Forecasting structural time series models and the kalman filter: Andrew Harvey, 1989, (Cambridge University Press), 554 pp., ISBN 0-521-32196-4

By Robert Fildes. 1992

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Recurrent Neural Filters: Learning Independent Bayesian Filtering Steps for Time Series Prediction (2019, methodology): “On top of efficient inference and uncertainty estimation, this decomposition of inference steps enables Bayes filters to be deployed in use cases beyond basic one-step-ahead prediction – with simple extensions for multistep prediction [17] and prediction in the presence of missing observations [16].”
The Tourism Forecasting Competition (2011, methodology): “There are a large number of potential ARIMA models—too many to allow us to estimate every possible combination of p, q, P and Q. Instead, we need a way to efficiently traverse the space of models in order to arrive at the model with the lowest AIC value. Hyndman and Khandakar (2008) proposed the fol”
Smoothing and trend detection in waterbird monitoring data using structural time-series analysis and the Kalman filter (2007, methodology): “For mathematical details about structural time-series analysis and the Kalman filter please refer to Harvey (1989) .”
Affine Term-Structure Models: Theory and Implementation (2001, methodology): “The sources used in this paper for the Kalman filter algorithm include Kim and Nelson (1999), Hamilton (1994), Harvey (1990), and Judge, Griffiths, Hill, Lütkepohl, and Lee (1985, Appendix C).”
A Survey of Econometric Methods for Mixed-Frequency Data (2013, methodology): “Φf,p 01×(5−p) I4 04×1 ] ; F2 = 05×10; (40)”
A general test for time dependence in parameters (2004, methodology): “A third type of time dependence in parameters allows βit to vary stochastically:1 1 2 1it i t tβ α α β ε−= + + (3) Modeling parameter variation of this sort is a signal-extraction problem that is amenable to the Kalman filter (Harvey, 1989).”
The stochastic volatility in mean model: empirical evidence from international stock markets (2002, methodology): “To obtain a maximum likelihood estimate for , which we denote by ̂, the loglikelihood is numerically maximised with respect to in a similar fashion as for Gaussian models; see Harvey (1989) and Koopman et.al (1995).”
What drives Credit Risk in Emerging Markets? The Roles of Country Fundamentals and Market Co-movements (2006, methodology): “…shows the diagnostic tests for each country; these are carried-out using the standardised one-step-ahead residuals defined as follows, according to Harvey (1989): ttt f/vv ~ = , where tv = Bt - 1ttB − and )vvar(f tt = (13) Though there is a lack of normality for all of the countries’ standardised…”
25 Years of IIF Time Series Forecasting: A Selective Review (2005, methodology): “The works on state space, Kalman filtering, and discrete/continuous time structural models by Harvey (1989), West & Harrison (1989, 1997) and Durbin & Koopman (2001) have had an impact on the time series literature.”
Convergence of European spot market prices for natural gas? A real-time analysis of market integration using the Kalman Filter (2006, background): “3 For a formal description see Harvey (1989) and Meinhold and Singpurwalla (1983). less than 6 euros per MWh; on the contrary, the minimum for the UK is 1.86 Euros/MWh.”
A meta-analysis of international tourism demand forecasting and implications for practice (2014, methodology): “8 variables (Harvey, 1990; Gonzalez & Moral, 1995). However, according to Turner and Witt (2001) and Kulendran and Witt (2003b), no evidence has yet emerged to suggest that the inclusion of explanatory variables improves the forecasting accuracy of the BSM model.”
Forecasting international quarterly tourist flows using error-correction and time-series models (1997, methodology): “The types of models fitted include seasonal autoregressive integrated moving average (ARIMA) models, Harvey's (1990) basic structural models (BSM), simple autoregressive (AR) models and regression based time series models.”
Equivalence and Differences Between Structural Equation Modeling and State-Space Modeling Techniques (2010, background): “Indeed, different options for handling unknown initial condition in statespace models have been proposed by various researchers over the years (e.g., Ausley & Kolm, 1985; De Jong, 1991; Harvey, 2001).”
Trend Filtering Methods for Momentum Strategies (2011, methodology): “Harvey (1989) shows that we can obtain a t+1|t directly from a t|t−1: a t+1|t = (Tt+1 −KtZt) a t|t−1 +Ktyt + (ct+1 −Ktdt) where Kt = Tt+1P t|t−1Z⊤t F −1 t is the matrix of gain.”