Search

With short-term interest rates at the zero lower bound, forward guidance has become a key tool for central bankers, and yet we know little about its effectiveness. This paper first empirically documents the impact of forward guidance announcements on a broad cross section of financial markets data and professional forecasts. We find that Federal Open Market Committee (FOMC) announcements containing forward guidance had heterogeneous effects depending on the other content of the statement. We show that once we control for these other elements, forward guidance had, on average, positive and meaningful effects on output and inflation expectations. Using this benchmark, we then show that standard medium-scale DSGE models tend to grossly overestimate the impact of forward guidance on the macroeconomy, a phenomenon we call the “forward guidance puzzle.” We explain why this is the case and show that incorporating a perpetual youth structure into the benchmark provides a possible resolution to the puzzle.

This paper presents a framework for quantifying uncertainty around point forecasts for GDP, inflation and house prices in Norway. The framework combines quantile regressions using a broad set of uncertainty indicators with a skewed t-distribution, allowing for time-variation and asymmetry in the uncertainty forecasts. This approach helps provide deeper insights into the macroeconomic uncertainty surrounding forecasts than more traditional time-series models, where uncertainty is usually symmetric and with limited time-variation. Formal tests, such as the log score and the Continuous Ranked Probability Score (CRPS), show that using informative indicators tend to improve density forecasts, particularity in the medium run.

This paper introduces the concept of Temporal Hierarchies for time series forecasting. A temporal hierarchy can be constructed for any time series by means of non-overlapping temporal aggregation. Predictions constructed at all aggregation levels are combined with the proposed framework to result in temporally reconciled, accurate and robust forecasts. The implied combination mitigates modelling uncertainty, while the reconciled nature of the forecasts results in a unified prediction that supports aligned decisions at different planning horizons: from short-term operational up to long-term strategic planning. The proposed methodology is independent of forecasting models. It can embed high level managerial forecasts that incorporate complex and unstructured information with lower level statistical forecasts. Our results show that forecasting with temporal hierarchies increases accuracy over conventional forecasting, particularly under increased modelling uncertainty. We discuss organisational implications of the temporally reconciled forecasts using a case study of Accident & Emergency departments.

We propose a novel copula approach to producing density forecasts of economic aggregates combining models using disaggregate data. Our copula approach is more flexible compared to existing techniques, because it is applicable to any econometric model that produces density forecasts. We construct a set of Monte Carlo studies to investigate the properties of the suggested approach. In our empirical application, we use the Norwegian index for goods consumption (VKI) and the Norwegian consumer price index for underlying inflation (CPI-ATE). We find that the copula approach compares well to alternative methods using recursive out-of-sample estimation.

We evaluate the performances of various methods for forecasting tourism data. The data used include 366 monthly series, 427 quarterly series and 518 annual series, all supplied to us by either tourism bodies or academics who had used them in previous tourism forecasting studies. The forecasting methods implemented in the competition are univariate and multivariate time series approaches, and econometric models. This forecasting competition differs from previous competitions in several ways: (i) we concentrate on tourism data only; (ii) we include approaches with explanatory variables; (iii) we evaluate the forecast interval coverage as well as the point forecast accuracy; (iv) we observe the effect of temporal aggregation on the forecasting accuracy; and (v) we consider the mean absolute scaled error as an alternative forecasting accuracy measure. We find that pure time series approaches provide more accurate forecasts for tourism data than models with explanatory variables. For seasonal data we implement three fully automated pure time series algorithms that generate accurate point forecasts, and two of these also produce forecast coverage probabilities which are satisfactorily close to the nominal rates. For annual data we find that Naive forecasts are hard to beat.

This paper proposes two new weighting schemes that average forecasts using different estimation windows to account for structural change. We let the weights reflect the probability of each time point being the most-recent break point, and we use the reversed ordered Cusum test statistics to capture this intuition. The second weighting method simply imposes heavier weights on those forecasts that use more recent information. The proposed combination forecasts are evaluated using Monte Carlo techniques, and we compare them with forecasts based on other methods that try to account for structural change, including average forecasts weighted by past forecasting performance and techniques that first estimate a break point and then forecast using the post break data. Simulation results show that our proposed weighting methods often outperform the others in the presence of structural breaks. An empirical application based on a NAIRU Phillips curve model for the United States indicates that it is possible to outperform the random walk forecasting model when we employ forecasting methods that account for break uncertainty.

