eThesis - School of Business electronic theses

OBJECTIVES OF THE STUDY The objective of this thesis is to analyze the profitability of a specific fixed income trading strategy in the U.S. Treasury market. The strategy in question, riding the yield curve, is based on the assumption that the strong form of the rational expectations hypothesis of the term structure of interest rates (REHTS) does not hold. In addition, the thesis assesses whether different filter rules can enhance the performance of the strategy. This part of the study is motivated by previous studies documenting violations of the REHTS, as well as the inconclusive findings on the profitability of riding the yield curve.

The second objective of the study is to scrutinize whether the returns from riding the yield curve explain fixed income hedge fund returns. The motivation for this part of the study stems from the findings of previous studies that fixed income hedge funds engage in risk arbitrage. Since hedge funds are often considered to be lacking in transparency, this thesis aims to extend previous literature by shedding light on the relatively little-known investment strategies of fixed income hedge funds.

DATA AND METHODOLOGY The study utilizes interest rate data consisting of the U.S. Treasury data spanning from June 1961 through June 2010. The hedge fund data consists of hedge fund index data as well as data relating to a particular risk factor model. The sample period for the hedge fund data spans from January 1994 to September 2009.

The ex post excess returns from riding the yield curve are calculated for trading strategies of various holding periods and maturities of the riding instrument, and the attractiveness of these strategies is assessed on a risk-adjusted basis by using the Sharpe ratio. Moreover, performance of the filter rules is analyzed in the context of riding the yield curve. Lastly, hedge fund index returns are regressed on a risk factor model amended with the returns from riding the yield curve.

RESULTS The main findings of the thesis are summarized in the following: firstly, it is found that riding the yield curve strategies have generally outperformed the buy-and-hold strategies over equal maturities; in other words, REHTS is clearly violated. The performance of the riding strategies has changed markedly over decades, and it is strongly linked to the prevailing interest rate environment. Secondly, it is found that the filter rules do not enhance the performance of the riding strategies. Thirdly, no positive relationship between riding returns and hedge fund returns is found. In conclusion, this study finds no evidence that the yield curve speculators would consistently ride the yield curve, although the ex post excess riding returns have on average been positive in the period under scrutiny.