No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
An Automatic Leading Indicator of Economic Activity: Forecasting GDP Growth for European Countries
Abstract
In the construction of a leading indicator model of economic activity, economists must select among a pool of variables which lead output growth. Usually the pool of variables is large, and selection of a subset must be carried out. In this paper we propose an `Automatic Leading Indicator' model. Rather than preselection, we use a dynamic factor model to summarise the information content of a pool of variables. Results show that the forecasting performance of our `Automatic Leading Indicator' model is signi#cantly better than that of traditional model selection criteria with VAR models. This study is carried out using quaterly data for France, Germany, Italy and the United Kingdom. KEYWORDS: Forecasting, State Space Models, Time Series. 1 Introduction The number of variables that can be used when forecasting output growth with a VAR model is limited. Given typical sample periods available in applied macroeconomics, forecasters must preselect a subset from a pool of variables...
... In general, methodologies can be divided in two broad categories characterised by the size of the set of explanatory variables taken into consideration. Stock and Watson (1991), Camba Mendez et al. (2001), Aruoba et al.. (2009, Aruoba and Diebold (2010) and Camacho and Perez-Quiros (2010) are some examples where a small number of wisely selected predictors is used. Then, under the assumption of non cross-correlated errors the factor models are estimated by maximum likelihood using the Kalman filter. ...
... BIC, HQ). The benchmark is the Automatic Leading Indicator (ALI) model in the spirit of Camba Mendez et al. (2001) which belongs to the category of small scale dynamic factor models. The rest of the paper is organised as follows: Section 2 briefly describes the methodologies, Section 3 is concerned with the forecasting algorithm, Section 4 discusses the results and Section 5 summarises the conclusions. ...
... The last methodology is the automatic leading indicator (ALI) model as introduced by Camba-Mendez et al. (2001). This is a small scale dynamic factor model and includes a two-step procedure where in the first step the factors are extracted and in the second stage a VAR model is estimated and used for forecasting. ...
This paper assesses the forecasting performance of various variable reduction and variable selection methods. A small and a large set of wisely chosen variables are used in forecasting the industrial production growth for four Euro Area economies. The results indicate that the Automatic Leading Indicator (ALI) model performs well compared to other variable reduction methods in small datasets. However, Partial Least Squares and variable selection using heuristic optimisations of information criteria along with the ALI could be used in model averaging methodologies.
... Engle et al. (1985) briey consider the possibility that their 2-factor model is not identied, but point out that lack of identiability does not cause problems with the EM algorithm that they employ, and don't consider the matter further. Camba-Mendez et al. (2001) prove identication for a k-factor model of the form x t = Bf t + ε t , under the following conditions ...
... In the case of dynamic factor models, this is complicated by the presence of lags. Geweke and Singleton (1981) and Camba-Mendez et al. (2001) have shown that restrictions on the factor loading, similar to those used in the case of static factor analysis, are sucient to identify the factors in the two-sided and one-sided factor model respectively. Therefore, identication is no more of a problem in the dynamic factor model than it is in the static model. ...
... Since autoregressions are identied from their unconditional second moments, all the parameters of the model are identied. Zero restrictions, or the unit restrictions of Camba-Mendez et al. (2001) are redundant in this case, provided that the irreducibility and rank assumptions on β(L) are satised. This is a remarkable result since it implies that the factor estimates from such models can be interpreted far more readily than is the case for static factor estimates. ...
... The recent revival of leading indicator models is largely due to the work of Stock and Watson, who proposed to extract, by means of dynamic factor analysis, from a large pool of variables a latent 'leading indicator', or an 'index of coincident indicators' as they call it, for the US economy, e.g. see (Stock and Watson, 1989;1991). 2 The 'automatic leading indicator' (ALI) model proposed by Camba-Mendez et al (2001) makes use of very similar techniques as in (Stock and Watson, 1989). 3 However, the angle of application has been re-oriented. ...
... 3 However, the angle of application has been re-oriented. Camba-Mendez et al (2001) focus their attention on shortterm forecasts of certain officially released variables of interest, e.g. real GDP growth of selected European countries. ...
... The ALI method used here is adopted from (Camba-Mendez et al, 2001). The method consists of two steps: factor extraction using a DFM and forecasting using a VAR. ...
This paper compares forecast performance of the ALI method and the MESMs and seeks ways of improving the ALI forecasts. Inflation and GDP growth form the forecast objects for comparison, using data from China, Indonesia and the Philippines. The ALI method is found to produce better short-term forecasts than those by MESMs in general, but the method is found to involve greater uncertainty in choosing indicators, mixing data frequencies and utilizing unrestricted VARs. Two possible improvements are found helpful to reduce the uncertainty: (i) give theory priority in choosing indicators and include theory-based disequilibrium shocks in the indicator sets; and (ii) reduce the VARs by means of the general → specific model reduction procedure.
... Some studies instead propose a "green GDP" measure (Talberth & Bohara, 2006). In any case, GDP continues to be the most widely used measure of economic growth (Camba-Mendez, Kapetanios, Smith, & Weale, 2001), and it is used in the current study. ...
... If there is a surplus, it can accumulate as wealth (i.e., GDP). As mentioned, at the national level, an increase in GDP connotes economic growth (Camba-Mendez et al., 2001). Meanwhile, fertility is achieved at the childbearing age, and reproductive behavior is affected by government policies, such as family-planning policies and the two-child policy . ...
The current study seeks to determine the type of population policy China should implement to effectively promote economic growth, especially in the context of its aging population. Using multiagent simulation technology, I integrate the two-child policy into a complex model featuring population, resources, and economic growth. My simulation results indicate that under the constraining effect of aging on economic growth, the implementation of a two-child policy alone could promote economic growth, albeit in a limited fashion: it would be more economically advantageous to combine a two-child policy with policy that promotes human capital growth. These findings provide new evidence regarding the relationship between population aging and economic growth. It is insufficient to emphasize only a liberalization of birth restrictions and promote population growth: rather, a combination of various policies would be more successful. A cluster of policies that aim to increase both fertility and investment in human capital can effectively curb economic recession otherwise caused by an aging population, and such a policy cluster could concurrently increase the supply of new labor and improve skills that appreciate with age.
... A second difficulty for predictions is a high volatility in the leading indexes. Caglayan and Xu (2016) show that this would occur for several leading indexes from about 2005 to 2012 in Germany, and Camba-Mendez et al. (2001) suggest that volatile periods would require rich models including several leading indicators. Volatility is also referred to as an "Investor fear gauge" (Xu and Zhou 2018). ...
... Smoothing of time series is discussed by Ozyildirim et al. (2010). In contrast, Camba-Mendez et al. (2001) use an intervention model to a priori filter out particular anomalous events. With the LL method, the LL algorithm detects such events as anomalies in the LL relations. ...
