Statistical Yield Curve Arbitrage in Eurodollar Futures Using Neural Networks

摘要

This paper describes a study of statistical yield curve arbitirage amongest eurodollar futures. A factor analysis indicates that the bulk of changes in eurodollar futures are accounted for by unpredictable parallel shifts in the yield curve which closely follow a random walk. However, the second and third factors correspond roughly to a tilt and a flex in the yield curve and these show evidence of a degree of predicability in the form of mean-reversion. We construct portfolios of eurodolalr futures which are immunised against the first two fators but exposed to the third and use both linear and non-linear techniques to model the expected return on these portfolios. The third factor is found to te predictable by non-linear, but ont bylinear, technique.