Forward premium anomaly in non-deliverable forward market

Abstract

Study aims to provide new evidence to existing academic literature on forward premium anomaly by examining the subject in the non-deliverable forward market. First hypothesis is created to examine and observe the unbiased forward rate theory functionality in sample data. Fama regression model is applied to test the hypothesis with 12 month over-the-counter non-deliverable forwards sample data between year 2004 – 2017. The sample data consist of five currencies which are examined against euro. The currencies are Brazilian real, Indian rupee, Indonesian rupiah, Korean won and Philippine peso. The second hypothesis is formed based on previous studies to examining influence of sample period affect to slope estimate of Fama regression model. Based on regression results in all sample currencies, the unbiased forward rate hypothesis is rejected and forward premium anomaly is evident in the non-deliverable forward market. Fama regression model slope coefficient estimate is determined to be biased estimator of future spot rate and have very little prediction power. Forward premium anomaly from corporate hedging perspective is discussed.