This thesis looks at investments in low-risk assets denominated in foreign currencies and discusses theoretical implications related to the findings. Specifically, it looks at Japanese yen (JPY) and US dollars (USD).
Chapter 1 provides some background information about the phenomenon called carry trade, which has been mentioned in the public debate about economic affairs in general and exchange rate movements in particular. Since the term carry trade is often used in a vague or undefined manner, I start out by gathering some information from various sources in order to present a short overview of what carry trade actually is. This is done in the first part of Chapter 1. I then continue by describing some aspects of the Japanese financial market in order to justify the choice of Japanese yen and US dollars for the empirical study made in the thesis.
In the main part of Chapter 1, I go on to check whether uncovered interest parity (relating interest rates and expected exchange rate changes) are supported using the data on Japanese yen and US dollars. I then proceed to explore if there was a forward discount bias in the period 1993 to 2008 in order to look at uncovered interest parity from another angle. This is done by removing currency exchange risk through the use of forward contracts.
My empirical findings do not support uncovered interest parity. In fact, the results are opposite of what the uncovered interest parity hypothesis suggests. In the case of uncovered interest parity, I find that when a currency gave a higher interest return than another, this tended to lead to an increase rather than a decrease in the value of the same currency over the same period. I also find that although shorter three-year periods within the 1993-2008 period exhibited mixed results, uncovered investments in USD three-month bonds tended to yield higher return over the longer 1993-2008 period when compared to similar uncovered three-month bond investments in JPY.
I find that covered investments in Japanese yen gave higher return than similar investments in US dollars. This result is robust in the sense that it applied to the observed period 1993-2008 as a whole, and also to shorter three-year time-spans within this period. This shows that the forward rate was distorted, or biased. I relate this to the forward discount bias which has been described in economic literature.
In the last part of Chapter 1, I discuss potential reasons for these findings and compare the findings to other studies by reviewing some of the literature in this field. I also look at the development of the Japanese financial market from the 1980s. That review indicates that although the Japanese financial market may have been less liquid in the past than the US market, deregulations and reforms have made the markets more similar.
Chapter 2 of the thesis goes a step further into a theoretical paradox called the Siegel paradox. The Siegel paradox describes investments in foreign currencies and predicts a forward discount bias. I study the Siegel paradox for currency movements by using a simplified example with USD and JPY. The rationale leading to the paradox is analyzed, and alternative ways of thinking are investigated by looking at a simple version of the two-envelope problem. I try to find similarities and differences between the Siegel paradox and the two-envelope problem in order to gain some insight into what may be the common cause of the two paradoxes. I find that Jensen’s inequality applies to both paradoxes. I find that the lack of a total value in the Siegel paradox and the existence of such a total value in the two-envelope problem may be one of the important differences between the two problems.
In Chapter 3, I discuss the relevance of the findings related to the paradoxes in Chapter 2 to interest rate parity and the forward discount bias found in Chapter 1. I suggest that the Siegel paradox is a purely theoretical misstatement which does not explain the forward discount bias observed in the market. Rather, I suggest that the forward discount bias is the result of other rational choices made by investors, like a risk premium and risk aversion among the investors combined with tax differences, trading costs and differences in liquidity. I suggest that investors that base their investment choices on the Siegel paradox are probably allocating their capital in an inefficient manner.
Finally, I indicate that further investigation should be made into resolving the Siegel paradox as this could possibly lead to a deeper understanding of risk, return and capital allocation in the financial markets.