Optimal probabilistic forecasts of integer-valued random variables are derived. The optimality is achieved by estimating the forecast distribution nonparametrically over a given broad model class and proving asymptotic efficiency in that setting. The ideas are demonstrated within the context of the integer autoregressive class of models, which is a suitable class for any count data that can be interpreted as a queue, stock, birth and death process or branching process. The theoretical proofs of asymptotic optimality are supplemented by simulation results which demonstrate the overall superiority of the nonparametric method relative to a misspecified parametric maximum likelihood estimator, in large but .nite samples. The method is applied to counts of wage claim benefits, stock market iceberg orders and civilian deaths in Iraq, with bootstrap methods used to quantify sampling variation in the estimated forecast distributions.

We describe some fast algorithms for reconciling large collections of time series forecasts with aggregation constraints. The constraints arise due to the need for forecasts of collections of time series with hierarchical or grouped structures to add up in the same manner as the observed time series. We show that the least squares approach to reconciling hierarchical forecasts can be extended to more general non-hierarchical groups of time series, and that the computations can be handled efficiently by exploiting the structure of the associated design matrix. Our algorithms will reconcile hierarchical forecasts with hierarchies of unlimited size, making forecast reconciliation feasible in business applications involving very large numbers of time series.

Damped trend exponential smoothing has previously been established as an important forecasting method. Here, it is shown to have close links to simple exponential smoothing with a smoothed error tracking signal. A special case of damped trend exponential smoothing emerges from our analysis, one that is more parsimonious because it effectively relies on one less parameter. This special case is compared with its traditional counterpart in an application to the annual data from the M3 competition and is shown to be quite competitive.

This paper argues that VAR models with cointegration and common cycles can be usefully viewed as observable factor models. The factors are linear combinations of lagged levels and lagged differences, and as such, these observable factors have potential for forecasting. We illustrate this forecast potential in both a Monte Carlo and empirical setting, and demonstrate the difficulties in developing forecasting "rules of thumb" for forecasting in multivariate systems.

This article studies a simple, coherent approach for identifying and estimating error correcting vector autoregressive moving average (EC-VARMA) models. Canonical correlation analysis is implemented for both determining the cointegrating rank, using a strongly consistent method, and identifying the short-run VARMA dynamics, using the scalar component methodology. Finite sample performances are evaluated via Monte-Carlo simulations and the approach is applied to model and forecast US interest rates. The results reveal that EC-VARMA models generate significantly more accurate out-of-sample forecasts than vector error correction models (VECMs), especially for short horizons.

Statistical models can play a crucial role in decision making. Traditional model validation tests typically make restrictive parametric assumptions about the model under the null and the alternative hypotheses. The majority of these tests examine one type of change at a time. This paper presents a method for determining whether new data continues to support the chosen model. We suggest using simulation and the kernel density estimator instead of assuming a parametric distribution for the data under the hull hypothesis. This leads to a more versatile testing procedure, one that can be applied to test different types of models and look for a variety of different types of divergences from the null hypothesis. Such a flexible testing procedure, in some cases, can also replace a range of tests that each test against particular alternative hypotheses. The procedure’s ability to recognize a change in the underlying model is demonstrated through AR(1) and linear models. We examine the power of our procedure to detect changes in the variance of the error term and the AR coefficient in the AR(1) model. In the linear model, we examine the performance of the procedure when there are changes in the error variance and error distribution, and when an economic cycle is introduced into the model. We find that the procedure has correct empirical size and high power to recognize the changes in the data generating process after 10 to 15 new observations, depending on the type and extent of the change.

