Multivariate Unobserved Components Research Articles

Searching for Permanent Changes in the U.S. Economy: A Particle Gibbs Sampling Approach with Particle Rejuvenation

The main purpose of this paper is to scrutinize the time-varying relative importance of the permanent and the transitory shocks to the U.S. real GDP. After accounting for the cointegration relation between real GDP, consumption, and investment, we find that i) the real permanent shock played a more prominent role than the nominal transitory shock from the 1950s to the 1960s, but the two shocks have almost equally contributed to the stochastic movements of real GDP since the 1970s; (ii) the long-run growth of the U.S. economy has substantially slowed since the recession in 2001. The annualized growth rate of real GDP has declined to approximately 1.6 %, falling from a peak of nearly 3.3 % in the last decade. For the empirical analysis, we employ a multivariate unobserved component model that accommodates stochastic volatility. The multivativariate model is estimated by an efficient particle Gibbs sampler with particle rejuvenation that simultaneously draws latent state variables all at once.

Model-Based Business Cycle and Financial Cycle Decomposition for Europe and the U.S.

We develop a multivariate unobserved components model to extract business cycle and financial cycle indicators from a panel of economic and financial time series of four large developed economies. Our model is flexible and allows for the inclusion of cycle components in different selections of economic variables with different scales and with possible phase shifts. We find clear evidence of the presence of a financial cycle with a length that is approximately twice the length of a regular business cycle. Moreover, cyclical movements in credit related variables largely depend on the financial cycle, and only marginally on the business cycle. Property prices appear to have their own idiosyncratic dynamics and do not substantially load on business or financial cycle components. Systemic surveillance policies should therefore account for the different dynamic components in typical macro financial variables.

Decomposing Beveridge Curve Dynamics By Correlated Unobserved Components

AbstractOur paper provides a consistent framework to study the structural or cyclical nature of Beveridge curve (BC) dynamics: We connect equilibrium unemployment theory to a flexible multivariate unobserved components model. We disentangle permanent and transitory components of all series determining the BC and its position. Cointegration and identification are addressed. The German curve is an ideal illustration as reforms of the institutional setting and the Great Recession were accompanied by a remarkable labour market development. We find an extraordinary increase in trend matching efficiency after the reforms, which testifies to a permanent improvement. Matching efficiency accounts for about half of the BCś inward shift. As tightness also increased, a persistent upward movement masked the inward shift.

On the identification of multivariate correlated unobserved components models

This letter analyses identification for multivariate unobserved components models with correlated trend and cycle innovations. We address both the order as well as the rank criteria. Identification is shown for lag structures with lengths larger than one. We also discuss UC models with common features and with cycles that allow for dynamic spillovers.

Modely nepozorovaných komponent pro odhad produkční mezery české ekonomiky

Output gap belongs between imporant indicator of central banks by their executing monetary policy despite output gap is an unobservable variable. Because of importance of accurate measurement of output gap, there are more methods for estimation. The aim of this paper is to discover if univariate and multivariate unobserved components models contribute to more precise measurement of output gap. For confirmation or refusal of this hypothesis, Clark's model (1989) and Kuttner's model (1994) will be used. All two models will be analysed and their advantages and disadvantages will be shown. Results of estimations will be compared with papers dealing with output gap for czech economy. Then, the ability of output gap to forecast inflation will be analysed. At last, multivariate unobserved componenets model will be labeled after some tests as the most suitable method for estimation of output gap.

Forecasting macroeconomic variables using collapsed dynamic factor analysis

We explore a new approach to the forecasting of macroeconomic variables based on a dynamic factor state space analysis. Key economic variables are modeled jointly with principal components from a large time series panel of macroeconomic indicators using a multivariate unobserved components time series model. When the key economic variables are observed at a low frequency and the panel of macroeconomic variables is at a high frequency, we can use our approach for both nowcasting and forecasting purposes. Given a dynamic factor model as the data generation process, we provide Monte Carlo evidence of the finite-sample justification of our parsimonious and feasible approach. We also provide empirical evidence for a US macroeconomic dataset. The unbalanced panel contains quarterly and monthly variables. The forecasting accuracy is measured against a set of benchmark models. We conclude that our dynamic factor state space analysis can lead to higher levels of forecasting precision when the panel size and time series dimensions are moderate.

House Price Determinants: Fundamentals and Underlying Factors

This study examines the key drivers of real house prices in the five main Euro area countries (Germany, France, Italy, Spain, and the Netherlands) and the Anglo–Saxon economies (the United Kingdom and the United States) from 1970 to 2010. Estimating the determinants of house prices is very important, as they significantly influence economic activity and financial stability. Therefore, a multivariate unobserved component model, introduced by Harvey, has been used to model house price fluctuations. This technique allows us to catch those price movements that are not fully explained by economic fundamentals. The empirical results, in fact, indicate that in addition to changes in real income, long-run interest rates, stock prices and inflation, the latent component has a significant role in explaining real house prices. The latent component reflects those factors that are not specifically observed, such as structural changes in markets and changing preferences.

