Minnesota-type adaptive hierarchical priors for large Bayesian VARs

Large Bayesian VARs with stochastic volatility are increasingly used in empirical macroeconomics. The key to making these highly parameterized VARs useful is the use of shrinkage priors. We develop a family of priors that captures the best features of two prominent classes of shrinkage priors: adaptive hierarchical priors and Minnesota priors. Like adaptive hierarchical priors, these new priors ensure that only ‘small’ coefficients are strongly shrunk to zero, while ‘large’ coefficients remain intact. At the same time, these new priors can also incorporate many useful features of the Minnesota priors such as cross-variable shrinkage and shrinking coefficients on higher lags more aggressively. We introduce a fast posterior sampler to estimate BVARs with this family of priors—for a BVAR with 25 variables and 4 lags, obtaining 10,000 posterior draws takes about 3 min on a standard desktop computer. In a forecasting exercise, we show that these new priors outperform both adaptive hierarchical priors and Minnesota priors.     

Macroeconomic forecasting with large Bayesian VARs: Global-local priors and the illusion of sparsity

A class of global-local hierarchical shrinkage priors for estimating large Bayesian vector autoregressions (BVARs) has recently been proposed. We question whether three such priors: Dirichlet-Laplace, Horseshoe, and Normal-Gamma, can systematically improve the forecast accuracy of two commonly used benchmarks (the hierarchical Minnesota prior and the stochastic search variable selection (SSVS) prior), when predicting key macroeconomic variables. Using small and large data sets, both point and density forecasts suggest that the answer is no. Instead, our results indicate that a hierarchical Minnesota prior remains a solid practical choice when forecasting macroeconomic variables. In light of existing optimality results, a possible explanation for our finding is that macroeconomic data is not sparse, but instead dense.     

Composite likelihood methods for large Bayesian VARs with stochastic volatility

Adding multivariate stochastic volatility of a flexible form to large vector autoregressions (VARs) involving over 100 variables has proved challenging owing to computational considerations and overparametrization concerns. The existing literature works with either homoskedastic models or smaller models with restrictive forms for the stochastic volatility. In this paper, we develop composite likelihood methods for large VARs with multivariate stochastic volatility. These involve estimating large numbers of parsimonious models and then taking a weighted average across these models. We discuss various schemes for choosing the weights. In our empirical work involving VARs of up to 196 variables, we show that composite likelihood methods forecast much better than the most popular large VAR approach, which is computationally practical in very high dimensions: the homoskedastic VAR with Minnesota prior. We also compare our methods to various popular approaches that allow for stochastic volatility using medium and small VARs involving up to 20 variables. We find our methods to forecast appreciably better than these as well.     

Bayesian model comparison for time‐varying parameter VARs with stochastic volatility

We develop importance sampling methods for computing two popular Bayesian model comparison criteria, namely, the marginal likelihood and the deviance information criterion (DIC) for time‐varying parameter vector autoregressions (TVP‐VARs), where both the regression coefficients and volatilities are drifting over time. The proposed estimators are based on the integrated likelihood, which are substantially more reliable than alternatives. Using US data, we find overwhelming support for the TVP‐VAR with stochastic volatility compared to a conventional constant coefficients VAR with homoskedastic innovations. Most of the gains, however, appear to have come from allowing for stochastic volatility rather than time variation in the VAR coefficients or contemporaneous relationships. Indeed, according to both criteria, a constant coefficients VAR with stochastic volatility outperforms the more general model with time‐varying parameters.     

Infinite-dimensional VARs and factor models

This paper proposes a novel approach for dealing with the ‘curse of dimensionality’ in the case of infinite-dimensional vector autoregressive (IVAR) models. It is assumed that each unit or variable in the IVAR is related to a small number of neighbors and a large number of non-neighbors. The neighborhood effects are fixed and do not change with the number of units (N), but the coefficients of non-neighboring units are restricted to vanish in the limit as N tends to infinity. Problems of estimation and inference in a stationary IVAR model with an unknown number of unobserved common factors are investigated. A cross-section augmented least-squares (CALS) estimator is proposed and its asymptotic distribution is derived. Satisfactory small-sample properties are documented by Monte Carlo experiments. An empirical illustration shows the statistical significance of dynamic spillover effects in modeling of US real house prices across the neighboring states.     