In this paper we propose a new set of multivariate stochastic models that capture time varying seasonality within the vector innovations structural time series (VISTS) framework. These models encapsulate exponential smoothing methods in a multivariate setting. The models considered are the local level, local trend and damped trend VISTS models with an additive multivariate seasonal component. We evaluate their performances for forecasting international tourist arrivals from eleven source countries to Australia and New Zealand.

A new innovations state space modeling framework, incorporating Box-Cox transformations, Fourier series with time varying coefficients and ARMA error correction, is introduced for forecasting complex seasonal time series that cannot be handled using existing forecasting models. Such complex time series include time series with multiple seasonal periods, high frequency seasonality, non-integer seasonality and dual-calendar effects. Our new modelling framework provides an alternative to existing exponential smoothing models, and is shown to have many advantages. The methods for initialization and estimation, including likelihood evaluation, are presented, and analytical expressions for point forecasts and interval predictions under the assumption of Gaussian errors are derived, leading to a simple, comprehensible approach to forecasting complex seasonal time series. Our trigonometric formulation is also presented as a means of decomposing complex seasonal time series, which cannot be decomposed using any of the existing decomposition methods. The approach is useful in a broad range of applications, and we illustrate its versatility in three empirical studies where it demonstrates excellent forecasting performance over a range of prediction horizons. In addition, we show that our trigonometric decomposition leads to the identification and extraction of seasonal components, which are otherwise not apparent in the time series plot itself.

In this paper we challenge the traditional design used for forecasting competitions. We implement an online competition with a public leaderboard that provides instant feedback to competitors who are allowed to revise and resubmit forecasts. The results show that feedback significantly improves forecasting accuracy.

One of the most widely used standard procedures for model evaluation in classification and regression is K-fold cross-validation (CV). However, when it comes to time series forecasting, because of the inherent serial correlation and potential non-stationarity of the data, its application is not straightforward and often omitted by practitioners in favor of an out-of-sample (OOS) evaluation. In this paper, we show that the particular setup in which time series forecasting is usually performed using Machine Learning methods renders the use of standard K-fold CV possible. We present theoretical insights supporting our arguments. Furthermore, we present a simulation study where we show empirically that K-fold CV performs favourably compared to both OOS evaluation and other time-series-specific techniques such as non-dependent cross-validation.

In There is evidence that exponential smoothing methods as well as time varying parameter models perform relatively well in forecasting comparisons. The aim of this paper is to introduce a new forecasting technique by integrating the exponential smoothing model with regressors whose coefficients are time varying. In doing this, we construct an exponential smoothing model with regressors by extending Holt's linear exponential smoothing model. We then translate it into an equivalent state space structure so that the parameters can be estimated via the maximum likely-hood estimation procedure. Due to the potential problem in the updating equation for the regressor coefficients when the change in regressor is too small, we propose an alternative structure of the state space model which allows the updating process to be put on hold until sufficient information is available. An empirical study of forecast accuracy shows that the new model performs better than the existing exponential smoothing model as well as the linear regression model.

We propose an econometric model for predicting the share of bank debt held by bankrupt firms by combining a novel set of firm-level financial variables and macroeconomic indicators. Our firm-level data include payment remarks in the form of debt collections from private agencies and attachments from private and public agencies and cover all Norwegian limited liability companies for the period 2010–2021. We use logistic Lasso regressions to select bankruptcy predictors from a large set of potential predictors, comparing a highly sparse variable selection criterion (“the one standard error rule”) with the minimum cross validation error (CVE) criterion. Moreover, we examine the implications of using debt shares as weights in the estimation and find that weighting has a large impact on variable selection and predictions and, generally, leads to lower out-of-sample prediction errors than alternative approaches. Debt weighting combined with sparse variable selection gives the best predictions of the risk of bankruptcy in firms holding high shares of the bank debt.

