Gumata, Nombulelo2020-12-062020-12-062020https://hdl.handle.net/10539/30348A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the Faculty of Commerce, Law and Management in the School of Economic and Business Sciences, Wits Business School, University of the Witwatersrand, Johannesburg, 2020This thesis consists of six self-contained empirical chapters that deal with different aspects of how global liquidity is transmitted into the South African economy. The chapters link excess global liquidity and risk shocks channels of transmission into the South African economy via aggregate and components of capital flow effects on domestic assets, money demand and credit growth. In addition, they link asset price cycles, financial conditions, labor market conditions and policy uncertainty to developments in household credit growth and household debt growth and how these affect the macro-economy. The chapters employ empirical approaches that range from the factor augmented sign restricted vector autoregressions (FAVAR) to properly identify structural shocks, threshold vector autoregressions (TVAR) to assess the asymmetric effects and nonlinearities induced by the thresholds and vector error correction models to estimate the mechanisms that assist the economy to adjust from deviations from the long-run (equilibrium path). The results in Chapter two show that all shocks to price and quantity measures of global liquidity lead to a significant increase in GDP growth and credit growth. However, inflation declines in response to price and quantity measures of global liquidity shocks consistent with the appreciation of the exchange rate exerting downward pressure on inflation. In addition, evidence in this chapter shows that the quantity measures of global liquidity, in particular asset purchases lead to higher peak responses in GDP growth. The inclusion of Japan data in the construction of the global liquidity factors results in higher peak responses in credit growth and GDP growth. Asset purchases by the Bank of Japan added to excess global liquidity and contributed to easier domestic credit and financial conditions. The responses of house price growth and equity price growth suggest that the effects of price and quantity measures of global liquidity result in much more pronounced effects on house price growth and equity price growth. Last, evidence in this chapter shows that the decline in global policy rates and investor risk perceptions coupled with an increase in the quantity measures of global liquidity loosened domestic financial conditions. The yield on the ten-year domestic government bond, the South Africa-United States (SA-US) yield spread, domestic sovereign credit risk across the five-year credit default swap (CDS) spreads and emerging market bond index (EMBI) spreads declined. This indicates that the lowering of global interest rates and the asset purchases loosened domestic financial conditions and the fiscal constraints. In a nutshell, we establish that global risk, monetary and financial conditions do spill-over and are transmitted into the South African economy via several channels. Unconventional monetary policy tools do indeed complement and reinforce the effects of conventional monetary policy interventions. Thus, our findings concur with several empirical studies in this area showing that unconventional monetary policy tools are a necessary addition to the monetary policy toolkit. Furthermore, unconventional monetary policy tools have implications for fiscal policy, exchange rates, monetary and financial stability mandates of central banks. Chapter three presents evidence that across various specifications and capital inflow categories, capital inflow shocks are more persistent and induce more variability in credit growth, house price growth, equity price growth, investment growth, commodity price growth and the current account. The results in this chapter confirm that capital inflows shocks have persistent effects compared to monetary policy shocks. The data confirms the “savings glut” hypothesis, capital inflows have larger and more persistent shock effects on investment growth, commodity price growth and the current account. Furthermore, the results show that the composition of capital inflows matters. Credit growth and house price growth are more responsive to equity inflows compared to debt inflows. Capital inflow shocks, in particular, equity inflows shocks, were the key contributors to increases in commodity price growth during the 2004 commodity price boom. House price growth, equity price growth and commodity price growth present additional important channels through which accommodative monetary policy and positive capital inflow shocks are transmitted via different categories of capital inflows to domestic credit markets. Therefore, we conclude that capital inflow shocks are more persistent compared to monetary policy shocks. Capital inflows, in particular equity inflows, induce more variability credit growth, house price growth, equity price growth, investment growth, commodity price growth and the current account. Although monetary policy shocks are transitory, by loosening the credit and financial constraints, they contribute to changes in sentiment and domestic supply and demand factors. The interaction of monetary and capital flow shocks affects asset price growth and real economic activity with implications for the exchange rate, price stability and financial stability mandates. Chapters four and five find that real money balances are positively linked to real income and negatively linked to the interest rate spreads as predicted by theory. Augmenting the money demand function with asset price returns and financial and housing wealth variables, yields the desired results of lowering the point elasticity for GDP in the money demand function. This means that the augmented money demand functions are much more able to explain developments in monetary aggregates via the substitution and portfolio balance effects. The cointegration relations derived from the money demand functions augmented with asset price returns, interest rate differential, financial and housing wealth variables, are able to pick up the role of asset price returns and the interest rate differentials post-the financial crisis which is absent in the conventional money demand function as it still indicates a substantial and persistent money demand short-fall. We conclude that the inclusion of asset returns, interest rate differential, and financial and housing wealth variables in the behaviour of the money demand function explains the effects of constant structural changes and financial innovation in the South African economy. In addition, developments in asset price