Volume 8, Issue 4

In this paper we propose a new class of tests for the martingale difference hypothesis based on the moment conditions derived by Bierens (1982). In contrast with the existing consistent tests, the proposed test has a standard limiting distribution and is easy to implement. Comparing with many commonly used autocorrelation- and spectrum-based tests, it has better power against a larger class of alternatives that may be serially correlated or uncorrelated. Moreover, this test does not rely on the assumption of conditional homoskedasticity and requires a weaker moment condition. Our simulations confirm that the proposed test is powerful against various linear and nonlinear alternatives and is quite robust to the failure of higher-order moments. Our empirical study on exchange rate returns also shows that the conclusion resulted from the proposed test is different from that of the conventional tests.

We introduce a flexible nonparametric technique that can be used to select weights in a forecast-combining regression. We perform a Monte Carlo study that evaluates the performance of the proposed technique along with other linear and nonlinear forecast-combining procedures. The simulation results show that when forecast errors are correlated across models, the nonparametric weighting scheme dominates. As a general rule, our simulation results suggest that the practice of combining forecasts, no matter the technique employed in selecting the combination weights, can yield lower forecast errors on average. An application to inflation forecasting is also presented to demonstrate the use of all forecast-combining techniques.

This paper analyses the nonlinear dynamic behaviour of intraday returns in the Eurostoxx50 cash and futures index which, given their relatively recent appearance, have not yet been analysed. We find that both return series show nonlinear individual dynamics that cannot exclusively be explained by the presence of conditional heteroskedasticity. Our findings also indicate nonlinear dynamic relationships between both market prices. The adjustment process to mispricing errors is nonlinear and shows periods of explosive behaviour. Finally, we distinguish between linear and nonlinear Granger causality and establish that the information flow is bi-directional both in the linear as well as in the nonlinear sphere.

In a heterogenous agents framework, we study a randomized version of Zeeman's market model with fundamental and momentum traders. Using methods from random dynamical systems theory, we examine convergence properties of invariant measures which correspond to market equilibria. It turns out that due to a stochastic stabilisation effect the market stays stable up to some critical value of speculative activity. If this threshold is surpassed, sudden trend reversals are possible without being induced by some exogenous shock.

Linear dynamic equilibrium correction mechanisms are shown to follow from the discretisation of continuous processes with steady-state solutions.

Many service industry firms strive hard to fill free capacity in order to cover their costs for a fixed capital stock. This paper presents a time series model where the capacity constraint is an integral part. The integer-valued autoregressive model builds on a simple idea of how daily time series arise for hotels and other similar establishments. Measures that follow naturally from the time series model are the occupancy probability and the duration of stay for the visitor. Empirically, we study the effects of price changes and a large festival, on these measures.