Modelling Time-Varying Parameters in Panel Data State-Space Frameworks: An Application to the Feldstein–Horioka Puzzle

Abstract

In this paper, we develop a State-Space framework for modeling panel time series data. Our research extends the simple canonical model generally employed in the literature, into a panel-data time-varying parameters framework, combining both fixed (either common and country-specific) and varying components. Under some specific circumstances, this setting can be understood as a mean-reverting panel time-series model, where the mean fixed parameter can, at the same time, include a deterministic trend. Regarding the transition equation, our structure allows for the estimation of different autoregressive alternatives, and include control instruments, whose coefficients can be setup either common or idiosyncratic. This is particularly useful to detect asymmetries among individuals (countries) to common shocks. We develop a GAUSS code that allows for the introduction of restrictions regarding the variances of both the transition and measurement equations. Finally, we use this empirical framework to test for the Feldstein-Horioka puzzle in a 17-country panel. The results show its usefulness to solving complexities in macroeconomic empirical research.

Capital mobility in Latin American and Caribbean countries: new evidence from dynamic common correlated effects panel data modeling

“…,Payne and Kumazawa (2005),Murthy and Anoruo (2010),Bangake and Eggoh (2012),Hernandez (2015), andKetenci (2015). Further, there are several studies that have investigated quantitatively the impact of financial flows on the FH puzzle by modeling algorithms focusing on financial risk analysis; for example,Bai and Zhang (2010),Kou et al (2014),Camarero et al (2019),Causevic (2020), andWang et al (2020).…”

50 Years of Capital Mobility in the Eurozone: Breaking the Feldstein-Horioka Puzzle

“…In this line, Ketenci (2018) using GMM techniques analyzes the impact of the 2008 financial crisis on the level of capital mobility, finding an insignificant impact. Finally, Camarero et al (2020) developed a time-varying parameters statespace model applied to panel data covering 12 EU countries, where they show the evolution along time of the FH coefficient as a by-product of a more general OECD sample. 8 Finally, the ECB performs periodical studies on the degree of financial integration in the Eurozone.…”

“…8 At this point, it is important to underline a caveat. Note that But and Morley (2017) and Camarero et al (2020) cover a sample of OECD countries following the tradition of the seminal literature on FH. Therefore, neither the econometric approach nor the sample are entirely comparable to the present paper.…”

“…Our contributions are the following: first, we reassess the FH puzzle for the Eurozone countries for the most extended sample available. Surprisingly, there is little empirical literature on the Feldstein-Horioka puzzle covering financial integration from the very beginning of the monetary integration process, the only exceptions to the best of our knowledge being Katsimi and Zoega (2016), Drakos et al (2018), ECB (2018ECB ( , 2020 and Camarero et al (2020), although they use different empirical approaches and cover a shorter time span compared to the present study; second, for this purpose, we take into account the non-stationarity properties of the variables in the same vein of recent literature both in large panels and time series; third, we relate our econometric methodology, that accounts for structural changes (breaks) in the long-run relationship, with the changes in the degree of financial integration and capital mobility as measured by the savings-retention coefficient in the FH equation; fourth, we deal with the expected degree of heterogeneity within the EU considering the existence of several country groups. The groupings are determined endogenously relying on the breaks found and signal a clear difference between core countries and the rest of the sample; fifth, we present robustness checks using in a complementary way both MG/PMG estimators 4 and rolling window regressions.…”

“… See, among others, Telatar et al (2007),Gomes et al (2008),Ma and Li (2016),Khan (2017) orCamarero et al (2020).…”