Ex-ante assessment of the Spanish Occupational Health and Safety Strategy (2007–2012) using a State Space framework

Spain is above the EU-27, EU-25 and EU-15 average in the number of serious work accidents, and it is estimated that Spain is five years behind the rest of Europe in reducing such accidents. To address the problem of the high number of occupational accidents, in 2007 the Spanish Occupational Health and Safety Strategy (SOHSS, 2007–2012) was launched, with the general aim of achieving a continuous and significant reduction in work accidents, approaching the EU average in occupational accidents and illness. This article is an attempt to assess the extent to which the aims of the SOHSS have been satisfied by predicting incidence rates for different levels of accident severity (slight, serious and fatal), accidents that do not require sick leave, and commuting accidents (slight, serious and fatal). With these objectives in mind, both Univariate and Multivariate Unobserved Components models are used in an enhanced State Space framework in order to deal with the irregular sampling interval of the data from 1998 to 2010.

Asset prices, credit and the business cycle

This paper uses the multivariate unobserved components model with phase shifts to analyse the interaction of interest rates, output, asset prices and credit in the US. We find close linkages amongst cyclical fluctuations in the variables.

How many lives can bloody and shocking road safety advertising save? The case of Spain

Abstract At the beginning of the 21st century, the punishment strategies used by the Spanish administration were considerably hardened to achieve a reduction in road accidents. This hardening could also be seen in the mass-media public advertising campaigns, with a marked shift from gentle messages to threat-based advertisements. The goal of this study is to evaluate the effectiveness of this radical change in terms of main road accident indicators and the time that the effects last using multivariate unobserved component models set up in a state space framework applied to monthly series for the 1980–2008 period. The main conclusion is that the effect on reducing road accidents is no greater when citizens are subjected to a greater level of threats in advertising campaigns than would on average be achieved using campaigns with a low level of threats. Secondly, the impact of bloody advertising in Spain is limited to the most serious accidents, those that cause deaths, either on highways or in built-up areas. Moreover, the positive effects progressively decrease as the average lifespan for significant effects of medium and high level campaigns on deaths on highways or in built-up areas was 8 months and almost 12 days. The results show that a reduction in numbers of deaths and injuries is always achieved when the level of harshness in the messages is increased after a period of several years of mild advertising.

Forecasting Macroeconomic Variables Using Collapsed Dynamic Factor Analysis

We explore a new approach to the forecasting of macroeconomic variables based on a dynamic factor state space analysis. Key economic variables are modeled jointly with principal components from a large time series panel of macroeconomic indicators using a multivariate unobserved components time series model. When the key economic variables are observed at a low frequency and the panel of macroeconomic variables is at a high frequency, we can use our approach for both nowcasting and forecasting purposes. Given a dynamic factor model as the data generation process, we provide Monte Carlo evidence for the finite-sample justification of our parsimonious and feasible approach. We also provide empirical evidence for a U.S. macroeconomic dataset. The unbalanced panel contain quarterly and monthly variables. The forecasting accuracy is measured against a set of benchmark models. We conclude that our dynamic factor state space analysis can lead to higher forecasting precisions when panel size and time series dimensions are moderate.

Taking Trends Seriously in DSGE Models: An Application to the Dutch Economy

We construct a new-Keynesian DSGE model tailored to the Netherlands and interpret it as a multivariate unobserved components model. We identify three major stochastic trends in the data — trends in general - purpose technology, investment-specific technology, and labor supply — and model them formally in our theoretical set-up. Our trend-cycle decomposition captures the data’s co-integrating properties without which long-run analysis — whether scenario analysis or forecasting — would likely be mis-specified. In particular, this approach appears to produce better-behaved posteriors for parameters along decision margins where traditional modeling imposes highly persistent but temporary shocks. The existence of permanent and temporary disturbances along the same margin broadens the scope for counter-factuals. Specifically, differences in short-run responses to the two types of shocks reflect smoothing motives and discounted valuation effects reminiscent of the Permanent Income Hypothesis.

From Many Series, One Cycle: Improved Estimates of the Business Cycle from a Multivariate Unobserved Components Model

We construct new estimates of potential output and the output gap using a multivariate approach that allows for an explicit role for measurement errors in the decomposition of real output. Because we include data on hours, output, employment, and the labor force, we are able to decompose our estimate of potential output into separate trends in labor productivity, labor-force participation, weekly hours, and the NAIRU. We find that labor-market variables—especially the unemployment rate—are the most informative individual indicators of the state of the business cycle. Conditional on including these measures, inflation is also very informative. Among measures of output, we find that although they add little to the identification for the cycle, the income-side measures of output are about as informative as the traditional product-side measures about the level of structural productivity and potential output. We also find that the output gap resulting from the recent financial crisis was very large, reaching -7 percent of output in the second half of 2009.

From Many Series, One Cycle: Improved Estimates of the Business Cycle from a Multivariate Unobserved Components Model

We construct new estimates of potential output and the output gap using a multivariate approach that allows for an explicit role for measurement errors in the decomposition of real output. Because we include data on hours, output, employment, and the labor force, we are able to decompose our estimate of potential output into separate trends in labor productivity, labor-force participation, weekly hours, and the NAIRU. We find that labor-market variables—especially the unemployment rate—are the most informative individual indicators of the state of the business cycle. Conditional on including these measures, inflation is also very informative. Among measures of output, we find that although they add little to the identification for the cycle, the income-side measures of output are about as informative as the traditional product-side measures about the level of structural productivity and potential output. We also find that the output gap resulting from the recent financial crisis was very large, reaching -7 percent of output in the second half of 2009.

