An Econometric Investigation into the Determinants of Disaggregated Capital Inflows into Developed and Emerging Market Economies
Description
The sample data set covers 92 countries – 65 Emerging market economies and 27 Advanced market economies, over the period from 1990 to 2015. Due to the data availability, some variables may have more observations than others . Since this data is for capital inflows – that is purchases of home assets by overseas residents less sales of home assets by overseas residents in the same period - the source countries are primarily, although not exclusively, the advanced economies, to other advanced economies or to the so-called emerging market economies.
Files
Steps to reproduce
A general model of international capital flows can be postulated based loosely on the portfolio balance approach which can include both domestic demand-pull and global-push factors. We use three different types of capital flows, direct flows, portfolio flows and other flows and as well as Total flows. Rather than emphasise the geographical sources of the variables we chose to emphasise their economic and institutional characteristics. We have used the following portfolio balance model where K, the vector of the inward stock of financial capital, depends on three essential characteristics as represented by the vectors of: returns, R, risk, D, and scale, S. The model to be estimated is Y=f(FO, TO, FD, PR, VIX, r, r*, RER, GDP, WGDP, CR, DB, MS, D(crisis)) Where Y are the measure of the capital flow, institutional features are country-specific factors such as the degree of trade openness, measured as net exports to GDP (TO), the extent of financial openness (FO), and the extent of financial development (FD). Also there is political risk, PR, the global stock market volatility index (VIX) , and a crisis dummy for 2008-09, D(crisis).The real variables are simply competitiveness (the real exchange rate, RER) the rates of domestic and world growth, (GDP) ̂ and (WGDP) ̂ respectively. The nominal variables, which reflect scale are the growth in the money stock, (MS) ̂, and credit available to the private sector, (CR) . R and r* are the domestic and foreign interest rate. The estimation is performed in two steps using Stata. In the first step, the equation is estimated for all three types of capital inflows – direct, portfolio, and other – for each of the advanced economies, the emerging market economies countries and total aggregates, using ordinary least squares (OLS), fixed effects (FE) and random effects (RE) estimators. Then two specific exclusion restrictions are tested. The insignificant variables dropped from the first stage are based on the non-rejection of the null hypothesis that the coefficients are jointly significant at 5% level of confidence, according to the F-test (for FE model) or Wald test (for RE model). Testing for error interdependence to justify the more efficient SUR procedure, using the Breusch Pagan test for error independence is performed on the correlation matrix residuals.