates causes mortgage credit and credit to the households to increase significantly.The coefficient for both the variables is significant at the 10% level.The REAL GDP variable appears to have a significant and inverse influence on credit in most of the categories of credit.This negative association suggests that the real growth in the economy did not support further credit extension to the various sectors of the economy.A closer inspection of the correlation matrix (not reported) of the independent variables shows that the REAL GDP variable is positively correlated to the CRISIS variable, with a correlation coefficient of 0.38.This positive correlation suggests that the slowdown in the economy was influenced by the global financial crisis, thereby causing a negative influence on the various categories of credit.
There is mild support for the effects of real effective exchange rates on credit.The ISC, MORTGAGE and TDC categories of credit are negatively affected by a devaluation of the Rand against the weighted average of 15 currencies.The strength of the correlations is significant at the 10% level.This observation suggests that the weakening of the Rand against major currencies may cause inflationary pressures, thereby reducing affordability for instalment sales, mortgage credit and total debt.The MORTGAGE and HOUSEHOLD credit categories are directly correlated with the INFLATION variable.The coefficients are positive and significant at the one and 10% levels respectively.The measure used for inflation is the CPI, which includes interest on mortgage bonds for the period prior to 1 January 2009.As mortgage costs increase, the CPI measure also increases; hence a positive association is expected.

The impact of the NCA and the global financial crisis
As shown in TABLE 3, the instalment sales credit category, which includes motor vehicle finance, is mildly affected by the implementation of the Act.The NCA2 variable is positive and significant at the 10% level.This implies that the Act had a positive impact on instalment sales credit provision.This is not surprising, as from June 2007 a deposit was no longer required to secure a motor vehicle loan.Furthermore, institutions that were not registered as credit providers were no longer allowed to provide loans with interest payments to their employees.
This restriction could have caused employees of non-registered credit providers to seek finance from other registered credit providers such as banks.Given these two observations, it is reasonable to expect an increase in the instalment sales credit variables.The global financial crisis appears to have had a strong and significant negative influence on the provision of instalment sales credit.The associated coefficient is negative and significant at the 1% level.
The LEASE variable is negatively affected by the implementation of the NCA.The coefficient of the NCA2 variable is negative and significant at the 5% level.Because lease finance involves costly equipment such as aircraft and other capital intensive equipment, it is reasonable to find this inverse association.The implementation of the NCA reduced the affordability of capital intensive equipment.A stronger negative correlation is observed between the CRISIS variable and leasing finance.The coefficient is statistically significant at all conventional levels.This negative association suggests that the global financial crisis reduced the demand for leasing finance for capital intensive equipment.
Mortgage loans increased during the period prior to the implementation of the Act.The associated coefficient is positive but insignificant.The implementation of the NCA exerts a positive influence on mortgage advances, but the associated coefficient is still insignificant.Furthermore, the household category, which comprises loans to households, was adversely affected by the global financial crisis.The CRISIS variable is negative and significant at the 5% level.
The OTHER loans category, which comprises bank overdrafts, credit cards and other conventional loans, is also impacted significantly by the announcement effects of the Act.The coefficient of the NCA1 variable is positive and significant at all conventional levels.This direct correlation may indicate that banks were writing up as much debt as possible in the period preceding the implementation of the Act.The same direct association is revealed for the NCA2 variable, although the relationship is significant at the 5% level.The global financial crisis exerted a significant negative influence on the OTHER loans category, suggesting a slowdown in the economy.
The announcement effects of the Act seem to have a positive and significant impact on the total credit to the private sector category.The coefficient of the NCA1 variable is statistically significant at the 5% level.This strong correlation suggests that the promulgation of the Act had a significant positive effect on total credit to the private sector.The implementation of the Act also contributed positively to the total credit to the private sector.The associated coefficient is significant at the 5% level.The global financial crisis reversed this trend by contributing negatively to the TCPS variable.The correlation is negative and significant at the 1% level.
The TDC category includes loans to both the private and the government sector.No significant correlations are observed with the NCA1 variable, although the NCA2 variable is statistically and positively correlated to the total domestic debt category.The global financial crisis had a negative influence on the TDC variable, although the coefficient is insignificant.These results suggest that the private sector was more affected by the banking regulation and the global financial crisis.Overall, it appears that the announcement of the promulgation of the NCA had a significant positive impact on the OTHER and TCPS categories, suggesting that credit cards, bank overdrafts and private sector credit categories were the most actively issued debt categories.
Owing to the fact that lease finance involves capital intensive equipment, the actual implementation of the Act slowed down the leasing finance category only.The global financial crisis exerted significant pressure on most categories of credit.

RECOMMENDATIONS FOR FURTHER STUDY
The purpose of this paper was to document the effects of the NCA and the global financial crisis on various forms of credit to the South African private and public sector.Future research could explore the effects of the global financial crisis, the NCA, and possibly other pieces of legislation on the efficiency of financial institutions.Furthermore, additional variables could be introduced to examine the effects of the NCA on the operating performance of financial institutions.Lastly, the relationship between the NCA and the creditworthiness of South African households could possibly shed more light on the effectiveness of the NCA.

CONCLUSION
This article has examined the impact of the promulgation and implementation of the new National Credit Act No. 34 of 2005 and the global financial crisis on credit provision by all monetary institutions in South Africa.The evidence emerging is that the promulgation of the Act caused monetary institutions to increase their lending for the OTHER loans category, which comprises credit cards, bank overdrafts and other conventional loans.The total credit to the private sector category also increased with the promulgation of the Act.Leasing finance has been adversely affected by the implementation of the Act.The global financial crisis has had a strong negative influence on most of the categories of credit.However, the total domestic credit category, which comprises loans to both the private and government sectors, appears to be less affected by the implementation of the Act and the global financial crisis.

FIGURE 1 :
FIGURE 1: Household debt to disposable income, 2005 -2010 Source: Authors' compilation based on statistical output from Stata version 11

FIGURE
FIGURE 2: Credit extension by all monetary institutions, 2006-2010 Source: Statistical output

TABLE 1 .
1   +  2   +  3   +  4   +  5 1  +  6 2  +  7   +   REAL INTEREST, log REAL EXCHANGE and INFLATION time series have a unit root problem.Because of this, regressions with these variables are estimated in first differences.To detect a possible multicollinearity problem, the variance inflation factors (VIF) for the independent variables are reported in TABLE 2. A VIF above 10 may indicate a multicollinearity problem.The mean VIF is 4.83, indicating that multicollinearity may not be a concern.Regressions were run by dropping the INFLATION variable, which has a VIF of 12.8, and the results are similar.
t dummy variable that captures the announcement effects of the NCA at time t

TABLE 1 : Results of the Augmented Dickey Fuller unit root test
Source: Statistical output