摘要：This paper investigates the nonlinearities in the behavior of jet fuel prices and air carrier yields as measured by revenue passenger miles(RPMs), where one RPM is defined as one passenger flown one mile in revenue traffic. It indicates that previous research might have overlooked the possibilities of nonlinear dynamics between these two series. Drawing on existing tests of nonlinearities and chaos, this paper first investigates the existence of chaotic behavior as the source of nonlinearities in the monthly prices of jet fuel and RPMs. The findings show strong evidence that the two series exhibit nonlinear dependencies. Evidence is found, however, that this behavior may be inconsistent with chaotic structure. We propose and estimate bivariate GARCH(1,1) and bivariate EGARCH(1,1) models to ascertain the flow of information between jet fuel prices and revenue passenger miles. Estimation results of the bivariate GARCH models offer evidence that the shock transmission between the two series is mainly asymmetric, that is that positive and negative shocks impart degree of volatility differently. It is shown that the positive shocks to jet fuel prices show a substantially higher reaction from the revenue passenger miles. The conclusion is that, RPMs are quite responsive to upward volatility in prices of jet fuel, while falling jet fuel prices may not translate into efficiency gains