A new automatic forecasting procedure is proposed based on a recent exponential smoothing framework which incorporates a Box-Cox transformation and ARMA residual corrections. The procedure is complete with well-defined methods for initialization, estimation, likelihood evaluation, and analytical derivation of point and interval predictions under a Gaussian error assumption. The algorithm is examined extensively by applying it to single seasonal and non-seasonal time series from the M and the M3 competitions, and is shown to provide competitive out-of-sample forecast accuracy compared to the best methods in these competitions and to the traditional exponential smoothing framework. The proposed algorithm can be used as an alternative to existing automatic forecasting procedures in modeling single seasonal and non-seasonal time series. In addition, it provides the new option of automatic modeling of multiple seasonal time series which cannot be handled using any of the existing automatic forecasting procedures. The proposed automatic procedure is further illustrated by applying it to two multiple seasonal time series involving call center data and electricity demand data.

This paper assesses the transmission of monetary policy in a large Bayesian vector autoregression based on the approach proposed by Banbura, Giannone and Reichlin (2010). The paper analyzes the impact of monetary policy shocks in the United States and Canada not only on a range of domestic aggregates, trade flows, and exchange rates, but also foreign investment income. The analysis provides three main results. First, a surprise monetary policy action has a statistically and economically significant impact on both gross and net foreign investment income flows in both countries. Against the background of growing foreign wealth and investment income, this result provides preliminary evidence that foreign balance-sheet channels might play an increasingly important role for monetary transmission. Second, the impact of monetary policy on foreign investment income flows differs considerably across asset categories and over time, suggesting that the investment instruments and the currency denomination of a country’s foreign assets and liabilities are potentially relevant for the way in which monetary policy affects the domestic economy. Finally, the results support existing evidence on the effectiveness of large vector autoregressions and the Bayesian shrinkage approach in addressing the curse of dimensionality and eliminating price and exchange rate puzzles.

We propose a new framework for evaluating predictive densities in an environment where the estimation error of the parameters used to construct the densities is preserved asymptotically under the null hypothesis. The tests offer a simple way to evaluate the correct specification of predictive densities, where both the model specification and its estimation technique are evaluated jointly. Monte Carlo simulation results indicate that our tests are well sized and have good power in detecting misspecification. An empirical application to density forecasts of the Survey of Professional Forecasters shows the usefulness of our methodology.

This paper provides a sectoral examination of the impact of trade policies and custom valuation procedures on estimating the dollar value of time varying import content of Japanese transplant automobiles. Using monthly data from 1985 to 1992, we introduce an error correction model (ECM) and a state space vector auto regression (SS-VAR) model to purify trade data of measurement errors induced by unobserveable prices and customs valuation procedures. Data show that US import of Japanese auto parts are elastic to the fleet of active Japanese automobiles in the U.S., inelastic to transplant production and that disequilibrium adjustments relative to transplant production are corrected in oneperiod. Further, changes in imports are responsive to the cyclical behavior of Big 3 production and the debt burden of automobile consumers. Moreover, we find that productivity trends in the automotive industry are not a significant determinant of imported parts. The model predicts that Japanese manufacturers will shift more production to the US in response to yen appreciation against the dollar. We show that import content decreased following the Fair Trade in Parts Act and the Omnibus Trade and Competitiveness Act of 1988, but increased shortly thereafter and predictions are that it will continue to increase. Therefore the empirical evidence suggests that direct trade policies designed to reduce the dollar value of import content and increase domestic sourcing of auto parts are not effective in the long run.

A new method of forecasting the pricing kernel, i.e., stochastic claim inflation or link ratio function, of incurred but not reported (IBNR) claims (in property-casualty insurance) from residuals in a dynamic claims forecast model is presented. We employ a pseudo Kalman filter approach by using claims risk exposure estimates to reconstruct innovations in stochastic claims development. Whereupon we find that the pricing kernel forecast is a product measure of the innovations. We show how these results impact performance measurement including but not limited to risk-adjusted return on capital by and through insurance accounting relationships for adjusted underwriting re-sults; and loss ratio or pure premium calculations. Additionally, we show how, in the context of Wold decomposition, diagnostics from our model can be used to compute signal to noise ratio for, and cross check, unobservable pricing kernels used to forecast claims. Furthermore, we prove that a single risk exposure factor connects seemingly unrelated specifications for loss link ratio, and claims volatility.Keywords: IBNR claims ladder; claims reserve forecast; stochastic claim inflation; claims risk exposure; link ratio function; property-casualty insurance; insurance accounting

