Abstract
This study investigates the attractiveness of carry trade strategies in the Israeli FX market during the period 1/2003 - 12/2014. We examine several Carry To Risk (CTR) popular measures and propose a new measure which is based on a time-varying currency risk premium. The predictive capability of the examined CTRs is examined on foreigners' holdings in Israeli government bonds using a proprietary data set. In order to estimate the currency risk premium we also assess the Covered Interest rate Parity (CIP) and the Uncovered Interest rate Parity (UIP) hypotheses. We find that the CIP hypothesis did not hold during the second half of the sample period while the UIP hypothesis did not hold along the entire sample period for a one-month horizon, but prevailed for the twelve-month horizon. We find, using various methodologies, that our proposed CTR measure predicted foreigners' positions in government bonds during the second sub-sample. Such a prediction was more robust and of longer duration than other popular CTR measures.