Measuring the Euro area output gap using a multivariate unobserved components model containing phase shifts

Several recent studies have used multivariate unobserved components models to identify the output gap and the non-accelerating inflation rate of unemployment. A key assumption of these models is that one common cycle component, such as the output gap, drives the cyclical fluctuations in all variables included in the model. This article also uses the multivariate approach to estimate the euro area output gap and the trends and cycles in other macroeconomic variables. However, it adopts a flexible way of linking the output gap to the cycle components in the other variables, in that we do not impose any leading or lagging restrictions between cycle components, as has been done in most previous studies. Our approach also allows us to assess the strength of cycle association and cross-correlation among cycle components using the model’s parameter estimates. Finally, we demonstrate that our multivariate model can provide a satisfactory historical output gap estimate and also a ‘real-time’ estimate for the aggregate euro area.

Can fear of going to jail reduce the number of road fatalities? The Spanish experience

The goal of this article is to evaluate the impact of the drastic Spanish Penal Code reform on the number of road deaths in Spain and the time that the effects might last. This is achieved by means of multivariate unobserved component models set up in a state space framework estimated using maximum likelihood. In short, with this reform Spain might be considered to be closing the final gap that kept it apart from other developed countries as far as the road accident rate is concerned. We have found two different types of effects on Spanish road traffic fatalities. Initially, a month before the reform was passed there was a 24.7 percent fall in Spanish road deaths. After the Bill had been passed and for the following thirteen months, the reduction stayed at a constant sixteen percent. This reform has reduced Spanish road fatalities by 534 in all between November 2007 and December 2008 and the effects will foreseeably continue during 2009.

Estimating the neutral real interest rate in an emerging market economy

This study estimates the neutral real interest rate and output gap jointly under two different multivariate unobserved components models. In the analyses, Kalman filter is applied to a small-scale macroeconomic model of the Turkish economy to estimate the unobserved variables for the period 1989 to 2005. In addition, two alternative specifications for neutral real interest rate are exploited in the analyses. The first model uses a simple random walk model for the neutral rate, whereas the second one employs more structural specification, which specifically links the neutral rate with the trend growth rate and the long-term course of the risk premium by adapting the model proposed by Laubach and Williams (2003) to allow for small open economy consideration. Model evaluations clearly indicate the use of more structural specification against random walk specification. Results point out that the variation in the long-term course of the risk premium could be an important determinant of the neutral real interest rate in Turkey. Though there is relatively high uncertainty surrounding the neutral rate estimates to use them directly in the policy-making process, estimates appear to be very useful for ex post monetary policy evaluations.

Modelling and forecasting occupational accidents of different severity levels in Spain

The control of accidents at the work place is a critical issue all over the world. The consequences of occupational accidents in terms of costs for the company in which the accidents take place is only one minor matter, being the social impact and the loss of human life the most controversial effects of this important problem. The methods used to forecast future evolution of accidents are often limited to trend estimations and projections, being the scientific literature on this topic rather scarce. This paper aims at showing and predicting the evolution of Spanish occupational accidents of different levels of severity, allowing the evaluation of the influence that preventive actions carried out by public administrations or private companies may have over the number of occupational accidents. Though some contributions may be found on this topic for Spain, this paper is the first contribution that forecast occupational accidents for different levels of severity using Multivariate Unobserved Components models developed in a State Space framework extended to deal with the irregular sampling interval of the data. Data from 1998 to 2009 have been used to test the efficacy of the forecasting system.

Output Fluctuations in the G-7: An Unobserved Components Approach

This paper proposes a multivariate unobserved components model to simultaneously decompose the real GDP for each of the G-7 countries into their respective trend and cycle components. In contrast to previous literature, our model allows for explicit correlation between all the contemporaneous trend and cycle shocks. We find that all the G-7 countries have highly variable stochastic permanent components for output, even once we allow for structural breaks. We also find that common restrictions on the correlations between trend and cycle shocks are rejected by the data. In particular, we find that correlations across permanent and transitory shocks are important both within and across countries.

An econometric analysis of the effects of the penalty points system driver's license in Spain

This article seeks to quantify the effects of the penalty points system driver's license during the 18-month period following its coming into force. This is achieved by means of univariate and multivariate unobserved component models set up in a state space framework estimated using maximum likelihood. A detailed intervention analysis is carried out in order to test for the effects and their duration of the introduction of the penalty points system driver's license in Spain. Other variables, mainly indicators of the level of economic activity in Spain, are also considered. Among the main effects, we can mention an average reduction of almost 12.6% in the number of deaths in highway accidents. It would take at least 2 years for that effect to disappear. For the rest of the safety indicator variables (vehicle occupants injured in highway accidents and vehicle occupants injured in accidents built-up areas) the effects disappeared 1 year after the law coming into force.