We test for bias and efficiency of the ECB inflation forecasts using a confidential dataset of ECB macroeconomic quarterly projections. We investigate whether the properties of the forecasts depend on the level of inflation, by distinguishing whether the inflation observed by the ECB at the time of forecasting is above or below the target. The forecasts are unbiased and efficient on average, however there is evidence of state dependence. In particular, the ECB tends to overpredict (underpredict) inflation at intermediate forecast horizons when inflation is below (above) target. The magnitude of the bias is larger when inflation is above the target. These results hold even after accounting for errors in the external assumptions. We also find evidence of inefficiency, in the form of underreaction to news, but only when inflation is above the target. Our findings bear important implications for the ECB forecasting process and ultimately for its communication strategy.

The authors examine the effectiveness of labor cost reductions as a means to stimulate economic activity and assesses the differences which may occur with the prevailing exchange rate regime. They develop a medium-scale three-region DSGE model and show that the impact of a cut in the employers’ social security contributions rate does not vary significantly under different exchange rate regimes. They find that both the interest rate and the exchange rate channel matters. Furthermore, the measure appears to be effective even if it comes along with a consumption tax increase to preserve long-term fiscal sustainability. Finally, they assess whether obtained theoretical results hold up empirically by applying the local projection method. Regression results suggest that changes in employers’ social security contributions rates have statistically significant real effects – a one percentage point reduction leads to an average cumulative rise in output of around 1.3 percent in the medium term. Moreover, the outcome does not differ significantly across the different exchange rate regimes.

We defend the forecasting performance of the FOMC from the recent criticism of Christina and David Romer. Our argument is that the FOMC forecasts a worst-case scenario that it uses to design decisions that will work well enough (are robust) despite possible misspecification of its model. Because these FOMC forecasts are not predictions of what the FOMC expects to occur under its model, it is inappropriate to compare their performance in a horse race against other forecasts. Our interpretation of the FOMC as a robust policymaker can explain all the findings of the Romers and rationalises differences between FOMC forecasts and forecasts published in the Greenbook by the staff of the Federal Reserve System.

Functional Signal plus Noise (FSN) models are proposed for analysing the dynamics of a large cross-section of yields or asset prices in which contemporaneous observations are functionally related. The FSN models are used to forecast high dimensional yield curves for US Treasury bonds at the one month ahead horizon. The models achieve large reductions in mean square forecast errors relative to a random walk for yields and readily dominate both the Diebold and Li (2006) and random walk forecasts across all maturities studied. We show that the Expectations Theory (ET) of the term structure completely determines the conditional mean of any zero-coupon yield curve. This enables a novel evaluation of the ET in which its 1-step ahead forecasts are compared with those of rival methods such as the FSN models, with the results strongly supporting the growing body of empirical evidence against the ET. Yield spreads do provide important information for forecasting the yield curve, especially in the case of shorter maturities, but not in the manner prescribed by the Expectations Theory.

We investigate alternative robust approaches to forecasting, using a new class of robust devices, contrasted with equilibrium correction models. Their forecasting properties are derived facing a range of likely empirical problems at the forecast origin, including measurement errors, implulses, omitted variables, unanticipated location shifts and incorrectly included variables that experience a shift. We derive the resulting forecast biases and error variances, and indicate when the methods are likely to perform well. The robust methods are applied to forecasting US GDP using autoregressive models, and also to autoregressive models with factors extracted from a large dataset of macroeconomic variables. We consider forecasting performance over the Great Recession, and over an earlier more quiescent period.

Using a unique dataset of 22.5 million news articles from the Dow Jones Newswires Archive, we perform an in depth real-time out-of-sample forecasting comparison study with one of the most widely used data sets in the newer forecasting literature, namely the FRED-MD dataset. Focusing on U.S. GDP, consumption and investment growth, our results suggest that the news data contains information not captured by the hard economic indicators, and that the news-based data are particularly informative for forecasting consumption developments.