We apply a relatively novel leading–lagging (LL) method to four leading and one lagging indexes for industrial production (IP) in Germany. We obtain three sets of results. First, we show that the sentiment-based ifo index performs best in predicting the general changes in IP (−0.596, range −1.0 to 1.0, −1.0 being best). The ZEW index is very close (−0.583). In third place comes, somewhat unexpectedly, the behavioral-based unemployment index (−0.564), and last comes order flow, OF (−0.186). Second, we applied the LL method to predefined recession and recovery time windows. The recessions were best predicted (−0.70), the recoveries worst (−0.32), and the overall prediction was intermediate (−0.48). Third, the method identifies time windows automatically, even for short time windows, where the leading indexes fail. All indexes scored low during time windows around 1997 and 2005. Both periods correspond to anomalous periods in the German economy. The 1997 period coincides with “the great moderation” in the US at the end of a minor depression in Germany. Around 2005, oil prices increased from 60 a barrel. There were few orders, and monetary supply was low. Our policy implications suggest that the ZEW index performs best (including recessions and recoveries), but unemployment and monetary supply should probably be given more weight in sentiment forecasting.
... The ability to track and even anticipate economic activity has been an ongoing endeavor of economic researchers. The debate over the use of a single or collection of leading indicators has been long standing (see Stock and Watson [1,2]; Cecchetti et al. [3]; Camba-Mendez et al. [4], among others). In addition to the use of leading indicators, the predictive content of financial assets in relation to economic activities has also been extensively examined (see Fama [5]; Cochrane [6]; Restoy and OPEN ACCESS Rockinger [7]; Boldrin et al. [8]; Estrella et al. [9]; Vassalou [10]; Duarte et al. [11]; Hong et al. [12]; Clements and Galvao [13]; Cooper and Priestley [14]; Marcellino and Schumacher [15], amongst others). ...
... and IP jt = α i + δ i t + a 1 BDI t + a 2 MSCI t + η jt (4) In addition, the long-run specifications incorporating oil prices are noted as follows in Equations (5) and (6): AP it = α i + δ i t + a 1 BDI t + a 2 OP t + ε it (5) and IP jt = α i + δ i t + a 1 BDI t + a 2 OP t + η jt (6) As in Equations (1) and (2), we are interested in the statistical significance of the a 1 coefficient, in light of the presence of the alternative indicators. The new panel cointegration test results incorporating the above specifications are reported in Table 5. ...
This empirical study analyzes the information and predictive content of the Baltic Dry Index (BDI) with respect to a range of financial assets and the macroeconomy. By using panel methodological approaches and daily data spanning the period 1985–2012, the empirical analysis documents the joint predictability capacity of the BDI for both financial assets and industrial production. The results reveal the role of the BDI in predicting the future course of the real economy, yielding a link between financial asset markets and the macroeconomy.
... To attain improvement in timeliness, literature uses a strategy that has been employed to forecast quarterly aggregates of monthly indicators based on VAR models such as Camba-Mendez et al. (2001). Another strategy, which is indeed familiar to statistical offices, considers estimation of GDP growth when at least for some indicators there may be an incomplete set of within-quarter information, perhaps only two months of published data are available, and the final month in the quarter must then be forecasted. ...
... Alternatives to principal components analysis are identification and estimation of the factors using a parametric model. For example, the state-space approach can be used when the set of indicator variables is quite small (say < 12); e.g., seeStock and Watson (1989) andCamba-Mendez et al. (2001).9 Boivin and Ng (2005) show that the key difference of these two approaches is that the latter approach extracts the factors from the unobserved common to all information variables component. ...
In this paper, simple regression estimates and factor-based models are utilised to produce forecasts for Bahrain quarterly gross domestic product growth. Using simulated out-of-sample experiments, we assess and compare the performance of the simple regression estimates, which exploit the available information on selected indicator variables, with factor-based estimates. These estimates use up to 65 variables to obtain new factors that embody most of the potential information and handle it in a systematic way following the Stock–Watson approach. Additionally, we compare the performance of the nowcast factor MIDAS with the quarterly factor (static-SW) models based on time-aggregated data, which neglect the most recent information. Our empirical findings can be summarised as follows. First, using more information does not help to produce more accurate results. Preselected indicator variables can clearly improve the forecast performance in comparison with the use of large dataset. Second, quarterly factor models are in general outperformed by the nowcast factor models that directly relate low-frequency data to those of high frequency. Third, the best forecasting performance can be reached using simple regression estimates with a handful of variables. However, it fails in density forecast evaluation tests. Thus, the alternative factor-MIDAS model is considered as an optimal model that passes all the performance evaluation tests. Fourth, concerning the difference between MIDAS projection methods, the results indicate that MIDAS with exponential distributed lag functions outperforms the MIDAS with unrestricted lag polynomials. The best performing projection based on the number of factors is the model with three factors. They can pick up the rapid switch in the utility of the indicators automatically. Finally, although the most accurate Flash estimates are obtained at 84 days, nowcasting using industrial production into a bridge equation witnessed insignificant loss in accuracy at 54 days only.
... Before Brexit and the COVID-19 pandemic, several research studies have been undertaken to forecast the UK economic growth [17]- [19]. However, The UK economy after Brexit and COVID-19 pandemic has been very dynamic, uncertain, and highly unpredictable [20]. ...
... Numerous studies have examined the macroeconomic situation's behavior and created models to predict it. Researchers proposed in [9] an automatic leading indicator for forecasting GDP growth in European countries. They found that the overall forecasting performance of the proposed indicator overperforms the traditional VAR and BVAR models. ...
Gross Domestic Product (GDP) is one of the key macroeconomic aggregates that measures the added value produced in a country during a period. In the contemporary world, macroeconomic uncertainty, among others due to the COVID-19 pandemic and the conflict in Ukraine, and GDP prediction remain important goals in public policy making. This study aims to predict Benin's GDP through a unidimensional statistical approach and machine learning techniques. For this purpose, GDP data were collected from the Central Bank of the West African States (BCEAO) website from 1960 to 2021. The predictions are based on comparing classical statistical and machine learning methods. For the classical statistical methods, we investigated the Autoregressive Integrated Moving Average (ARIMA) and Error Trend Seasonality (ETS) forecasting models. As for the machine learning methods, the K-Nearest Neighbors (KNN) and Long Short-Term Memory (LSTM) forecasting models proved to be sound. The findings revealed that the statistical models (ARIMA and ETS) better predict Benin's GDP. However, machine learning models (KNN and LSTM) also provide a wide range of results that can be used to analyze Benin's economic growth.
... Since GDP (Gross Domestic Product) is a key indicator of marketrelated economic activities in a country, we use provincial GDP per capita to estimate provincial socioeconomic developments (51). Since business activities (design, production, and technical marketing) require provincial government regulation as a guide (38,39), the food producers constitute a source of pressure upon government for We obtain the census data for FISV from the China Food Industry Yearbook (52). ...
Introduction
According to China’s Food Safety Law of 2015, the filing of food safety enterprise standards is a policy innovation led by p9rovincial governments in China. However, there are significant differences in the development of the “Food Safety Enterprise Standard Filing Policy” between provincial governments across the country. This study aims to explore the internal mechanisms driving autonomous innovation by provincial governments in the absence of administrative pressure from the central government, to better understand the policy innovation mechanism in the Chinese context.
Methods
Crispy Set Qualitative Comparative Analysis (csQCA) method is used to identify the innovation mechanism.
