Abstract
An important issue in modelling economic time series is whether key unobserved components representing trends, seasonality and calendar components, are deterministic or evolutive. We address it by applying a recently proposed Bayesian variable selection methodology to an encompassing linear mixed model that features, along with deterministic effects, additional random explanatory variables that account for the evolution of the underlying level, slope, seasonality and trading days. Variable selection is performed by estimating the posterior model probabilities using a suitable Gibbs sampling scheme. The paper conducts an extensive empirical application on a large and representative set of monthly time series concerning industrial production and retail turnover. We find strong support for the presence of stochastic trends in the series, either in the form of a time-varying level, or, less frequently, of a stochastic slope, or both. Seasonality is a more stable component, although in at least 60 % of the cases we were able to select one or more stochastic trigonometric cycles. Most frequently the time variation is found in correspondence with the fundamental and the first harmonic cycles. An interesting and intuitively plausible finding is that the probability of estimating time-varying components increases with the sample size available. However, even for very large sample sizes we were unable to find stochastically varying calendar effects.
Similar content being viewed by others
Notes
The results are based on 80,000 draws, with burn-in sample of 20,000.
References
Bell WR, Hillmer SC (1983) Modeling time series with calendar variation. J Bus Econ Stat 78:526–534
Bell WR, Martin DEK (2004) Modeling time-varying trading day effects in monthly time series. In ASA Proceedings of the joint statistical meetings, American Statistical Association, Alexandria
Busetti F, Harvey AC (2003) Seasonality tests. J Bus Econ Stat 21:420–436
Canova F, Hansen BE (1995) Are seasonal patterns constant over time? A test for seasonal stability. J Bus Econ Stat 13:237–252
Casella G, Girón FJ, Martínez ML, Moreno E (2009) Consistency of bayesian procedures for variable selection. Annal Stat 37:1207–1228
Dagum EB, Quenneville B (1993) Dynamic linear models for time series components. J Econ 55:333–351
Dagum EB, Quenneville B, Sutradhar B (1993) Trading-day variations multiple regression models with random parameters. Int Stat Rev 60:57–73
De Jong DN, Whiteman CH (1991) The case for trend-stationarity is stronger than we thought. J Appl Econ 6:413–421
Dickey DA, Fuller WA (1979) Distribution of the estimators for autoregressive time series with a unit root. J Am Stat Assoc 74:427–431
Durbin J, Koopman SJ (2002) A simple and efficient simulation smoother for state space time series analysis. Biometrika 89:603–615
Durbin J, Koopman SJ (2012) Time series analysis by state space methods, 2nd edn. Oxford University Press, Oxford
Frühwirth-Schnatter S, Wagner H (2010) Stochastic model specification search for Gaussian and partial non-Gaussian state space models. J Econ 154:85–100
George EI, McCulloch R (1993) Variable selection via Gibbs sampling. J Am Stat Assoc 88:881–889
Geweke J (1992) Evaluating the accuracy of sampling-based approaches to the calculation of posterior moments. In: Heckman JJ, Leamer E, Bernado JM, Berger JO, Dawid AP, Smith AFM (eds) Bayesian statistics 4. University Press, Oxford, pp 169–193
Geweke J (2005) Contemporary bayesian econometrics and statistics. Wiley Series in Probability and Statistics, New York
Ghysels E, Osborn D (2001) The econometric analysis of seasonal time series. Cambridge University Press, Cambridge
Gower JC, Hand DJ (1996) Biplots. monographs on statistics and applied probability. Chapman & Hall, London
Grassi S, Proietti T (2014) Characterising economic trends by bayesian stochastic model specification search. Comput Stat Data Anal 71:359–374
Greenacre M (2010) Biplots in practice. Fundación BBVA, Mexico
Harvey AC (1989) Forecasting, structural time series and the Kalman filter. Cambridge University Press, Cambridge
Hylleberg S (1992) Modelling seasonality. Oxford University Press, Oxford
Hylleberg S, Engle RF, Granger CWJ, Yoo BS (1990) Seasonal integration and cointegration. J Econom 44:215–238
Hylleberg S, Pagan A (1997) Seasonal integration and the evolving seasonals model. Int J Forecast 13:329–340
Kim S, Shephard N, Chib S (1998) Stochastic volatility: likelihood inference and comparison with ARCH models. Rev Econ Stud 65:361–393
Koop G (1992) Objective bayesian unit root tests. J Appl Econom 7:65–82
Koop G, van Dijk HK (2000) Testing for integration using evolving trend and seasonals models: a Bayesian approach. J Econom 97:261–291
Kwiatkowski D, Phillips PCB, Schmidt P, Shin Y (1992) Testing the null hypothesis of stationarity against the alternative of a unit root. J Econom 54:159–178
La Grange A, Roux N, Gardner-Lubbe S (2009) Biplotgui: interactive biplots in r. J Stat Softw 30:1–37
Ley E, Steel MFJ (2009) On the effect of prior assumptions in bayesian model averaging with applications to growth reression. J Appl Econom 4:651–674
Nerlove M, Grether DM, Carvalho JL (1979) Analysis of economic time series: a synthesis. Academic Press, New York