We use a novel data set covering all domestic debit card transactions in physical terminals by Norwegian households, to nowcast quarterly Norwegian household consumption. These card payments data are free of sampling errors and are available weekly without delays, providing a valuable early indicator of household spending. To account for mixed-frequency data, we estimate various mixed-data sampling (MIDAS) regressions using predictors sampled at monthly and weekly frequency. We evaluate both point and density forecasting performance over the sample 2011Q4-2020Q1. Our results show that MIDAS regressions with debit card transactions data improve both point and density forecast accuracy over competitive standard benchmark models that use alternative high-frequency predictors. Finally, we illustrate the benefits of using the card payments data by obtaining a timely and relatively accurate nowcast of the first quarter of 2020, a quarter characterized by heightened uncertainty due to the COVID-19 pandemic.

This paper provides an overview of how to use \big data" for economic research. We investigate the performance and ease of use of different Spark applications running on a distributed file system to enable the handling and analysis of data sets which were previously not usable due to their size. More specifically, we explain how to use Spark to (i) explore big data sets which exceed retail grade computers memory size and (ii) run typical econometric tasks including microeconometric, panel data and time series regression models which are prohibitively expensive to evaluate on stand-alone machines. By bridging the gap between the abstract concept of Spark and ready-to-use examples which can easily be altered to suite the researchers need, we provide economists and social scientists more generally with the theory and practice to handle the ever growing datasets available. The ease of reproducing the examples in this paper makes this guide a useful reference for researchers with a limited background in data handling and distributed computing.

We evaluate the role of financial conditions as predictors of macroeconomic risk first in the quantile regression framework of Adrian et al. (2019b), which allows for non-linearities, and then in a novel linear semi-structural model as proposed by Hasenzagl et al. (2018). We distinguish between price variables such as credit spreads and stock variables such as leverage. We find that (i) although the spreads correlate with the left tail of the conditional distribution of GDP growth, they provide limited advanced information on growth vulnerability; (ii) nonfinancial leverage provides a leading signal for the left quantile of the GDP growth distribution in the 2008 recession; (iii) measures of excess leverage conceptually similar to the Basel gap, but cleaned from business cycle dynamics via the lenses of the semi-structural model, point to two peaks of accumulation of risks – the eighties and the first eight years of the new millennium, with an unstable relationship with business cycle chronology.

The positive relationship between real exchange rates and natural resource income is well understood and studied. However, climate change and the transition to a lower-carbon economy now challenges this relationship. We document this by proposing a novel news media-based measure of climate change transition risk and show that when such risk is high, major commodity currencies experience a persistent depreciation and the relationship between commodity price fluctuations and currencies tends to become weaker.

This paper investigates whether economic activity dynamics predict GDP revisions using panel data from 15 OECD countries. We find that economic activity predicts GDP revisions: early releases tend to overestimate GDP growth during slowdowns –and vice-versa. We also find that the source of the predictability could be related to the sampling of information collection. Finally, the predictability comes from short-term economic activity dynamics rather than business cycle position.

Facing several economic and financial uncertainties, assessing accurately global economic conditions is a great challenge for economists. The International Monetary Fund proposes within its periodic World Economic Outlook report a measure of the global GDP annual growth, that is often considered as the benchmark nowcast by macroeconomists. In this paper, we put forward an alternative approach to provide monthly nowcasts of the annual global growth rate. Our approach builds on a Factor-Augmented MIxed DAta Sampling (FA-MIDAS) model that enables (i) to account for a large monthly database including various countries and sectors of the global economy and (ii) to nowcast a low-frequency macroeconomic variable using higher-frequency information. Pseudo real-time results show that this approach provides reliable and timely nowcasts of the world GDP annual growth on a monthly basis.