Results
This study found that provinces with good provincial economic resources and strong government capabilities are prone to policy innovation, and the influence of internal factors of provincial governments is stronger than that of external factors.
Discussion
When provincial economic resources and capacity are weak, endogenous factors in the province also help achieve proactive policy innovation by provincial governments. The research results reveal how provincial governments construct local policies in the absence of administrative pressure from the central government.
... Initial literature used basic Autoregression (AR), Vector Autoregression (VAR), and Autoregressive Integrated Moving Average (ARIMA) statistical models for forecasting. With time works of literature started using more complex models like Chow-Lin related series technique (Abeysinghe & Rajaguru, 2004), automatic leading indicator model (Camba-Mendez et al., 2001), alternative factor models (Schumacher & Breitung, 2008), and Bridge models (Baffigi et al., 2004) for forecasting. Pieces of literature have also focused on the use of alternate datasets for this purpose, like the use of business tendency surveys data for forecasting short-term GDP growth for Sweden (Hansson et al., 2005). ...
... GDP growth statistics has been a primary way in which national economic performance is tracked since it is the aggregate statistic of all economic activity and it captures a broader coverage of the economy than other macro-economic variables (Wabomba, 2016). Moreover, there is demand for forecasts of major economic variables since the monetary policy in some countries is set with reference to expectations of output growth (Camba, 2001). Financial markets participants who rely on the impact of such policy frameworks to make decisions in the market also have a keen interest on such predictions of GDP. ...
Background:
Gross Domestic Product (GDP) is the market value of goods and services produced within a selected geographical area usually a country in a selected interval in time often a year and can be measured and forecasted in different ways for use by governments and other market participants .Specific users of information on GDP analysis include the United Nations Sustainable Development Goal assessment whose key indicator is economic growth as measured by GDP and the joint International Monetary Fund-World Bank methodology for conducting standardized debt-sustainability analyses in low-income countries.
Objective:
The main objective of this study was to assess the superiority as suggested by Literature of a Hybrid Autoregressive Integrated Moving Average(ARIMA) and feed forward Artificial Neural Network (ANN) model over a pure ARIMA model in forecasting Kenya s GDP. Methods: The ARIMA and the additive ANN-ARIMA Hybrid model is used to forecast absolute GDP values and the comparative forecast accuracy is tested using the RMSE and visualization plots. The Box-Jenkins methodology is used to fit the ARIMA model while the feed-forward Neural Network Autoregressive(NNAR) structure is used to model the neural network portion of the hybrid model.
Results:
The data analysis results indicate that the Hybrid model made up of the ARIMA(2,2,1) model and the NNAR (5,2) model does not outperform the pure ARIMA model in forecasting Kenya's short term out of sample GDP based on the results in which a value higher than the ARIMA model's Root Mean Squared Error (RMSE) value is obtained. The ARIMA forecasts outperform the Hybrid model forecasts by 30% on the same basis. However when the two models are compared against an industry benchmark-the IMF GDP forecasts, the Hybrid model yields forecasts closer to this benchmark than the ARIMA forecasts. In conclusion, the Hybrid model has great potential to compete favourably with the ARIMA model in forecasting short term GDP, if a method that accurately rather than arbitrarily specifies the optimum NNAR parameters is developed to reduce the computation costs of repeated trials that has been used in this study to obtain the Neural 2 Network specifications of the Hybrid model. This is further supported by the Diebold and Mariano statistic that pointed towards equal forecast accuracy ability between the two models.
... The existing literature has mainly focused on predicting GDP growth in the United States (Batchelor and Dua 1992;Batchelor and Dua 1998;Stock and Watson 2002;Clements and Galvão 2008;Marcellino 2008;Clark 2011;Clark and Ravazzolo 2015;Barsoum and Stankiewicz 2015;). On the other hand, a large majority of studies have developed predictions of GDP growth for the euro area and in specific countries of Europe, mainly for Scandinavian economies (Bergstroöm 1995;Camba-Mendez et al. 2001;Hansson et al. 2005;Martinsen et al. 2014;Smets et al. 2014;Kapetanios et al. 2016;Marcellino et al. 2016;Claveria et al. 2019). For their part, Ferrara et al. 2015) developed GDP growth prediction models in 19 OECD countries. ...
Precise macroeconomic forecasting is one of the major aims of economic analysis because it facilitates a timely assessment of future economic conditions and can be used for monetary, fiscal, and economic policy purposes. Numerous works have studied the behavior of the macroeconomic situation and have developed models to forecast them. However, the existing models have limitations, and the literature demands more research on the subject given that the accuracy of the models is still poor, and they have only been expanded for developed countries. This paper presents a comparison of methodologies for GDP growth forecasting and, consequently, new forecasting models of GDP growth have been constructed with the ability to estimate accurately future scenarios globally. A sample of 70 countries was used, which has allowed the use of sample combinations that consider the regional heterogeneity of the warning indicators. To the sample under study, different methods have been applied to achieve a high accuracy model, comparing Quantum Computing with Deep Learning procedures, being Deep Neural Decision Trees, which has provided excellent prediction results thanks to large-scale processing with mini-batch-based learning and can be connected to any larger Neural Networks model. Our model has a great potential impact on the adequacy of macroeconomic policy, providing tools that help to achieve macroeconomic and monetary stability at the global level, and creating new methodological opportunities for GDP growth forecasting.
... y los que usan indicadores económicos(Camba-Mendez et al., 2001). Entre los métodos más modernos están los factoriales dinámicos(Bánbura y Rünstler, 2011) y los que emplean datos de panel a corto plazo(Angelini et al., 2011).Soytas y Sari (2003) realizaron una recopilación de los diferentes trabajos que se hicieron hasta la fecha relativos a la relación causal entre la generación o el22 La primera versión oficial de la estimación del crecimiento del PIB en la zona euro la proporciona de manera preliminar Eurostat.. Esta se publica seis semanas después del final del trimestre de referencia. ...
El objetivo de esta tesis es comprobar cómo una cartera de proyectos similares financiados mediante Project Finance puede contener un riesgo no diversificable en las colas muy superior al que cabría esperar del análisis individualizado de cada uno de los proyectos que la componen. Para ello, en primer lugar, se estudia el comportamiento del proyecto de manera aislada. En segundo lugar, se analiza una cartera de proyectos considerando distintas fuentes de incertidumbre que afectan al riesgo de cada uno de los proyectos que se combinan mediante una cópula de variables aleatorias para generar la distribución de los rendimientos de los proyectos.
... Their study shows that the prediction of GDP using ANNs outperforms the existing linear models. In the time series domain, Camba-Mendez, Kapetanios, Smith, & Weale (2001) proposed the Automatic Leading Indicator Model, which exhibited better performance than vector autoregression (VAR) models for predicting GDP growth of European countries. The use of Markov regimes switching model, by Yan & Qiu (2007), for China's quarterly GDP growth ratio series has demonstrated some success to represent the nonlinear characteristics present in this prediction problem. ...
