Forecasting economic time series using flexible versus fixed specification and linear versus nonlinear econometric models

Nine macroeconomic variables are forecast in a real-time scenario using a variety of flexible specification, fixed specification, linear, and nonlinear econometric models. All models are allowed to evolve through time, and our analysis focuses on model selection and performance. In the context of real-time forecasts, flexible specification models (including linear autoregressive models with exogenous variables and nonlinear artificial neural networks) appear to offer a useful and viable alternative to less flexible fixed specification linear models for a subset of the economic variables which we examine, particularly at forecast horizons greater than 1-step ahead. We speculate that one reason for this result is that the economy is evolving (rather slowly) over time. This feature cannot easily be captured by fixed specification linear models, however, and manifests itself in the form of evolving coefficient estimates. We also provide additional evidence supporting the claim that models which ‘win’ based on one model selection criterion (say a squared error measure) do not necessarily win when an alternative selection criterion is used (say a confusion rate measure), thus highlighting the importance of the particular cost function which is used by forecasters and ‘end-users’ to evaluate their models. A wide variety of different model selection criteria and statistical tests are used to illustrate our findings.     

Threshold models for trended time series

This paper presents the theoretical development of a new threshold autoregressive model based on trended time series. The theoretical arguments underlying the model are outlined and a nonlinear economic model is used to derive the specification of the empirical econometric model. Estimation and testing issues are considered and analysed. Additionally we apply the model to the empirical investigation of U.S. GDP.This paper is the result of work carried out for the author's Ph.D. thesis. I would like to thank Hashem Pesaran for his help, encouragement and insights during the preparation of this paper. I also thank Gary Koop and Sean Holly for helpful comments. Financial assistance from the Economic and Social Research Council is gratefully acknowledged.     

The impact of flood dynamics on property values

There is evidence that environmental amenities and dis-amenities may be nonlinearly related to property valuation. This would bring inconsistency in estimating environmental variables of the hedonic price model. To explore the linearity of the relationship this study analyses spatial and temporal variation impacts of the 2011 Brisbane flood on property markets using semi-parametric estimation. The results show that most environmental variables impacts on property values nonlinearly, and in particularly distance to the river, indicating that the amenity value of being close to a river outweighs the flood risks. The estimation of the combined impact of elapsed time and neighbourhood income indicates that the flood risk impact on property markets disappears over time.     

Non-linear dynamics in international resource markets: Evidence from regime switching approach

This paper examines the nonstationary and nonlinear features of the non-renewable resource markets: the crude oil (US West Texas Intermediate and UK Brent), bituminous coal and natural gas markets. In particular, we achieve this goal by using the Markov switching unit root regression. This approach is attractive because it allows price to switch between stationary and nonstationary regimes (partial nonstationarity). It also allows price to switch between two stationary regimes (varied stationarity) or to switch between two nonstationary regimes (varied nonstationarity). The results of a range of non-linear tests show that the independently and identically distributed (i.i.d.) hypothesis or the random walk hypothesis is untenable for the non-renewable resource prices. The results from the Markov regression indicate that, in the cases of US West Texas Intermediate, UK Brent as well as bituminous coal, prices are characterized by the local nonstationarity in both regimes, and therefore varied nonstationarity is sustained. The price of natural gas is characterized by partial nonstationarity, indicating that this market is inconsistent with the efficient market.     

Comparing linear and nonlinear forecasts for stock returns

We compare the out-of-sample performance of monthly returns forecasts for two indices, namely the Dow Jones (DJ) and the Financial Times (FT) indices. A linear and a nonlinear artificial neural network (ANN) model are used to generate the out-of-sample competing forecasts for monthly returns. Stationary transformations of dividends and trading volume are considered as fundamental explanatory variables in the linear model and the input variables in the ANN model. The comparison of out-of-sample forecasts is done on the basis of forecast accuracy, using the Diebold and Mariano test [J. Bus. Econ. Stat. 13 (1995) 253.], and forecast encompassing, using the Clements and Hendry approach [J. Forecast. 5 (1998) 559.]. The results suggest that the out-of-sample ANN forecasts are significantly more accurate than linear forecasts of both indices. Furthermore, the ANN forecasts can explain the forecast errors of the linear model for both indices, while the linear model cannot explain the forecast errors of the ANN in either of the two indices. Overall, the results indicate that the inclusion of nonlinear terms in the relation between stock returns and fundamentals is important in out-of-sample forecasting. This conclusion is consistent with the view that the relation between stock returns and fundamentals is nonlinear.     

Nonlinearity and time-variation in the monetary model of exchange rates

In this paper we analyze the existence of nonlinear relationships between macroeconomic fundamentals and exchange rates for some major industrialized countries using an error correction model with time-varying parameters for the post Bretton Woods period. We find that inflation rate differentials with respect to the US inflation rate are the driving forces for the nonlinear relationships in the monetary model for exchange rates for the data from Germany, the UK, Canada, France and Italy. In addition to the variables in the traditional monetary model, also the relative interest rates are relevant in determining exchange rate changes only when the inflation differentials are either very large or very small. In contrast to previous studies we find significant long-run effects in the error correction representation of the monetary model for exchange rates when the nonlinear dynamics is taken into account in the analysis.     

Adaptive market efficiency of agricultural commodity futures contracts

In this paper we investigate the adaptive market efficiency of the agricultural commodity futures market, using a sample of eight futures contracts. Using a battery of nonlinear tests, we uncover the nonlinear serial dependence in the returns series. We run the Hinich portmanteau bicorrelation test to uncover the moments in which the nonlinear serial dependence, and therefore adaptive market efficiency, occurs for our sample.     

Simultaneous Adjustment of Quantities and Prices: An Example of Hamiltonian Dynamics

In a well-known essay first published in 1953, Goodwin analyzed the dynamic adjustment of quantities and prices to long-run equilibrium values, in a set ofn ‘Walrasian’ markets. He treated the crossed adjustment of prices and quantities as a linear Hamiltonian vector field. In more recent work, Goodwin introduced non-linear perturbations in his multi-sectoral adjustment model. He assumed that real consumption depends non-linearly on relative prices. This paper shows the following: (1) Goodwin's behavioural hypotheses are compatible with the assumption that agents maximize; (2) if the dynamic process is Hamiltonian, then symplectic coordinate changes are essential tools of analysis; (3) if the real wage is rigid and returns to scale are not constant, then the Hamiltonian model can generate chaotic transients or, in extreme cases, pure chaotic motions.     

Exchange rate volatility and regime change: A Visegrad comparison

We analyze exchange rate volatility in the Visegrad Four countries during the period in which they abandoned tight regimes for more flexible ones. We account for path dependency, asymmetric shocks, and movements in interest rates. In addition, we allow for a generalized error distribution. The overall findings are that path-dependent volatility has a limited effect on exchange rate developments and that the introduction of floating regimes tends to increase exchange rate volatility. During the period of flexible regimes, volatility was mainly driven by surprises. Asymmetric effects of news tend to decrease volatility under the floating regime. Interest differentials impact exchange rate volatility contemporaneously under either regime, although we find no intertemporal effect of interest differentials. Journal of Comparative Economics 34 (4) (2006) 727–753.