In this article, it is analysed the multihorizon predictive power of the Hybrid New Keynesian Phillips Curve (HNKPC) making use of a compact-scale Global VAR for the headline inflation of six developed countries with different inflationary experiences; covering from 2000.1 until 2014.12. The key element of this article is the use of direct measures of inflation expectations--Consensus Economics--embedded in a Global VAR environment, i.e. modelling cross-country interactions. The Global VAR point forecast is evaluated using the Mean Squared Forecast Error (MSFE) statistic and statistically compared with several benchmarks. These belong to traditional statistical modelling, such as autoregressions (AR), the exponential smoothing model (ES), and the random walk model (RW). One last economics-based benchmark is the closed economy univariate HNKPC. The results indicate that the Global VAR is a valid forecasting procedure especially for the short-run. The most accurate forecasts, however, are obtained with the AR and especially with the univariate HNKPC. In the long-run, the ES model also appears as a better alternative rather than the RW. The MSPE is obviously affected by the unanticipated effects of the financial crisis started in 2008. So, when considering an evaluation sample just before the crisis, the GVAR also appears as a valid alternative in the long-run. The most robust forecasting devices across countries and horizons result in the univariate HNKPC, giving a role for economic fundamentals when forecasting inflation.

In this paper, we revisit bear market predictability by employing a number of variables widely used in forecasting stock returns. In particular, we focus on variables related to the presence of imperfect credit markets. We evaluate prediction performance using in-sample and out-of-sample tests. Empirical evidence from the US stock market suggests that among the variables we investigate, the default yield spread, inflation, and the term spread are useful in predicting bear markets. Further, we find that the default yield spread provides superior out-of-sample predictability for bear markets one to three months ahead, which suggests that the external finance premium has an informative content on the financial market.

This article studies the pruned state-space system for higher-order perturbation approximations to dynamic stochastic general equilibrium (DSGE) models.We show the stability of the pruned approximation up to third order and provide closed-form expressions for first and second unconditional moments and impulse response functions. Our results introduce generalized method of moments (GMM) estimation and impulse-response matching for DSGE models approximated up to third order and provide a foundation for indirect inference and simulated method of moments (SMM). As an application, we consider a New Keynesian model with Epstein-Zin preferences and two novel feedback effects from long-term bonds to the real economy, allowing us to match the level and variability of the 10-year term premium in the U.S. with a low relative risk aversion of 5.

We investigate the role played by the media in the expectations formation process of households. Using a novel news-topic-based approach we show that news types the media choose to report on, e.g., fiscal policy, health, and politics, are good predictors of households' stated inflation expectations. In turn, in a noisy information model setting, augmented with a simple media channel, we document that the underlying time series properties of relevant news topics explain the time-varying information rigidity among households. As such, we not only provide a novel estimate showing the degree to which information rigidities among households varies across time, but also provide, using a large news corpus and machine learning algorithms, robust and new evidence highlighting the role of the media for understanding inflation expectations and information rigidities.

We present new methodology and a case study in use of a class of Bayesian predictive synthesis (BPS) models for multivariate time series forecasting. This extends the foundational BPS framework to the multivariate setting, with detailed application in the topical and challenging context of multistep macroeconomic forecasting in a monetary policy setting. BPS evaluates—sequentially and adaptively over time—varying forecast biases and facets of miscalibration of individual forecast densities for multiple time series, and—critically—their time-varying inter-dependencies. We define BPS methodology for a new class of dynamic multivariate latent factor models implied by BPS theory. Structured dynamic latent factor BPS is here motivated by the application context—sequential forecasting of multiple U.S. macroeconomic time series with forecasts generated from several traditional econometric time series models. The case study highlights the potential of BPS to improve of forecasts of multiple series at multiple forecast horizons, and its use in learning dynamic relationships among forecasting models or agents. Supplementary materials for this article, including a standardized description of the materials available for reproducing the work, are available as an online supplement.

