A risk-based approach to evaluate currency risk premiums

Abstract

This work focuses on the theory of uncovered interest rate parity and particularly on whether findings can be disrupted by differences in bonds’ risk level. Previous literature has rejected the theory indicating that higher interest rate currencies appreciate to lower interest rate currencies. In my research I have used a proven and very effective system on evaluating the link between exchange rates, interest rates and risk premium. This method was developed by Lusting and Verdelhan (2007): I have built five portfolios of foreign currency excess returns on the basis of the foreign interest rates: the first portfolio always contains the lowest interest rate currencies and the last portfolio contains the highest interest rate currencies. The method used to pursue my results is not the key innovation in my work. The peculiarity of my research is that I introduced the risk variable. All the researches carried out in the past considered currencies and countries from all over the world. However, what I strongly believe is that we cannot compare countries whose bonds are perceived as more (or less) risky by investors. In the work of Lusting and Verdelhan (2007), as well in many others, the research is carried out comparing US bonds to countries including Bangladesh, Egypt, Kenya, Namibia, Nigeria. Of course the uncovered interest rate parity condition is supposed to work only with same-risky bonds, hence, we cannot expect to prove empirically the theory by comparing US bonds to Egypt ones. In my research I considered the Euro-Dollar exchange rate over the past decade and I compared US bonds to six Euro-Zone bonds (called A group) from countries which are as same rated as USA: Austria, France, Germany, Netherlands, Belgium. To prove my theory, and to demonstrate how the risk premium can be twisted when comparing countries which are not comparable, I created a second group of Euro-Zone countries (called B group). This group is composed by Italy, Spain, Portugal, Ireland and Greece. As both of the groups have the same currency, the Euro, I insulated the currency effect and this allowed me to concentrate on the “rating” effect. The results empirically estimates an inverse relationship between interest rate and exchange rate for the A group, thus supporting the theory of uncovered interest rate parity. This finding are further confirmed by irregular results of the B group, which partially trace what outlined in other past works, and hence prove the misleading effect resulting from a research which does not consider the differences in bonds’ risk level.