The failure of the cost of carry model to explain the price of the Australian Share Price Index Futures contract has been an enigma for some time. This first part of this study develops a theory which explains why this is so and adopts the perspective that it is the basis which is priced. The Transaction Cost Limits pricing framework argues that arbitrage is mostly important at the edges of the no arbitrage band, but it has little implication for prices within the band. (However it could affect the variance of the basis). The no arbitrage band is examined in some detail and it is postulated that the band widens with time to maturity, is mis-specified in efficiency studies, and because the boundaries are "fuzzy" it is impossible to precisely define.
As these processes appear to be mostly insignificant to the futures price, other processes were examined. Due to the large width of the Australian no arbitrage band, the spot price level (basis = 0), remains within the no arbitrage band for some time. As the spot price is an unbiased estimator of future prices, it was postulated the two prices shouldn't significantly differ in the long term, (at least for the duration of the near contract). Various non-stationary, autoregressive expectations factors were discussed but the basic conclusion remained unchanged. It was also postulated that volatility of the basis should decrease with decreasing time to maturity. When coupled with inter contract arbitrage, interest rate factors are not expected to be significant to the basis in the near contract, but they are significant in the far contracts, consequently the near and next to near contracts should have significantly different properties. Empirical testing confirmed the basic properties expected, supporting the general validity of the models and processes proposed. Of particular importance to hedgers is the strong time dependency in the basis series which has been postulated and described, and the tendency of the basis to return to zero well before maturity.