Support Vector Regression (SVR) has already been proved to be one of the most referred and used machine learning technique in various fields. In this study, we have addressed a predictive-cum-prescriptive analysis for finalizing fund allocations by the Government at center to the schemes under Central Plan and to the schemes under States and Union Territories Plan, with a goal to maximize Gross Value Added (GVA) at factor cost. Here, we have proposed a hybrid machine learning model comprising of OFS (Orthogonal Forward Selection), TLBO (Teaching Learning Based Optimization) and SVR for the prediction of GVA at factor cost. In this model, referred as OFS–TLBO–SVR hybrid model, SVR is at the core of prediction mechanism, OFS is for identifying the relevant features, and TLBO is to support in optimizing the free parameters of SVR and again TLBO is used for optimizing the governable attributes of data.
... Up to now, there is no econometric theory for the decomposition of the static factors into lagged dynamic factors as in (1). 7 But the empirical literature shows that the inclusion of lags of the factors may improve the forecasting ability of the models. 8 Therefore, in order to obtain such a dynamic factor model, the literature suggests to use empirical criteria such as forecasting accuracy to determine the number of dynamic factors q and the corresponding lags p. ...
This paper discusses a large-scale factor model for the German economy, Following the recent literature, a data set of 121 time series is used to determine the factors by principal component analysis. The factors enter a linear dynamic model for German GDP. To evaluate its empirical properties, the model is compared with alternative univariate and multivariate models. These simpler models are based on regression techniques and considerably smaller data sets. Empirical forecast tests show that the large-scale factor model almost always encompasses its rivals. Moreover, out-of-sample forecasts of the large-scale factor model have smaller prediction errors than the forecasts of the alternative models. However, these advantages are not statistically significant, as a test for equal forecast accuracy shows. Therefore, the efficiency gains of using a large data set with this kind of factor models seem to be limited.
... Indeed, when many potential predictors are available, one can follow basically two different routes: either only a small number of predictors are selected and considered in the forecasting model, or some way of reducing the dimensionality of the forecasting problem must be found via data reduction techniques. In fact, many alternative techniques have been proposed in the literature to exploit the information conveyed by many predictors, e.g. by building leading economic indicators (see De Leeuw, 1991;Camba-Mendez et al., 2001;Stock and Watson, 2003;Banerjee et al., 2005;Marcellino, 2006, among others). Other approaches rely on multivariate techniques, notably reduced rank regressions and canonical correlations analysis (see, e.g. ...
In recent years there has been a growing interest in using consumer surveys as leading indicators of real economic activity, especially to assess the future path of private consumption. However, consumer surveys data amount to large data sets, often condensed in rather ad hoc measures of the so-called consumer sentiment. In this paper we are interested in evaluating the pros and cons deriving from using consumer surveys and canonical correlations to forecast private consumption components in real time. In the present framework , there are different reasons to consider canonical correlations. (1) Canonical correlations lead to optimal forecasting linear combinations ; (2) the solution offered by canonical correlations should be more flexible than that offered by, for example, vector autoregres-sions (VARs), given that the information set has not to be uniformly updated to time t − 1; (3) it should be possible to use a much larger set of potentially important predictors than allowed by VAR analysis ; (4) beside producing forecasts of private consumption growth, using canonical correlations we are able to offer an alternative measure of the consumer sentiment. The forecasting performance of the proposed methods is compared to that of some widely used benchmarks using Italian data.
... These indicators include quantitative indicators, such as industrial production, confidence surveys and composite indicators. The forecast properties of business-cycle indicators have been examined by Parigi and Schlitzer (1995), Camba-Mendez et al. (2001), Baffigi, Golinelli, and Parigi (2002), Banerjee, Marcellino, and Masten (2003), Mourougane and Roma (2003), Rünstler and Sédillot (2003), Sédillot and Pain (2003), Gayer (2005) and Golinelli and Parigi (2007) for a number of OECD countries, which has shown that short-term forecasts of real GDP growth derived from such indicators usually perform properly. ...
This paper evaluates short-term forecasts of real GDP in the Euro area derived from the CESifo Economic Climate indicator (WES) in terms of forecast accuracy. We compare the forecast properties of the WES with those of monthly composite indicators. Considering the WES is interesting because (i) it is exclusively based on the assessment of economic experts about the current economic situation, and (ii) it is timely released within the quarter on a quarterly basis. The empirical analysis is carried out under full information , which means that the competing monthly indicators are known for the entire quarter, and under incomplete information. Our findings exhibit that the forecast power of the WES is comparatively proper.
... Bandholz/Funke (2001) estimate a multivariate state-space model with and without regime-switching to determine a business cycle factor. Breitung/Jagodzinski (2002) give an overview about other composite 1 See Geweke (1977), Sargent/Sims (1977), Stock/Watson (1991), Camba-Mendez et al. (2001). 2 See Stock/Watson (1999). ...
This paper discusses a large-scale factor model for the German economy. Following the recent literature, a data set of 121 time series is used to determine the factors by principal component analysis. The factors enter a linear dynamic model for German GDP. To evaluate its empirical properties, the model is compared with alternative univariate and multivariate models. These simpler models are based on regression techniques and considerably smaller data sets. Empirical forecast tests show that the large-scale factor model almost always encompasses its rivals. Moreover, out-of-sample forecasts of the large-scale factor model have smaller prediction errors than the forecasts of the alternative models. However, these advantages are not statistically significant, as a test for equal forecast accuracy shows. Therefore, the efficiency gains of using a large data set with this kind of factor models seem to be limited.
... Since the seminal work of Stock and Watson, the single-index model, also called the dynamic factor model, has been widely used. The recent literature includes Camba-Mendez et al . (2001) and Garcia-Ferrer and Poncela (2002), who modify the model to forecast GDP growth for several European economies, and Bandholz and Funke (2003), who use the model to develop leading and coincident indicators of economic activity in Germany. Fukuda and Onodera (2001) and Chen and Lin (2000) apply the model to Japan and to Taiwan, respect ...
... Since the seminal work of Stock and Watson, the single-index model, also called the dynamic factor model, has been widely used. The recent literature includes Mendez et al . (2001) and Ferrer and Poncela (2002), who modify the model to forecast GDP growth for several European economies , and Bandholz and Funke (2003), who use the model to develop leading and coincident indicators of economic activity in Germany. Fukuda and Onodera (2001) and Chen and Lin (2000) apply the model to Japan and to Taiwan, respectively. ...
... The main factor model used in the past to extract dynamic factors from economic time series has been a state space model estimated using maximum likelihood. This model was used in conjunction with the Kalman filter in a number of papers carrying out factor analysis (see, among others, Stock and Watson (1989) and Camba-Mendez et al (2001)). However, maximum likelihood estimation of a state space model is not practical when the dimension of the model becomes too large due to the computational cost. ...
The estimation of dynamic factor models (DFMs) for large sets of variables has attracted considerable attention recently, due to the increased availability of large datasets. Yet, identification issues com-bined with non-parametric estimation prevented a structural interpre-tation of the results. In this paper we propose a method for identifying shocks and factors in parametric DFMs for large datasets. We use a state space formulation of the model for estimation and computation of impulse response functions. The same set of identification meth-ods used in the structural VAR literature can then be applied in this context. An extensive set of simulation experiments provides support for the theoretical results. Finally, we apply the new structural DFM to identify the driving forces of the US economy using data for about 150 macroeconomic variables.
