[This paper develops and tests arbitrage bounds for a combination of two option spread positions known as a box spread. This strategy involves the simultaneous use of four options and creates a position that is equivalent to riskless lending. The no-arbitrage conditions are compared to existing arbitrage bounds and are tested using Chicago Board Options Exchange data.]

This paper develops and tests arbitrage bounds for a combination of two option spread positions known as a box spread. This strategy involves the simultaneous use of four options and creates a position that is equivalent to riskless lending. The no-arbitrage conditions are compared to existing arbitrage bounds and are tested using Chicago Board Options Exchange data.

The purpose of this paper is to provide a test of a state‐dependent multinomial model of intertemporal changes in the term structure of interest rates. The theoretical background for the model comes from Ho and Lee (1986) . The current paper extends their model in several significant ways. First, we perform diagnostic tests on the data to demonstrate that the empirical results reject a binomial model in favor of a trinomial one. After theoretically deriving the appropriate trinomial model, the current paper extends their model to allow for state‐dependent shifts which are determined by the set of ex ante observable state variables. The methodology for the study utilizes OLS regressions to identify the exogenous explanatory variables which drive the hypothesized trinomial process of term structure evolution. The empirical tests indicate that the set of state variables explains a significant portion of the variability in the shifts of the term structure over time. The model also identifies and quantifies a set of variables which impact on changes in the term structure of interest rates.