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.     

Forecasting industrial production and the early detection of turning points

In this paper we propose a simple model to forecast industrial production in Italy up to 6 months ahead. We show that the forecasts produced using the model outperform some popular forecasts as well as those stemming from an ARIMA model used as a benchmark and those from some single equation alternative models. We show how the use of these forecasts can improve the estimate of a cyclical indicator and the early detection of turning points for the manufacturing sector. This is of paramount importance for short-term economic analysis.The first draft of this paper was mostly prepared while Claudio Lupi was working as a Research Director at ISAE. The present version of the paper is a revision of a preliminary draft that was circulated under the same title as ISAE working paper 20/01. The authors would like to thank two anonymous referees, audiences at ISAE, Dublin (22nd International Symposium on Forecasting), and Ente Einaudi (econometrics seminars) for comments. Fruitful discussions and suggestions from Gianluca Cubadda, Sergio de Nardis, John FitzGerald, Antonio García-Ferrer, Franco Peracchi, and Tommaso Proietti are gratefully acknowledged. None of them is responsible for any remaining error. The opinions expressed in this paper are solely the responsibility of the authors and should not be interpreted as reflecting the views of ISAE or its staff.First revision received: Ocotber 2002/Final revision received: May 2003     

A vector error-correction forecasting model of the US economy

Any research or policy analysis in economics must be consistent with the time-series properties of observed macroeconomic data. Numerous previous studies reinforce the need to specify correctly a model’s multivariate stochastic structure. This paper discusses in detail the specification of a vector error correction forecasting model that is anchored by long-run equilibrium relationships suggested by economic theory. The model includes six variables––the CPI, the GDP price index, real money balances (M1), the federal funds rate, the yield on long-term (10-year) government bonds, and real GDP––and four cointegrating vectors. The accuracy of VECM model forecasts for individual, univariate time series during for the 1990s is comparable to forecasts made by government agencies and private forecasters, perhaps because many forecasters share a similar implicit, long-run steady-state growth model of the economy. Judged by multivariate statistics that account for forecast-error covariance, VECM forecasts are found to be somewhat more accurate than a naïve random-walk alternative.     

Forecasting house prices for the four census regions and the aggregate US economy in a data-rich environment

This article considers the ability of large-scale (involving 145 fundamental variables) time-series models, estimated by dynamic factor analysis and Bayesian shrinkage, to forecast real house price growth rates of the four US census regions and the aggregate US economy. Besides the standard Minnesota prior, we also use additional priors that constrain the sum of coefficients of the VAR models. We compare 1- to 24-months-ahead forecasts of the large-scale models over an out-of-sample horizon of 1995:01–2009:03, based on an in-sample of 1968:02–1994:12, relative to a random walk model, a small-scale VAR model comprising just the five real house price growth rates and a medium-scale VAR model containing 36 of the 145 fundamental variables besides the five real house price growth rates. In addition to the forecast comparison exercise across small-, medium- and large-scale models, we also look at the ability of the ‘optimal’ model (i.e. the model that produces the minimum average mean squared forecast error) for a specific region in predicting ex ante real house prices (in levels) over the period of 2009:04 till 2012:02. Factor-based models (classical or Bayesian) perform the best for the North East, Mid-West, West census regions and the aggregate US economy and equally well to a small-scale VAR for the South region. The ‘optimal’ factor models also tend to predict the downward trend in the data when we conduct an ex ante forecasting exercise. Our results highlight the importance of information content in large number of fundamentals in predicting house prices accurately.     

Dynamic panel data models: a guide to micro data methods and practice

This paper reviews econometric methods for dynamic panel data models, and presents examples that illustrate the use of these procedures. The focus is on panels where a large number of individuals or firms are observed for a small number of time periods, typical of applications with microeconomic data. The emphasis is on single equation models with autoregressive dynamics and explanatory variables that are not strictly exogenous, and hence on the Generalised Method of Moments estimators that are widely used in this context. Two examples using firm-level panels are discussed in detail: a simple autoregressive model for investment rates; and a basic production function.JEL Classification: C23This paper is prepared for a special issue of the Portuguese Economic Journal. My understanding of this subject has benefitted immeasurably from the input of colleagues, notably Manuel Arellano, Richard Blundell, Costas Meghir, Steve Nickell and Frank Windmeijer. I thank Joao Santos Silva, Frank Windmeijer and other participants at the Cemmap/ESRC Econometric Study Group workshop in London on 22 February 2002 for detailed comments on an earlier draft. Financial support from the ESRC Centre for the Microeconomic Analysis of Public Policy is gratefully acknowledged.     