... Early applications of dynamic factor models are described in Sargent and Sims (1977) and Geweke (1977). Recent examples are Watson (1989, 2002), Camba-Mendez, Kapetanios, Smith and Weale (2001), and the Generalized Dynamic Factor Model of Forni, Hallin, Lippi and Reichlin (2000) and Forni and Lippi (2001). ...
... The literature on forecasting GDP using these variables is extensive. For example, Comba-Mendez et al. (2001) use short-term interest rates, real effective exchange rates, and stock price indices etc. to forecast GDP of selected European countries. Fagan et al. (2001) and Dreger and Marcellino (2003) construct medium-scale macroeconometric models for the Euro-area economic variables such as private consumption, fixed capital formation, nominal exchange rate, domestic and foreign nominal interest rate, real interest rate, etc. Banerjee et al. (2005) use not only Euro-area series but also US macroeconomic variables to conduct a detailed evaluation of the properties of a large set of leading indicators for predicting Euroarea GDP. ...
This study provides the first attempt to examine the ability of the price of fine wine to forecast the Gross Domestic Product (GDP) for the major developed countries. Considering the limitation of a linear Granger causality test in detecting nonlinear causal relationships, a nonlinear Granger causality test is also employed. The results from our nonlinear causality test show that this new variable contains useful information to forecast GDP for the US, the UK, and Australia, suggesting that we may include it as a forecasting variable in GDP forecasting models, especially nonlinear models, for these three countries.
... Since the seminal work of Stock and Watson, the single-index model, also called the dynamic factor model, has been widely used by many other researchers. The recent literature includes Camba-Mendez (2001) and Garcia-Ferrer and Poncela (2002), who modify the model to forecast GDP growth for European countries. For Germany in particular, Bandholz and Funke (2003) use the model to develop leading/ coincident indicators of economic activity in the country. ...
This paper applies the single-index dynamic factor model developed by Stock and Watson (1991) to construct current-quarter estimates of economic activity in Hong Kong. The Hang Seng index, a residential property price index, retail sales and total exports are used as coincident indicators. Principal Component Analysis is first used to obtain an impression of the common component of the indicator series. This component and the dynamic factor identified by the Stock-Watson methodology are strongly correlated and seem to account for economic fluctuations in Hong Kong reasonably well.
... 18 For a relatively similar treatment of dynamics at the level of factor extraction but in the domain of frequencies, see Forni et al. [2002]. 19 See for example: Watson [1998, 1999], Camba-Mendez et al. [1999], Angelini et al. [2001a]. 20 Let us recall that Eurostat GDP flash estimate is released about 45 days after the end of the quarter estimated, hence, the use of the term coincident for information available shortly before its release. ...
This paper introduces a statistical model for short-term GDP forecasting based on approximate dynamic factors, Stock and Watson methodology, extracted from a very large number of leading indicators at several lags. Given that factor extraction is performed on many series from all countries of the euro area, the common component to all predictors reflects the overall business cycle of the euro area and can accordingly provide a good proxy for euro area GDP.
... Camba-Mendez, Kapetanios, Smith, and Weale (2001) propose an automatic leading indicator approach (ALI) based on dynamic factor models. In a twostep approach, they show for Germany and other European countries that the on average ALI forecasts better than traditional VAR and BVAR models with traditional leading indicators. ...
... A more recent development in the construction of BCIs is the use of dynamic factor models. Earlier applications of dynamic factor models are Sargent and Sims (1977) and Geweke (1977) and recent examples are Watson (1989, 2002), Camba-Mendez, Kapetanios, Smith and Weale (2001) . The second is EuroCOIN of Altissimo et al. (2001), an indicator constructed using the GDFM and the third is our own indicator EuroIJR in which the GDFM is applied to the (limited) set of components of the TCB coincident and the leading indicators for the three Euro area countries. ...
Business cycle indicators are used to get a timely and frequent description of the state of the economy and its likely development in the near future. This paper discusses two methods for constructing business cycle indicators, the traditional NBER method and a recently developed dynamic factor model, and compares these methods for the Euro area business cycle. The results suggest that a reliable indicator can be constructed from a limited number of series that are selected using economic logic.
This paper constructs a coincident indicator for the United Kingdom employing—for the first time—a novel set of “faster” indicators recently published by the Office for National Statistics. Most of these variables were first released during the COVID‐19 pandemic with the aim of facilitating economics and social research by more timely indicators. The empirical evidence suggests that a coincident indicator based on a novel weekly dataset successfully captures the economic conditions in real time. We further use this indicator in an out‐of‐sample macroeconomic nowcasting exercise targeting monthly economic growth, prices and retail sales.
Ekonomik aktivitede zaman içerisinde
yaşanan inişli çıkışlı hareketler iktisadi
dalgalanmalar olarak tanımlanmakta iken çıktı
düzeyindeki dalgalanmaların ileriki dönemlerdeki
seyri hakkında bilgiler içeren göstergelere
ekonomik göstergeler denilmektedir. İktisadi
dalgalanmaların kısa dönemli öngörüsü, yaklaşan
resesyonların tespiti ve buna yönelik gerekli
ö n l e m l e r i n a l ı n m a s ı a ç ı s ı n d a n ö n e m l i
görülmektedir. Bu çalışmada, 2002Q1-2016Q3
dönemleri arası aylık ve üç aylık toplam 29
göstergeden yararlanılarak GSYH tahmin modeli
oluşturma denemesi yapılmıştır. Bu amaçla, biri
bağımlı değişken olarak GSYH, diğeri bağımsız
değişken olarak bir ekonomik gösterge olmak üzere
iki değişkenli VAR tahmin modelleri tahmin
edilmiştir. Sonra, elde edilen gösterge bazlı bireysel
tahminlerden aritmetik ortalama ve medyan alma
yöntemleriyle tahmin birleştirme teknikleri
uygulanmıştır. Son olarak, elde edilen tahminlerin
ortalama hataları referans bir AR tahmininden elde
edilen hataların ortalamasıyla karşılaştırılarak
tahmin performanslarına bakılmıştır. Sonuç olarak,
BIST100 endeksi, döviz kuru ve tüketici fiyat
endeksi gibi verilerin, GSYH'nin kısa dönemli
konjonktürel hareketlerinin tahmininde yararlı
bilgiler içerdiği görülmüştür. Ayrıca, tahmin
birleştirme yöntemlerinin pek çok göstergeden elde
edilen bireysel tahminlerden daha başarılı sonuçlar
verdiği de tespit edilmiştir.
As India steps in the fourth decade of globalization, we can look at the fruits of globalization more objectively. The relationship of globalization with growth is much more complex with economies opting for protectionism. The aim of the paper is to identify the pace of globalization in the Indian economy by capturing the portion of economic growth explained by foreign factors. The factor‐induced domestic and foreign factor loadings are used in the time‐varying estimations to suggest the evidence of slowbalization in India. The models supplemented with stochastic volatility help in examining the structural changes in the economy.