Valuation Accuracy and Infinity Horizon Forecast: Empirical Evidence from Europe

This paper focuses on the assumptions of infinite-horizon forecasting in the field of firm valuation. The estimate of long-run continuing values is based on the hypothesis that companies should have reached the steady state at the end of the period of explicit forecasts. It is argued that the equivalence between cash accounting and accrual accounting is the way of verifying the steady-state assumption, defined as the state when a firm earns exactly its cost of capital, i.e., what we would expect in pure-competition settings. From this definition, we derive that the "ideal" growth rate to use in steady state is equal to the reinvestment rate times Weighted Average Cost of Capital. To validate our approach, we collect a sample of 784 analyst valuations and compare how the implied target prices deviate from what the target prices would have been using the "ideal" steady-state growth rates. Using Logit and Cox regression models, we find that this deviation has predictive value over the probability that actual market price reaches the target price over the following 12-month period—the smaller the deviation the greater is the likelihood that the market price reaches the target price.     

Structural VARs and noninvertible macroeconomic models

We resume the line of research pioneered by C. A. Sims and Zha (Macroeconomic Dynamics, 2006, 10, 231–272) and make two novel contributions. First, we provide a formal treatment of partial fundamentalness—that is, the idea that a structural vector autoregression (VAR) can recover, either exactly or with good approximation, a single shock or a subset of shocks, even when the underlying model is nonfundamental. In particular, we extend the measure of partial fundamentalness proposed by Sims and Zha to the finite‐order case and study the implications of partial fundamentalness for impulse‐response and variance‐decomposition analysis. Second, we present an application where we validate a theory of news shocks and find it to be in line with the empirical evidence.     

企业需求预测准确性存在的问题及其应对策略

预测由来已久,它源自人们对未来各种美好事物的向往与追求。中国民间不少古谚,如：“运筹谋划,须先预测未来”、“多算胜,少算不胜,而况于不算”等都体现了预测的思想;而在世界历史上也有很多预测,如：“哥德巴赫猜想”、“玛雅预言”等对后世有关领域的研究和发展产生了很大的影响。随着全球市场一体化和顾客需求多样化的发展,市场竞争已经从地区的单一性向区域性、     

Forecasting economic and financial variables with global VARs

This paper considers the problem of forecasting economic and financial variables across a large number of countries in the global economy. To this end a global vector autoregressive (GVAR) model, previously estimated by Dees, di Mauro, Pesaran, and Smith (2007) and Dees, Holly, Pesaran, and Smith (2007) over the period 1979Q1–2003Q4, is used to generate out-of-sample forecasts one and four quarters ahead for real output, inflation, real equity prices, exchange rates and interest rates over the period 2004Q1–2005Q4. Forecasts are obtained for 134 variables from 26 regions, which are made up of 33 countries and cover about 90% of the world output. The forecasts are compared to typical benchmarks: univariate autoregressive and random walk models. Building on the forecast combination literature, the effects of model and estimation uncertainty on forecast outcomes are examined by pooling forecasts obtained from different GVAR models estimated over alternative sample periods. Given the size of the modelling problem, and the heterogeneity of the economies considered–industrialised, emerging, and less developed countries–as well as the very real likelihood of possibly multiple structural breaks, averaging forecasts across both models and windows makes a significant difference. Indeed, the double-averaged GVAR forecasts perform better than the benchmark competitors, especially for output, inflation and real equity prices.     

Strategy,Valuation, and Forecast Accuracy: Evidence from Italian Strategic Plan Disclosures

Using a sample of 264 strategic plan presentations by Milan Stock Exchange firms during 2001–2012, we present evidence of both a security price reaction and an increase in the accuracy of analysts’ earnings forecasts pursuant to plan disclosure. In the cross-section, the information content of the plan disclosures and the accuracy increase are incrementally associated with the extent of forward-looking narrative disclosures in the plan, after controlling for other disclosures within and outside the plan presentation and the fact that the firm has self-selected into the sample. Both quantitative and qualitative narrative disclosures are informative to investors and analysts. The results are driven by narrative disclosures about company strategy and action plans rather than about the business environment in which the company operates. Our study informs the current debate on the use of voluntary comprehensive, integrated, long-run-oriented strategic plan disclosure as a potential complement for disclosures such as quarterly earnings forecasts that have been described as an example of ‘short-termism’.     

