[pt] O presente estudo busca entender os principais determinantes da flutuação de preços do Bitcoin através de variáveis relacionadas à força de mercado, tecnologia, reconhecimento público e variáveis macroeconômicas, estimando os coeficientes do vetor de correção de erros (VECM) e do procedimento autoregressivo de defasagens distribuídas (ARDL). Os resultados apresentaram que o reconhecimento público não possui impactos significantes sobre o preço de mercado do Bitcoin, enquanto as forças de mercado, fatores tecnológicos e as variáveis macroêconomicas apresentam impacto significativo em pelo menos um dos modelos utilizados. / [en] The present study seeks to understand the main determinants of Bitcoin price fluctuation through variables related to market forces, technology, public recognition and macroeconomic variables, estimating the coefficients of the error correction vector (VECM) and the autoregressive procedure of distributed lags (ARDL). The results showed that public recognition does not have significant impacts on the market price of Bitcoin, while market forces, technological factors and macroeconomic variables have a significant impact on at least one of the models used.

This study examines the indirect linkages between foreign direct investment (FDI) and economic growth in South Africa utilising 36 years’ (1980-2016) time series data obtained from the South African Reserve Bank (SARB). South Africa’s economy has been experiencing unsteadiness in recent years. Despite the government’s execution of different strategic initiatives to draw in FDI into South Africa, the country’s FDI remains lower than that of other emerging economies. Domestic investment by government, public corporations and the private sector is also relatively unsteady. Slow economic growth has put tremendous weight on the government to borrow externally for developmental purposes. This study tests two models – model I and model II. In model I, real GDP per capita (RGDP) is the dependent variable and foreign direct investment (FDI), domestic investment (DI), real exchange rate (EXR) and foreign debt (FD) are modelled as explanatory variables while in model II, FDI is the dependent variable and RGDP, DI, EXR and FD are modelled as explanatory variables. Domestic investment is sub-divided into credit to the domestic private sector (CPS), public investment (PI) by public corporations and government investment expenditure (GOVIN). The analysis of the relationship was carried out using econometric methods such as the Augmented Dickey-Fuller (ADF) and Phillips Perron (PP) unit root tests to identify the order of integration of the variables. The bounds cointegration test was applied to establish the long-term association among variables. The Autoregressive Distributed Lag (ARDL) model was utilised to test the long-run and short-run equilibrium conditions. Diagnostic tests were employed to check the model adequacy and the Granger causality tests were utilised to establish the causal relationships among variables. The discoveries from the ADF and PP tests uncovered that all the variables are non-stationary at level but became stationary at first differences. The bounds tests suggest that there is a long-run relationship and cointegration between variables. Following the presence of cointegration, the outcomes from ARDL model uncovered that FDI, CPS and GOVIN have a positive relationship with RGDP in the long run (crowding-in effect), while, a negative relationship occurs between PI, FD, EXR and RGDP in the long run (crowding-out effect) in model I. In model II, the outcomes revealed that RGDP, CPS, and PI have a positive relationship with FDI in the long run (crowding-in effect). Then again, the outcomes presented a negative connection between GOVIN, FD and v © Pamba, D, University of South Africa 2020 EXR to FDI in the long run (crowding-out effect). The short-run estimate of the coefficient of the error correction term (ECM) in model I and model II are statistically significant and negative. The negative indication of the error correction term shows a backward movement towards long-run equilibrium from short-run disequilibrium. In model I, the short-run coefficient results uncovered that FDI, lagged PI and lagged EXR are positively linked with RGDP (crowding-in effect). Then again, lagged CPS and lagged GOVIN are inversely related to RGDP (crowding-out effect). In model II, the short-run coefficient of FDI is certainly related to GOVIN (crowding-in effect). FDI, on the other hand, indicated a negative relationship with PI in the short run (crowding-out effect). The Granger causality tests for the variables uncovered a unidirectional causal connection running from RGDP to FDI and from FDI to RGDP in both models. The outcomes obtained for RGDP and FDI models pass all the diagnostic tests on serial correlation, normality and heteroscedasticity. The test for adequacy performed on the residuals demonstrates that they are homoscedastic and have no serial correlation, signifying that the model is acceptable. The Cumulative Sum (CUSUM) tests show that the extracted models are structurally steady and remain within the 5 percent level of critical bounds. / Economics / M. Com. (Economics)