Econometric policy evaluation and expectations

Recently large-scale econometric models have been criticized as inappropriate tools for policy evaluation. One critical point often mentioned is the way expectations are formed. This paper shows that the way the process of expectations formation is considered does not play a crucial role. Instead, the ability to model agents' reactions to changes in policy is the essence of the critique. As long as government behaviour is sufficiently accessible in the model, applied econometrics may still use large-scale systems for which rational expectations solutions would be troublesome if not impossible to reach.     

Error correction modelling and dynamic specifications as a conduit to outperforming the random walk in exchange rate forecasting

The proposition that dynamic exchange rate models can outperform the random walk in out-of-sample forecasting, in the sense that they produce lower mean square errors, is examined and disputed. By using several dynamic versions of three macroeconomic exchange rate models, it is demonstrated that dynamic specifications outperform the corresponding static models but improvement in the forecasting power may not be sufficient for the dynamic models to perform better than the random walk. The results are explained by suggesting that any dynamic specification or transformation of the static model leads to the introduction of a lagged dependent variable, which in effect is a random walk component. The analysis leads to the conclusion that it is implausible to aim at beating the random walk by augmenting a static model with a random walk component.     

Validation of modeled pharmacoeconomic claims in formulary submissions

Abstract

Modeled or simulated claims for costs and outcomes are a key element in formulary submissions and comparative assessments of drug products and devices; however, all too often these claims are presented in a form that is either unverifiable or potentially verifiable but in a time frame that is of no practical use to formulary committees and others who may be committed to ongoing disease-area and therapeutic-class reviews. On the assumption that formulary committees are interested in testable predictions for product performance in target populations and ongoing disease area and therapeutic reviews, the methodological standards that should be applied are those that are accepted in the natural sciences. Claims should be presented in a form that is amenable to falsification. If not, they have no scientific standing. Certainly one can follow ISPOR-SMDM standards for validating the assumptions underpinning a model or simulation. There is clearly an important role for simulations as an input to policy initiatives and developing claims for healthcare interventions and testable hypotheses; however, one would not evaluate such claims on the realism or otherwise of the model. The only standard is one of the model’s ability to predict outcomes successfully in a time frame that is practical and useful. No other standard is acceptable. This sets the stage for an active research agenda.     

Worker heterogeneity and labor market volatility in matching models

Shimer demonstrated that aggregate productivity shocks in a standard matching model cause fluctuations in key labor market statistics—such as the job-finding rate, the vacancy/unemployment ratio, and the unemployment rate—that are too small by an order of magnitude [Shimer, R., 2005. The cyclical behavior of equilibrium unemployment and vacancies. American Economic Review 95 (1) 25–49]. This paper shows that when the standard model is extended to allow for worker heterogeneity, it exhibits considerably greater volatility. In the model, marginal workers, whose productivity only slightly exceeds the value of their alternative use of time, constitute a disproportionate share of unemployment on average, and that share rises when aggregate conditions deteriorate. These composition effects cause firms to open fewer vacancies during downturns.     

Assessing the predictive power of financial spreads in the euro area: does parameters instability matter?

In the empirical literature there is wide consensus that financial spreads cannot constitute a broadly based assessment on future output growth and inflation because the bivariate estimated regressions are not stable over time and lead to relatively poor out-of-sample forecasting performance (e.g. J Econ Liter 41:788–829, 2003). This conclusion arised for the USA, as well as for several European countries. In this paper we check whether the marginal predictive content of some financial spreads (the slope of the yield curve, the reverse yield gap and the credit spread) for macroeconomic forecasting in the euro area can be recovered using techniques taking into account potential parameters instability. We set up a quarterly Bayesian vector autoregression model with time-varying coefficients, comprising both target variables, as well as other monetary policy indicators, to serve as a benchmark. Then, the properties of the spreads as leading indicators are assessed by augmenting this benchmark BVAR with the spreads, one at a time. We find time variation of the coefficients to be a relevant issue in our model, especially for forecasting output growth, but financial spreads continue to have no or negligible marginal predictive content for both output growth and inflation. Overall, our results confirm that there is no ready-to-use financial spread that can replace an encompassing multivariate model for the prediction of target variables in the euro area.      