The link between globalization and economic growth is getting complex as the propagators of globalization are opting protectionism. The paper attempts to identify the time-varying dimension of globalization in India from Q2 1996 to Q3 2019. The aim is to capture the portion of growth explained by domestic and foreign factors suggesting the pace of globalization. The factor-induced domestic and foreign loadings are used in Time-Varying Parameter Regression and Time-Varying Parameter Autoregressive approach to indicate the evidence for slowbalization in India. The models combined with stochastic volatility, aid in capturing the structural changes in the economy in a robust manner.
This paper investigates the volatility linkage between energy and agricultural futures returns and how this linkage responds to external macroeconomic shocks. A framework combining the VARMA-BEKK-GARCH model and the Permanent-Transitory decomposition technology is employed to detect the volatility transmission, to decompose the volatility linkage into permanent and transitory components, and to examine the underlying determinants of the transitory volatility linkage. We have the following findings. A bidirectional volatility linkage between energy and agricultural futures returns exists and becomes more pronounced in recent years. The bidirectional linkage results from the co-movement effect induced by external shocks rather than from the substitution effect induced by the biofuel industry, and is not weakened by the shale gas revolution. Serving as proxies for external shocks from the world economy, trade, and financial markets, the CRB, BDI, and USDX indices provide strong explanatory power for the transitory volatility linkage, and the futures of these indices can be used to effectively and inexpensively hedge against the risks of the portfolios involving energy and agricultural futures.
We consider the problem of macroeconomic forecasting for China. Our objective is to determine whether well-established forecasting models that are commonly used to compute forecasts for Western macroeconomies are also useful for China. Our study includes 19 different forecasting models, ranging from simple approaches such as the naive forecast to more sophisticated techniques such as ARMA, Bayesian VAR, and factor models. We use these models to forecast two different measures of price inflation and two different measures of real activity, with forecast horizons ranging from 1 to 12 months, over a period that stretches from March 2005 to December 2018. We test null hypotheses of equal mean squared forecasting error between each candidate model and a simple benchmark. We find evidence that AR, ARMA, VAR, and Bayesian VAR models provide superior 1-month-ahead forecasts of the producer price index when compared to simple benchmarks, but find no evidence of superiority over simple benchmarks at longer horizons, or for any of our other variables.
Forecasting GDP growth is essential for effective and timely implementation of macroeconomic policies. This paper uses a principal component augmented Time Varying Parameter Regression (TVPR) approach to forecast real aggregate and sectoral growth rates for India. We estimate the model using a mix of fiscal, monetary, trade and production side-specific variables. To assess the importance of different growth drivers, three variants of the model are tried, namely, Demand-side, Supply-side and Combined models. We also find that TVPR model consistently outperforms constant parameter principal component augmented regression model and Dynamic Factor Model in terms of forecasting performance for all the three specifications.
Statistical agencies release their preliminary estimates of quarterly GDP growth with a publication delay that ranges from four to eight weeks. Given this lack of timeliness, nowcasting methods have been developed to produce early estimates of GDP growth during the ongoing quarter. As a practical illustration of these methods, I apply several small-scale nowcasting models, including a dynamic factor model, to produce estimates of Russian GDP growth for the first quarter of 2016. I then compare the nowcasting performance of the dynamic factor model against naïve AR- and ADLmodels using pseudo out-of-sample forecasting errors. The results indicate Russia’s GDP contraction slowed in the first quarter of 2016. The dynamic factor model outperforms the naïve models,
displaying better nowcasting prediction accuracy for Russian GDP.
Business tendency surveys (BTS) carried out by the statistical institute INSEE are intensively used for the short-term forecasting of the French economic activity. In particular, the service BTS has been used together with the industry BTS for the short-term forecasting of GDP growth since Bouton and Erkel-Rousse (2003-2004) showed that the former survey contained a specific piece of information on GDP growth with respect to the latter survey. However, it remained to be demonstrated that this specific piece of information permits one to significantly improve the quality of short-term GDP forecasts with respect to models involving variables from the industry survey exclusively. More generally, the predictive accuracy of models based on the two surveys with respect to simpler autoregressive (AR) models deserved to be assessed.
We, therefore, perform a real-time out-of-sample analysis that consists in estimating and, then, simulating miscellaneous kinds of models (VAR and univariate multistep models) aimed at the short-term forecasting of the quarterly GDP growth rate. Some BTS-based models encompass industry and service data, others either service or industry data exclusively. The predictive accuracy of these three kinds of models is compared to that of simple AR models; that of models including service and industry survey data is also compared to that of models based on data from one of the two surveys exclusively. Predictive accuracy tests (Harvey, Leybourne and Newbold, 1997, Clark-West, 2007) are performed up to four-quarter-forecast horizons. To assess the robustness of the results, we carry out both recursive and rolling estimations as well as three
tests (differing by the method used to estimate the variance of the test statistics’ numerators) for each couple of competing forecasts. The results establish the usefulness of the two BTS, as well as the limited but significant contribution of the service survey in addition to the industry survey especially in the months when long enough service series are available (namely January, April, July, and October). The industry survey, nonetheless, appears to predominate over the service survey as a source of leading indicators of GDP growth.
This paper uses a forecasting model for real economic activity for a group of emerging economies (Brazil, India, Mexico and Russia) based on the information contained in their capital markets. We forecast the industrial production in emerging markets throughout different time horizons using information contained in stock and fixed-income markets. Our results suggest that fixed-income and stock markets do not reveal information regarding future economic growth in Brazil, Mexico and Russia. In the case of India, the yield spread explain part of the variation of the economic activity, but the stock market does not have predictive power.
Celem artykułu jest zbadanie, jaki wpływ na jakość krótkoterminowych prognoz produkcji przemysłowej ma zastosowanie w modelu wskaźników cyklu koniunkturalnego. Badanie przeprowadzone zostało na gospodarce polskiej przy użyciu metodologii zaproponowanej przez M. Kleina.
Many scientists have analyzed a relation among economic and financial market variables. One of the purposes of these researches, and probably the most important one, is to find a better prediction of future economic changes. Economic theory suggests a strong link between stock market and economic activity. The question is whether the stock market is a predictor of future economic activity measured as growth of country’s gross domestic product. In this paper we analyze which economic sectors, represented by stock market sector indices, could have most impact on GDP. The aim of this paper is to analyze whether some sectors are more important than others while analyzing GDP change. The study uses data for the period 2000 Q2 – 2012 Q1 of the U.S. seasonally adjusted GDP and Dow Jones indices and data for the period 2001 Q1 – 2012 Q1 of France seasonally adjusted GDP and Euronext CAC indices. In order to find a relation between selected stock market sector indices and GDP, we will use cross correlation analysis. Our findings support the theory that stock market is a leading indicator for economic growth. In France stock market appears to be a stronger indicator for economic growth compared to the U.S. The results revealed that seasonally adjusted GDP growth lagged behind changes in stock market indices four quarters in France and three quarter in the U.S. In the U.S. worst predictive capabilities for GDP growth come from utilities, oil and gas sectors. Industrial and financial sectors gave the best cross correlation results with GDP growth. In France telecommunication, utilities sectors have the worst predictive capabilities, and consumer services and health care sectors gave the best cross correlation results on GDP growth, compared with others. © 2014, Kauno Technologijos Universitetas. All rights reserved.