Nyblom J, Makelainen T (1983) Comparisons of tests for the presence of random walk coefficients in a simple linear model. J Am Stat Assoc 78:864–865
O’Hagan A (1995) Fractional bayesian factors for model comparison. J R Stat Soc, Ser B. 57:99–118
Peña D, Tiao GC, Tsay RS (2001) A course in time series anal. Wiley, New York
Phillips PCB (1991) To criticize the critics: an objective bayesian analysis of stochastic trends. J Appl Econom 6:333–364
Phillips PCB, Perron P (1988) Posterior odds testing for a unit root with data-based model selection. Biometrika 75:335–346
Phillips PCB, Perron P (1994) Testing for a unit root in time series regression. Econom Theory 10:774–808
Proietti T, Grassi S (2012) Bayesian stochastic model specification search for seasonal and calendar effects. In: Bell William R, Holan Scott H, McElroy Tucker S (eds) Economic time series: modeling and seasonality. CRC Press, Boca Raton, pp 431–455
Schotman PC, van Dijk HK (1991) On bayesian routes to unit roots. J Appl Econom 6:387–401
Sims CA (1988) Bayesian skepticism on unit root econometrics. J Econ Dyn Control 12:463–474
Sims CA, Uhlig H (1991) Understanding unit rooters: a helicopter tour. Econometrica 59:1591–1599
Strickland CM, Martin GM, Forbes CS (2007) Parameterisation and efficient MCMC estimation of non-Gaussian state space models. Comput Stat Data Anal 97:2911–2930
Wecker WE, Ansley CF (1983) The signal extraction approach to nonlinear regression and spline smoothing. J Am Stat Assoc 78:81–89
West M, Harrison J (1997) Bayesian Forecast Dyn Models. Springer, New York
Zellner A (1978) Seasonal analysis of economic time series. Department of Commerce-Bureau of the Census, New York
Zellner A (1983) Applied time series analysis of economic data. Department of Commerce-Bureau of the Census, New York
Acknowledgments
The authors wish to thank the reviewers and the editor for their careful review and constructive criticism that led to several improvements in the paper. Tommaso Proietti gratefully acknowledges financial support by the Italian Ministry of Education, University and Research (MIUR), PRIN research project 2010–2011 - prot. 2010J3LZEN, Forecasting economic and financial time series: understanding the complexity and modelling structural change. Both authors acknowledge support from CREATES - Center for Research in Econometric Analysis of Time Series (DNRF78), funded by the Danish National Research Foundation.
Author information
Authors and Affiliations
Corresponding author
Appendix A: markov chain monte carlo estimation
Appendix A: markov chain monte carlo estimation
This appendix provides details of the Gibbs sampling scheme outlined at the end of Sect. 3. Recall that the linear mixed model was defined as \(y = Z_\varUpsilon \psi _\varUpsilon + \epsilon \), where \(y\) and \(\epsilon \) are vectors stacking the values \(\{y_t\}\) and \(\{\epsilon _t\}\), respectively, and the generic row of matrix \(Z_\varUpsilon \) contains the relevant subset of the explanatory variables.
Step a. is carried out by sampling the indicators with probabilities proportional to the conditional likelihood of the regression model, as
which is available in closed form (see below). All the \(2^{10}\) combinations of indicators are sampled jointly in a multimove way as in FS-W, conditional on the latent process \(\alpha \).
Under the normal-inverse Gamma conjugate prior for \((\psi _\varUpsilon , \sigma _\epsilon ^2)\)
where \(D_{\varUpsilon }\) is a diagonal matrix with elements \(\kappa _\mu , \kappa _A,\) etc., steps b. and c. are carried out by sampling from the posteriors
where
Finally,
see e.g. Geweke (2005), where \(\Gamma (\cdot )\) denotes the Gamma function.
The sample from the posterior distribution of the latent states, conditional on the model and its parameters, in step d., is obtained by the conditional simulation smoother proposed by Durbin and Koopman (2002).
The draws of the parameters \(\beta _\mu \), \(\beta _A\), \(\beta _{sj}, j = 1, \ldots , 6\), \(\beta _{TD}\) are obtained by performing a final random sign permutation. This is achieved by drawing independently Bernoulli random variables \(\mathsf{B}_\mu \), \(\mathsf{B}_A\), \(\mathsf{B}_{sj}, j = 1, \ldots , 6, \mathsf{B}_{TD}\) with probability 0.5, and recording \((-1)^{\mathsf{B}_{\mu }} (\sigma _\eta , \tilde{\mu }_{t})\), \((-1)^{\mathsf{B}_{A}} (\sigma _\zeta , \tilde{A}_{t}, a_t)\), etc.
The starting values are obtained by iterating for the full model (with all the indicators being equal to 1) the above GS scheme 1000 times, with initial values of the hyperparameters in \(\psi _\varUpsilon \) set equal to zero.
Under a fractional prior the marginal likelihood is evaluated as follows:
where \(q\) is the dimension of \(\psi _{\varUpsilon }\), and \(m\) is as given above. The free parameter in the fractional prior specification is \(b\) that in our study is set to be \(10^{-4}\) and \(10^{-5}\).
A key assumption is that \(\sigma ^2_\epsilon \) is strictly greater than zero, i.e. the irregular component is always present.
Rights and permissions
About this article
Cite this article
Proietti, T., Grassi, S. Stochastic trends and seasonality in economic time series: new evidence from Bayesian stochastic model specification search. Empir Econ 48, 983–1011 (2015). https://doi.org/10.1007/s00181-014-0821-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-014-0821-y