Gender Differences in Italian Children's Capabilities

This paper analyzes children's well-being using the capability approach, with a special focus on gender differences. The two areas analyzed are the capability of senses, imagination, and thought; and the capability of play. Using data from the 2008 Multipurpose Survey on Daily Life released by the Italian National Institute of Statistics, a structural equation model is estimated in which the capabilities are defined as latent variables that are intrinsically interrelated. For each capability, a set of indicators of functionings is utilized and the effects of individual and social conversion factors – including parents’ unpaid work, their level of education, and employment status – are analyzed. The model is applied to Italian girls and boys ages 6–10 in 2008. The analysis confirms that the two capabilities are interrelated. Policies aimed at improving children's achievements in education also improve the capability of play and vice versa. Differences by gender occur in the factors’ effects.     

Models as speech acts: the telling case of financial models

This paper intends to bring Austinian themes into methodological discussion about models. Using Austinian conceptual vocabulary, I argue that models perform actions in and outside of the academic field. This multiplicity of fields induces a variety of felicity conditions and types of performed actions. If for example, an inference from a model is judged according to some epistemological criteria in the scientific field, the representation of the world which the model carries will not be judged by the same criteria outside the scientific field. A model can be considered as a standard in a strict scientific framework, while not being used as part of public policies, or vice versa. However, we focus on the dynamics between different fields.     

From ecological footprint to ecological rent: An economic indicator for resource constraints

This paper takes as a starting point a combination of an input-output model with a national Ecological Footprint account for Germany in the spirit of Wiedmann et al. [Wiedmann, T., Minx, J., Barrett, J., Wackernagel, M., 2006. Allocating ecological footprints to final consumption categories with input-output analysis. Ecological Economics. 56, 28-48]. Footprint as well as Biocapacity is dealt with at the industry level. Gross output of each industry and final demand for each industry can then be split up into a share that is reconcilable with Biocapacity and another share that corresponds to the ‘Ecological Deficit’. The Ecological Footprint concept is extended in this study by introducing the additional biocapacity necessary for sustaining the given level of economic activity. It is assumed that each industry had to rent the corresponding areas and to apply a given technology in order to make this additional land biophysically productive. That results in an additional technology for each industry leading to an increase in costs and prices. The new price level is directly linked to the share of output that corresponds to Biocapacity overshooting, which is defined by the ‘Ecological Deficit’. Economic indicators can be derived by measuring the income difference brought about by the price increase. This difference corresponds to a Ricardian rent which is due to resource constraints on output growth.     

The predictive value of subjective labour supply data: A partial-adjustment model with measurement error

This paper tests the predictive value of subjective labour supply data for adjustments in working hours over time. The idea is that if subjective labour supply data help to predict working hours, the subjective data must contain at least some information on individual labour supply preferences. In this paper, I formulate a partial-adjustment model that allows for measurement error in the observed variables. Applying estimation methods that are developed for dynamic panel data models, I find evidence for a predictive power of subjective labour supply data concerning desired working hours in the German Socio-Economic Panel 1988–1995.I wish to thank John Haisken-DeNew, Astrid Kunze, Markus Pannenberg, Winfried Pohlmeier, Frank Windmeijer, Rainer Winkelmann, my former colleagues at IZA and two anonymous referees for their valuable comments. The author gratefully acknowledges DIW for providing the data.First version received: December 2001/Final version received: December 2003     

A competitive model of economic geography

Summary. Most of the literature argues that competitive analysis has nothing interesting to say about location. This paper argues, to the contrary, that a competitive model can have something interesting to say about location, provided that locations are not identical and transportation costs are not zero. To do this, it constructs a competitive intertemporal general equilibrium model and applies it to a suggestive example of migration.Received: 25 August 2003, Revised: 18 December 2003, JEL Classification Numbers: D5, R0.Our interest in this topic has been stimulated over the years by many conversations with Marcus Berliant. We thank an anonymous referee for exceptionally careful and useful comments. Financial support from the UCLA Academic Committee on Research (Ellickson, Zame) and the National Science Foundation (Zame) is gratefully acknowledged. Views expressed here are those of the authors and do not necessary reflect the views of any funding agency.     