In this paper, we evaluate the role of a set of variables as leading indicators for Euro-area inflation and GDP growth. Our leading indicators are taken from the variables in the European Central Bank's (ECB) Euro-area-wide model database, plus a set of similar variables for the US. We compare the forecasting performance of each indicator ex post with that of purely autoregressive models. We also analyse three different approaches to combining the information from several indicators. First, ex post, we discuss the use as indicators of the estimated factors from a dynamic factor model for all the indicators. Secondly, within an ex ante framework, an automated model selection procedure is applied to models with a large set of indicators. No future information is used, future values of the regressors are forecast, and the choice of the indicators is based on their past forecasting records. Finally, we consider the forecasting performance of groups of indicators and factors and methods of pooling the ex ante single-indicator or factor-based forecasts. Some sensitivity analyses are also undertaken for different forecasting horizons and weighting schemes of forecasts to assess the robustness of the results.
In this paper we perform and evaluate, in sample, some methodologies for building of coincident indicators focusing on the detection of business cycle of the Industrial activity. Specifically, we try a version of coincident indicator based on the coincident indicator for economy of the TCB (The Conference Board), a modified version in which the volatility is modeled, a Stock-Watson approach and a Mariano-Murasawa approach. We conclude that, in general, the TCB index out-performs the others methods tested, being competitive just with the modified method. Furthermore, The TCB Traditional has the tendency to comprehend the recessives periods of the others, in correspondence with the known recessive periods for industrial activity.
Product diversity, which is highly important in economic systems, has been
highlighted by recent studies on international trade. We found an empirical
pattern, designated as the "S-shaped curve", that models the relationship
between economic size (logarithmic GDP) and export diversity (the number of
varieties of export products) on the detailed international trade data. As the
economic size of a country begins to increase, its export diversity initially
increases in an exponential manner, but overtime, this diversity growth slows
and eventually reaches an upper limit. The interdependence between size and
diversity takes the shape of an S-shaped curve that an be fitted by a logistic
equation. To explain this phenomenon, we introduce a new parameter called
"substitutability" into the list of capabilities or factors of products in the
tri-partite network model (i.e., the country-capability-product model) of
Hidalgo et al. As we observe, when the substitutability is zero, the model
returns to Hidalgo's original model but failed to reproduce the S-shaped curve.
However, in a plot of data, the data increasingly resembles an the S-shaped
curve as the substitutability expands. Therefore, the diversity ceiling effect
can be explained by the substitutability of different capabilities.
The present paper compares expected inflation to (econometric) inflation forecasts based on a number of forecasting techniques from the literature using a panel of ten industrialized countries during the period from 1988 to 2007. To capture expected inflation, we develop a recursive filtering algorithm that extracts unexpected inflation from real interest rate data, even in the presence of diverse risks and a potential Mundell–Tobin-effect.The extracted unexpected inflation is compared to the forecasting errors of ten econometric forecasts. In addition to the standard AR(p) and ARMA(1,1) models, which are known to have the best performance on average, we also employ several Phillips curve-based approaches, VAR, dynamic factor models and two simple model averaging approaches.Finally, we show that the quality of the expectations clearly matches the quality of the forecasts derived with the techniques that are currently proposed for this purpose in the economic literature.Highlights► We develop a new panel filter to identify inflation expectations. ► We compare expectations to a set of econometric models of inflation. ► Expectations perform well in terms of precision compared to all benchmark models. ► Expectations include information that is not included in most forecasting models. ► However, expectations do not exploit all available information.
We consider games with incomplete information a la Harsanyi, where the payoff of a player depends on an unknown state of nature as well as on the profile of chosen actions. As opposed to the standard model, players' preferences over state--contingent utility vectors are represented by arbitrary functionals. The definitions of Nash and Bayes equilibria naturally extend to this generalized setting. We characterize equilibrium existence in terms of the preferences of the participating players. It turns out that, given continuity and monotonicity of the preferences, equilibrium exists in every game if and only if all players are averse to uncertainty (i.e., all the functionals are quasi--concave). We further show that if the functionals are either homogeneous or translation invariant then equilibrium existence is equivalent to concavity of the functionals.
i) A review has been made of spectral analysis and its relation with other branches of time‐series analysis. A detailed account has been given of the methods available for estimating the spectral density, the band spectrum and the integrated spectrum; suggestions have also been made for confidence intervals for the latter two estimators.
(ii) Emphasis has been laid throughout upon the difficulties which are met in practice and gaps in the theoretical structure have been indicated. In applications to physical problems, it is suggested that an empirical statistical approach is not enough by itself and that more realistic descriptions of each particular phenomenon should be attempted.
(iii) Reference has already been made in Section 9 to a sampling investigation in which it is proposed to apply the techniques mentioned in this paper to a large number of artificially constructed series of the type given by (1).
The paper formulates and estimates a single-factor multivariate time series model. The model is a dynamic generalization of the multiple indicator (or factor analysis) model. It is shown to be a special case of the general state space model and can be estimated by maximum likelihood methods using the Kalman filter algorithm. The model is used to obtain estimates of the unobserved metropolitan wage rate for Los Angeles, based on observations of sectoral wages within the Standard Metropolitan Statistical Area. Hypothesis tests, model diagnostics, and out-of-sample forecasts are used to evaluate the model.
This paper extends previous work on the information in the term structure about future real economic growth. For the U.S. and Germany, and to a lesser extent for the U.K., we find evidence that the long end of the term structure has information about future growth of industrial production beyond expectations about future monetary policy. We also find evidence that foreign term structures can forecast domestic low frequency movements in economic activity especially in countries that experience high and variable rates of inflation.
This paper presents analytical, empirical and simulation results concerning inference about the moments of nondifferentiable functions of out-of-sample forecasts and forecast errors. Special attention is given to the measurement of a model's predictive ability using the test of equal mean absolute error. Tests for equal mean absolute error and mean square error are used to evaluate predictions of excess returns to the S & P 500 composite. Simulations indicate that appropriately constructed tests for equal mean absolute error can provide more accurately sized and more powerful tests than inappropriately constructed tests for equal mean absolute error and mean square error.
Given two sources of forecasts of the same quantity, it is possible to compare prediction records. In particular, it can be useful to test the hypothesis of equal accuracy in forecast performance. We analyse the behaviour of two possible tests, and of modifications of these tests designed to circumvent shortcomings in the original formulations. As a result of this analysis, a recommendation for one particular testing approach is made for practical applications.
The problem of selecting one of a number of models of different dimensions is treated by finding its Bayes solution, and evaluating the leading terms of its asymptotic expansion. These terms are a valid large-sample criterion beyond the Bayesian context, since they do not depend on the a priori distribution.
