Thesis advisor: Peter Ireland / The literature on real business cycles finds that one reason why emerging economies are more volatile than developed small open economies is that they face greater financial frictions. Indeed, according to several measures of financial depth and access, financial systems in emerging countries are on average less developed than those in developed small open economies. Despite the lag in financial development, private credit, particularly unsecured credit to households, has been steadily increasing during the last two decades in emerging countries in Latin America. During this period of rising credit, various countries in the region observed an increase in the size of their middle income class population and the emergence of the vendor financing channel in their consumption credit market. Estimates by the World Bank suggest that the share of middle class households increased from 20.9 % in 1995 to 40.7 % in 2010. In addition, the share of poor households was approximately halved and reached 23.4 % at the end of this 15 year period. This phenomenon not only increased credit demand but also motivated the entry of new suppliers in the consumer credit market in countries like Mexico, Colombia, Chile and Brazil. In spite of a significant decline in unemployment in recent years, the lack of formal employment and poor credit history were still impeding many individuals from gaining access to consumer finance from traditional financial institutions. In order to enable new middle class shoppers access items typically offered by large retail stores, the retailers themselves started offering credit. In this dissertation, I study the relationship between middle class size, unsecured credit markets and aggregate consumption volatility in emerging countries. In the first chapter of this thesis, we examine the link between middle class size and consumption growth volatility using a sample of middle income countries. In the second chapter, we study the effect of an expansion of the middle class on vendor financing incentives and unsecured credit supply on its extensive margin. In the third chapter, I study business cycle implications of a reduction in the share of financially excluded households in an emerging economy. In the first chapter, I empirically examine the effect of middle income class size on consumption growth volatility in emerging countries. Using a panel data of middle income countries, I find that a larger middle class size tends to increase aggregate consumption growth volatility, particularly at lower levels of financial system depth. Financial development plays a significant role in determining the sign of the marginal effect of middle class size on aggregate volatility. Unlike emerging countries, the effect of the size of the middle class and the role of financial development on consumption volatility in developed countries is ambiguous. The key message of this analysis is that as more households escape poverty thresholds and reach the middle income class status in developing and emerging economies, it becomes more important to deepen financial systems from the perspective of aggregate consumption volatility. In the second chapter, I explore through the lens of a theoretical model, potential reasons triggering an increase in credit supplied by the non traditional financial sector, i.e vendors, at the extensive margin. I find that a reduction in the average risk of default and an increase in the market size of credit customers raise vendor financing incentives. This model rationalizes the observation that the improvement of economic conditions of the low-income and financially constrained households potentially led to increased credit supply by vendors in several countries of Latin America. In the third chapter, I study business cycle implications of a decline in household financial exclusion in a dynamic general equilibrium model suitable for emerging economies. Using Mexico as a case study, I estimate the model with Bayesian methods for the period 1995 to 2014. Standard measures of predictive accuracy suggest that the extended business cycle model with limited credit market participation outperforms a model with zero financial exclusion. The results of the estimation suggest that a rise in credit market participation in an emerging economy increases aggregate volatility of key macroeconomic aggregates, and that financial frictions play a key role in this relationship. I confirm this prediction by re-estimating the model for Mexico after splitting the sample into two non- overlapping decades. A key implication derived in this chapter is that a reduction of financial exclusion within an emerging country may lead to higher consumption growth volatility and trade balance volatility, and that fewer financial frictions dampen the marginal effect. As household financial access increases in these countries, a greater need for improving broad financial development measures arises.

The objective of this research is to investigate the existence of structural changes in unsecured lending time series data and analyse the impact thereof on trends in consumer demand for unsecured credit spanning the years from 2008 to 2015. This is achieved by identifying dates when structural changes occurred over this period. The identified structural break date is linked to an influential economic event or monetary policy change that took place in South Africa of which the impact on three unsecured credit categories are analysed. Unsecured credit growth in South Africa has been subjected to intensive scrutiny since the inception of the National Credit Act (Act No. 34 of 2005) by various regulatory bodies. In 2012 the National Credit Regulator (NCR) commissioned a research study into examining the impact that the National Credit Act (Act No. 34 of 2005) has had on the consumer credit market. The empirical part of this study involved the gathering of time series data on unsecured loans approved, unsecured credit granted per income category and unsecured credit granted from the National Credit Regulator (NCR) database and performing descriptive and econometric analysis. The Zivot-Andrews (1992) and augmented Dickey-Fuller tests determined the break dates which were linked to a significant economic event while the one sample t-test of means compared average loan values before and after the break date. Results of the study indicate that the break dates determined coincided with economic events and monetary policy changes in South Africa, such as the collapse of African Bank, the implementation of the National Credit Amendment Act, prime interest rate movements and the introduction of a debt counselling program by the government. These events, coupled with stricter lending criteria and no further loans being granted to customers already more than three months in arrears, restrained the uptake of unsecured loans to lower and middle income groups. The introduction of new affordability criteria and increasing interest rates in 2014 further negatively impacted demand for unsecured loans. However, higher income earners had the advantage of being able to apply for higher loan amounts. Findings could be used by monetary policymakers and financial institutions to constantly monitor credit trends, improve credit assessment techniques and review lending criteria. / Business Management / M. Com. (Financial Management)