Adapting and testing the Fama and French model,with some variations of company characteristics

We examine whether the Fama and French (1992) (F&F) model can be adapted to become a more versatile and flexible tool, capable of incorporating variations of company characteristics in a more dynamic form. For this, the risk factors are reconstructed at the end of each reading of monthly data. We argue that, over time, the evaluation of a company may change as a result of variations in its market price, size or book price, and we are aware that the F&F model does not accurately reflect these dynamics. Our results show that the adapted model is able to capture the behaviour of a greater number of stocks than the original F&F model and risk factors are more significant when building them through our procedure. In addition, we carry out these adaptations during a period of instability in financial markets.     

Tax tilting and politics: Some theory and evidence for Latin America

A government budget deficit can exist for at least two possible reasons: tax smoothing and/or tax tilting. Under tax-smoothing, deficits are temporary phenomena resulting from the decision not to vary the tax rate in response to fluctuations in government spending (as a share of output). This is done in order to minimize the distortionary cost of taxes. Tax tilting occurs whenever the government has an incentive to discount the losses to society from taxes at a higher rate than society discounts them; hence it delays taxes or advances spending introducing an upward trend in total government debt. This paper develops a model that implies that tax-tilting tends to increase with political risk. An increase in political risk, measured by the probability of losing power, increases the rate at which the government discounts the future, causing government policy to be relatively more myopic. Hence it delays taxes or advances spending and its deficit increases. Using data from a panel of 19 Latin-American countries for the period 1984–2009, the paper presents estimation results that strongly support the proposition that an increase in political risk increases the degree of tax-tilting.     

Business models for material efficiency services: Conceptualization and application

Despite the abundant research on material flows and the growing recognition of the need to dematerialize the economy, business enterprises are still not making the best possible use of the many opportunities for material efficiency improvements. This article proposes one possible solution: material efficiency services provided by outside suppliers. It also introduces a conceptual framework for the analysis of different business models for eco-efficient services and applies the framework to material efficiency services. Four business models are outlined and their feasibility is studied from an empirical vantage point. In contrast to much of the previous research, special emphasis is laid on the financial aspects. It appears that the most promising business models are ‘material efficiency as additional service’ and ‘material flow management service’. Depending on the business model, prominent material efficiency service providers differ from large companies that offer multiple products and/or services to smaller, specialized providers. Potential clients (users) typically lack the resources (expertise, management's time or initial funds) to conduct material efficiency improvements themselves. Customers are more likely to use material efficiency services that relate to support materials or side-streams rather than those that are at the core of production. Potential client organizations with a strategy of outsourcing support activities and with experience of outsourcing are more keen to use material efficiency services.     

Nonlinear dynamics: Evidence for a small stock exchange

This paper models the main stock index of the Vienna Stock Exchange with daily data from 1986 to 1992. We find that returns are nonnormal and show linear and nonliner dependence. On that basis we compare the fit of alternative specifications of Generalized Autoregressive Conditional Heteroscedasticity (GARCH) to the Markov-Switching approach. The models are evaluated with diagnostic tests on the standardized residuals. We consider evidence for deterministic structures and for infinite variance. Our main result is that a parsimonious model from the GARCH – class can generate the statistical properties of daily returns. The behavior of the two types of models with respect to temporal aggregation is found to differ significantly. First version received: January 1996/Final version received: December 1997     

Stability of supply function equilibria implications for daily versus hourly bids in a poolco market

We consider a supply function model of a poolco electricity market where demand varies significantly over a time horizon such as a day and also has a small responsiveness to price. Although there are equilibria yielding prices at peak that are close to Cournot prices, it is known that the wider the range of demand uncertainty the narrower the range of such supply function equilibria. Here we show that such equilibria are also typically unstable and consequently would be difficult to sustain in practice. This strengthens the results of Green and Newbery by ruling out many equilibria that have high prices. We demonstrate this result both theoretically under somewhat restrictive assumptions and also numerically using both a three-firm example system and a five-firm example system having generation capacity constraints. Hence, this reinforces the conclusion that the market outcome is significantly influenced by a requirement that offers into the poolco be consistent over the time horizon. This result contrasts with markets where bids can be changed on an hourly basis, where Cournot prices are possible outcomes. The stability analysis has important policy implications for the design of day-ahead electricity markets. The stability perspective also provides a narrowing of the equilibrium selection that strengthens empirical analysis.