dis-equilibria can undermine macroeconomic stability. We therefore conclude that there is valuable, and policy relevant information contained in the augmented money demand functions. Furthermore, evidence in Chapter five reinforces the idea that financial cycles take longer to adjust. This is contained in the evidence of the slow speed of adjustment in the money demand disequilibria in models augmented with real wealth and asset prices. This means that the money demand disequilibrium takes longer when considering the role of real financial wealth than is the case under the conventional money demand function. The inclusion of wealth and asset price variables matters for the economic and policy interpretation of the money demand function. The impact of positive shocks to real asset prices and real wealth, in particular equity prices, results in a higher decline in the demand for real money balances compared to, for example real house prices. From the interaction of monetary policy and financial stability perspective, the policy implication is that the augmented money demand function can capture aspects related to asset price inflation and wealth channels of the monetary policy transmission mechanism. Chapter six estimates a financial conditions index for South Africa and uses it in the empirical estimations. We find that the constructed the financial conditions index (FCI) is positively related to lending conditions, meaning that tight financial conditions imply tight lending conditions. In addition, the FCI is negatively related to the consumer confidence and business confidence indices. Tight financial conditions lead to lower consumer and business confidence. Heightened policy and economic uncertainty as captured by the economic policy uncertainty index, leads to tight financial conditions for prolonged periods. Evidence shows that the loan supply and aggregate supply shocks played a dominant role in the evolution of total and sectorial credit growth. However, post2009 aggregate supply shocks reinforced by loan supply shocks played a dominant role in depressing credit growth. Furthermore, tight financial conditions and heightened economic policy uncertainty shocks interact and are mainly transmitted via aggregate demand shocks as well. Tight financial conditions and heightened economic policy uncertainty shocks play a significant role in propagating adverse loan supply shocks and have depressing effects on aggregate demand and credit growth. Furthermore, we establish the existence of the balance sheet channel of monetary policy via the housing market. Mortgage advances credit growth declines more due to a contractionary monetary policy shock and house price growth declines more in response to adverse aggregate supply shock. The implication of these results is that credit growth to the private sector declines more mainly due to the deterioration in balance sheets of households and the non-performing loans (NPLs) of banks compared to the shock effects of the decline in the demand for credit. This does not mean that there the absence of the traditional monetary transmission mechanism. Rather, that the response of mortgage advances credit growth, house price growth and NPLs suggests that, a contractionary monetary policy shock is propagated via the deterioration in the collateral values and the borrowers’ net worth impact on the loan supply channel. The policy implication is that because housing plays an important role in the business cycle as a component of investment growth, wealth and demand via the consumption channel, policy interventions that support the banking system as well as the real economy need to be prioritized. Chapter seven finds that adverse aggregate demand and loan supply shocks are mainly transmitted via the labor markets and are propagated by credit conditions. The simultaneous occurrence of positive employment growth, labor productivity growth and lower credit risk as captured by the non-performing loans shocks loosens the credit constraints and allows banks to supply more credit to meet aggregate demand. Furthermore, employment creation results in immediate responses in the household gearing as income shocks emanating from the labor market dynamics have an important relationship with credit growth, NPLs and the household debt growth burden in general. The vector error correction model (VECM) error correction terms suggest that currently household debt is in shortfall by as much as 10 per cent. The rate at which household debt growth adjusts from the disequilibrium position is terribly slow at less than one per cent per quarter. The transition functions show that the household debt growth threshold occurs between 8.8 per cent and 13.4 per cent. Empirical evidence in this chapter shows that periods of high household debt growth regimes, GDP growth and employment growth are persistently high compared to periods of low household debt growth regimes. The implications of the findings for the price and financial stability mandates are that household debt growth levels matter for GDP growth. The channels of transmission are largely via the labor market and are amplified by prevailing credit conditions. The loan supply shocks act as amplifiers of aggregate demand shocks and induce asymmetric effects on the macro-economy. Therefore, both credit demand and supply hypothesis are confirmed by the data. But the credit supply hypothesis seems to have dominated the evolution of household debt growth pre-2009. Furthermore, loan supply shocks were propagated by heightened credit risk as NPLs made negative contributions to household debt. Increases in labor productivity growth, GDP growth, employment growth and household debt growth are the key variables that assist in the adjustment of household debt growth from the dis-equilibrium positions. In addition, the decline in NPLs contributes more to the adjustment of household debt growth from the dis-equilibrium compared to the prime rate. This is because heightened credit risk limits the ability of banks to supply credit and widens the lending rate margins. The household debt growth dis-equilibrium positions, thresholds and the transition functions between the low and high regimes matter for the conduct of macro-prudential, monetary, and financial stability policies. The household debt growth dis-equilibrium positions convey information about the state of credit demand driven inflationary pressures, the concentration and build-up of financial risks. In addition, the household debt growth dis-equilibrium positions act as useful guides for the conduct of macro-prudential, monetary, and financial stability policies, over and above guiding the conduct of fiscal and macro-economic policy interventions.enThesis