Large time‐varying parameter VARs: A nonparametric approach

In this paper we introduce a nonparametric estimation method for a large Vector Autoregression (VAR) with time‐varying parameters. The estimators and their asymptotic distributions are available in closed form. This makes the method computationally efficient and capable of handling information sets as large as those typically handled by factor models and Factor Augmented VARs. When applied to the problem of forecasting key macroeconomic variables, the method outperforms constant parameter benchmarks and compares well with large (parametric) Bayesian VARs with time‐varying parameters. The tool can also be used for structural analysis. As an example, we study the time‐varying effects of oil price shocks on sectoral U.S. industrial output. According to our results, the increased role of global demand in shaping oil price fluctuations largely explains the diminished recessionary effects of global energy price increases.     

The Persistence and Forecast Accuracy of Earnings Components in the USA and Japan

Not all components of earnings are expected to provide similar information regarding future earnings. For example, basic financial statement analysis indicates that the persistence of ordinary income should be greater than the persistence of special, extraordinary, or discontinued operations. Because the market assigns higher multiples to earnings components that are more persistent, differentiating earnings components on the basis of relative persistence would appear to be useful. A focus on relative predictive value is consistent with research findings and user recommendations on separating earnings components that are persistent or permanent from those that are transitory or temporary. This paper examines the persistence and forecast accuracy of earnings components for retail and manufacturing companies listed in the world's two largest equity markets; the USA and Japan. We find the forecast accuracy of earnings in both the USA and Japan increases with greater disaggregation of earnings components. The results further indicate that the improvements in forecast accuracy due to earnings disaggregation are greater in the USA than in Japan. The greater emphasis and more detailed guidelines for reporting earnings components in the USA produce a better differentiation in the persistence of earnings components resulting in greater forecast improvements from earnings disaggregation.     

The Use of Forecast Accuracy Indicators to Improve Planning Quality: Insights from a Case Study

Abstract

Accounting studies have analyzed rolling forecasts and similar dynamic approaches to planning as a way to improve the quality of planning. We complement this research by investigating an alternative (complementary) way to improve planning quality, i.e. the use of forecast accuracy indicators as a results control mechanism. Our study particularly explores the practical challenges that might emerge when firms use a performance measure for forecast accuracy. We examine such challenges by means of an in-depth case study of a manufacturing firm that started to monitor sales forecast accuracy. Drawing from interviews, meeting observations and written documentation, we highlight two possible concerns with the use of forecast accuracy: concerns related to the limited degree of controllability of the performance measure and concerns with its goal congruence. We illustrate how organizational actors experienced these challenges and how they adapted their approach to forecast accuracy in response to them. Our empirical observations do not only shed light on the possibilities and challenges pertaining to the use of forecast accuracy as a performance measure; they also improve our understanding of how specific qualities of performance measures apply to ‘truth-inducing’ indicators, and how the particular organizational and market context can shape the quality of performance measures more generally.     

Volatility activity: Specification and estimation

The paper examines volatility activity and its asymmetry and undertakes further specification analysis of volatility models based on it. We develop new nonparametric statistics using high-frequency option-based VIX data to test for asymmetry in volatility jumps. We also develop methods for estimating and evaluating, using price data alone, a general encompassing model for volatility dynamics where volatility activity is unrestricted. The nonparametric application to VIX data, along with model estimation for S&P index returns, suggests that volatility moves are best captured by an infinite variation pure-jump martingale with a symmetric jump compensator around zero. The latter provides a parsimonious generalization of the jump-diffusions commonly used for volatility modeling.     

Identifying structural VARs from sparse narrative instruments: Dynamic effects of US macroprudential policies

We study identification in Bayesian proxy VARs for instruments that consist of sparse qualitative observations indicating the signs of shocks in specific periods. We propose the Fisher discriminant regression and a non-parametric sign concordance criterion as two alternative methods for achieving correct inference in this case. The former represents a minor deviation from a standard proxy VAR, whereas the non-parametric approach builds on set identification. Our application to US macroprudential policies finds persistent declines in credit volumes and house prices together with moderate declines in GDP and inflation and a widening of corporate bond spreads after a tightening of capital requirements or mortgage underwriting standards.     

Minimum Distance Estimation and Testing of DSGE Models from Structural VARs*

The aim of this paper is to complement the minimum distance estimation–structural vector autoregression approach when the weighting matrix is not optimal. In empirical studies, this choice is motivated by stochastic singularity or collinearity problems associated with the covariance matrix of impulse response functions. Consequently, the asymptotic distribution cannot be used to test the economic model's fit. To circumvent this difficulty, we propose a simple simulation method to construct critical values for the test statistics. An empirical application with US data illustrates the proposed method.