Twin Deficit Problem and Feldstein – Horioka Hypothesis in Turkey: ARDL Bound Testing Approach an Investigation of Causality

The purpose of this study is to investigate the twin deficit problem for the Turkisheconomy using bounds testing approach known as ARDL (autoregressive distributed lag)to cointegration method and Toda and Yamamoto (1995) Granger causality test. The studycovers the annual data from 1974 to 2010. The cointegration test results suggest that thevariables are moving together in the long run. The positive value on the budget deficitcoefficient implies that Turkey has a twin deficit problem. The presence of twin deficit hasalso been supported by the Toda-Yamamota causality test results. As expected, the negativeand less than 1 investment coefficient indicates Feldstein-Horioka hypothesis holds.Further, it indicates that Turkey could not well-integrated into the international capitalmarkets with the model revealing that one fifth of the investments are financed throughforeign savings