This paper examines the forecasting performance of various leading economic indicators and composite indexes since 1988. in particular during the onset of the 1990 recession. The primary focus is on an experimental recession index (tile "XRI"). a composite index which provides probabilistic forecasts of whether the U.S. economy will be in a recession six months hence. After detailing its construction, the paper examines the out-of-sample performance of the XRI and a related forecast of overall economic growth. the experimental leading index (XLI). These indexes performed well from 1988 through the summer of 1990 - for example. in June 1990 the XLI model forecasted a .4% (annual rate) decline in the experimental coincident index from June through September. when in fact the decline was only slightly greater,.8%. However. the XLI failed to forecast the sharp declines of October and November 1990. After exploring several possible explanations. we conclude that one important source of the forecast error was the use of financial variables during a recession that was not associated with a particularly tight monetary policy. Financial indicators -- and the experimental index -- were not alone. however. in failing to forecast the 1990 recession, An examination of 45 economic indicators shows that almost all failed to forecast the 1990downturn. and the few that did provided unclear signals before the recessions of the 19705 and 1980s.
The system of Leading and Coincident Economic Indicators, currently maintained by the U.S. Department of Commerce (DOC), was developed as part of the NBER research program on business cycles over fifty years ago. This paper uses recent developments in econometric methodology and computing technology to take a fresh look at this system. The result is three experimental indexes. The first, constructed using a dynamic factor model, is numerically similar to the current index of coincident indicators maintained by the DOC. The second, an alternative index of leading indicators, is designed to forecast the growth in the DOC index over a six month horizon. The third-a "Recession Index"-estimates the probability that the economy will be in a recession six months hence. Only two of the seven series in the proposed leading index are used by the DOC to construct their index. Of these new series, interest rates (a public-private risk premium and the slope of the yield curve) are found to be particularly useful predictors of future economic activity.
The stylized facts of macroeconomic time series can be presented by fitting structural time series models. Within this framework, we analyze the consequences of the widely used detrending technique popularized by Hodrick and Prescott (1980). It is shown that mechanical detrending based on the Hodrick-Prescott filter can lead investigators to report spurious cyclical behavior, and this point is illustrated with empirical examples. Structural time-series models also allow investigators to deal explicitly with seasonal and irregular movements that may distort estimated cyclical components. Finally, the structural framework provides a basis for exposing the limitations of ARIMA methodology and models based on a deterministic trend with a single break. Copyright 1993 by John Wiley & Sons, Ltd.
This paper seeks to address the policy issue of the usefulness of financial spreads as indicators of future inflation and output growth in the countries of the European Union, placing a particular focus on out-of-sample forecasting performance. Such analysis is of considerable relevance to monetary authorities, given the breakdown of the money/income relation in a number of countries and following increased emphasis of domestic monetary policy on control of inflation following the broadening of the ERM bands. The results confirm that for some countries, financial spread variables do contain some information about future output growth and inflation, with the yield curve and the reverse yield gap performing best. However, the relatively poor out-of-sample forecasting performance and/or parameter instability suggests that the need for caution in using spread variables for forecasting in EU countries. Only a small number of spreads contain information, and improve forecasting in a manner which is stable over time.
Does co-integration help long-term forecasts? In this paper, we use simulation, real data sets, and multi-step-ahead post-sample forecasts to study this question. Based on the square root of the trace of forecasting error-covariance matrix, we found that for simulated data imposing the 'correct' unit-root constraints implied by co-integration does improve the accuracy of forecasts. For real data sets, the answer is mixed. Imposing unit-root constraints suggested by co-integration tests produces better forecasts for some cases, but fares poorly for others. We give some explanations for the poor performance of co-integration in long-term forecasting and discuss the practical implications of the study. Finally, an adaptive forecasting procedure is found to perform well in one- to ten-step-ahead forecasts. Copyright 1996 by John Wiley & Sons, Ltd.
The paper examines the performance of four multivariate volatility models, namely CCC, VARMA-GARCH, DCC and BEKK, for the crude oil spot and futures returns of two major benchmark international crude oil markets, Brent and WTI, to calculate optimal portfolio weights and optimal hedge ratios, and to suggest a crude oil hedge strategy. The empirical results show that the optimal portfolio weights of all multivariate volatility models for Brent suggest holding futures in larger proportions than spot. For WTI, however, DCC and BEKK suggest holding crude oil futures to spot, but CCC and VARMA-GARCH suggest holding crude oil spot to futures. In addition, the calculated optimal hedge ratios (OHRs) from each multivariate conditional volatility model give the time-varying hedge ratios, and recommend to short in crude oil futures with a high proportion of one dollar long in crude oil spot. Finally, the hedging effectiveness indicates that DCC (BEKK) is the best (worst) model for OHR calculation in terms of reducing the variance of the portfolio.
A number of interest rates and interest rate spreads have been found to be useful in prediction the course of the economy. We compare the predictive power of some of these suggested interest rate variables for nine indicators of real activity and the inflation rate. Our results are consistent with those of Stock and Watson (1989) and Friedman and Kuttner (1989), who found that the spread between the commercial paper rate and the Treasury bill rate has been a particularly good predictor. We present evidence that this spread is informative not so much because it is a measure of default risk (which has been the usual presumption), but because it is an indicator of the stance of monetary policy; for example, during the "credit crunches" of the l960s aid the 1970s, the commercial paper -- Treasury bill spread typically rose significantly. We also show that, possibly because of charges in monetary policy operating procedures aid in financial markets, this spread appears r to be a less reliable predictor than it used to be.
A positive slope of the yield curve is associated with a future increase in real economic activity: consumption (nondurables plus services), consumer durables, and investment. It has extra predictive power over the index of leading indicators, real short-term interest rates, lagged growth in economic activity, and lagged rates of inflation. It outperforms survey forecasts, both in-sample and out-of-sample. Historically, the information in the slope reflected, inter alia, factors that were independent of monetary policy and, thus, the slope could have provided useful information both to private investors and to policymakers. Copyright 1991 by American Finance Association.
The authors show that the CUSUM test of the stability over time of the coefficients of a linear regression model, which is usually based on recursive residuals, can also be applied to ordinary least squares residuals. The authors derive the limiting null distribution of the resulting test and compare its local power to that of the standard procedure. It turns out that neither version is uniformly superior to the other. Copyright 1992 by The Econometric Society.
- W Ploberger
- W Kramer
Ploberger, W., and W. Kramer 1992: The CUSUM Test with OLS Residuals,"
Econometrica, 60, 2711285.
- A C Harvey
Harvey, A. C. 1993: Time Series Models. Harvester and Wheatsheaf.
- W Ploberger
- W Kramer
Ploberger, W., and W. Kramer 1992: The CUSUM Test with OLS Residuals,"
Econometrica, 60, 271 285.
NBER Macroeconomics annual, 4, 351 394. 1991: A Probability Model of the Coincident Economic Indicators
- J H Stock
- M Watson
Stock, J. H., and M. Watson 1989: New Indexes Sic of Coincident and Leading
Indicators," NBER Macroeconomics annual, 4, 351 394.
1991: A Probability Model of the Coincident Economic Indicators," in Leading
Economic Indicators, ed. by G. Moore, and K. Lahiri. Cambridge University Press.
1993: A Procedure for Predicting Recessions with Leading Indicators: Econometric Issues and Recent Experience," in Business Cycles, Indicators and Forecasting, ed.
by J. H. Stock, and M. Watson. University of Chicago Press.