We develop efficient simulation techniques for Bayesian inference on switching GARCH models. Our contribution to existing literature is manifold. First, we discuss different multi-move sampling techniques for Markov Switching (MS) state space models with particular attention to MS-GARCH models. Our multi-move sampling strategy is based on the Forward Filtering Backward Sampling (FFBS) applied to an approximation of MS-GARCH. Another important contribution is the use of multi-point samplers, such as the Multiple-Try Metropolis (MTM) and the Multiple trial Metropolize Independent Sampler, in combination with FFBS for the MS-GARCH process. In this sense we ex- tend to the MS state space models the work of So [2006] on efficient MTM sampler for continuous state space models. Finally, we suggest to further improve the sampler efficiency by introducing the antithetic sampling of Craiu and Meng [2005] and Craiu and Lemieux [2007] within the FFBS. Our simulation experiments on MS-GARCH model show that our multi-point and multi-move strategies allow the sampler to gain efficiency when compared with single-move Gibbs sampling., 38 pages, 7 figures

In this paper, we extract common factors from a cross-section of U.S. macro-variables and Treasury zero-coupon yields. We find that two macroeconomic factors have an important predictive content for government bond yields and excess returns. These factors are not spanned by the cross-section of yields and are well proxied by economic growth and real interest rates.

Temperature plays a key role in driving the electricity demand. We adopt the “recency effect”, a term drawn from psychology, to represent the fact that the electricity demand is affected by the temperatures of the preceding hours. In the load forecasting literature, the temperature variables are often constructed in the form of lagged hourly temperatures and moving average temperatures. In the past, computing power has limited the amount of temperature variables that can be used in a load forecasting model. In this paper, we present a comprehensive study to model the recency effect using a big data approach. We take advantage of modern computing power to answer a fundamental question: how many lagged hourly temperatures and/or moving average temperatures are needed in a regression model in order to capture the recency effect fully without compromising the forecasting accuracy? Using a case study based on data from the load forecasting track of the Global Energy Forecasting Competition 2012, we first demonstrate that a model with the recency effect outperforms its counterpart (a.k.a. Tao’s Vanilla Benchmark Model) by 18% to 21% for forecasting the load series at the top (aggregated) level. We then model the recency effect in order to customize load forecasting models at the bottom level of a geographic hierarchy, again showing a superiority over the benchmark model of 12% to 15% on average. Finally, we discuss four different implementations of the recency effect modeling by hour of a day. In addition, this paper also presents two interesting findings: 1) the naive models are not useful for benchmark purposes in load forecasting at aggregated level due to their lack of accuracy; and 2) slicing the data into 24 pieces to develop one model for each hour is not necessarily better than building one interaction regression model using all 24 hours together.

Summary This paper introduces classification tree ensembles (CTEs) to the banking crisis forecasting literature. I show that CTEs substantially improve out-of-sample forecasting performance over best-practice early-warning systems. CTEs enable policymakers to correctly forecast 80% of crises with a 20% probability of incorrectly forecasting a crisis. These findings are based on a long-run sample (1870–2011), and two broad post-1970 samples which together cover almost all known systemic banking crises. I show that the marked improvement in forecasting performance results from the combination of many classification trees into an ensemble, and the use of many predictors. Copyright © 2016 John Wiley & Sons, Ltd.

Although many macroeconomic time series are assumed to follow nonlinear processes, nonlinear models often do not provide better predictions than their linear counterparts. Furthermore, nonlinear models easily become very complex and difficult to estimate. The aim of this study is to investigate whether simple nonlinear extensions of autoregressive processes are able to provide more accurate forecasting results than linear models. Therefore, simple autoregressive processes are extended by means of nonlinear transformations (quadratic, cubic, sine, exponential functions) of lagged time series observations and autoregression residuals. The proposed forecasting models are applied to a large set of macroeconomic and financial time series for 10 European countries. Findings suggest that these models, including nonlinear transformation of lagged autoregression residuals, are able to provide better forecasting results than simple linear models. Thus, it may be possible to improve the forecasting accuracy of linear models by including nonlinear components. This is especially true for time series that are positively tested for nonlinear characteristics and longer forecast horizons.

Using forecasts of the inflation rate in South Africa, we study the rationality of forecasts and the shape of forecasters’ loss functions. When we study micro-level data of individual forecasts, we find mixed evidence of an asymmetric loss function, suggesting that inflation forecasters are heterogeneous with respect to the shape of their loss functions. We also find strong evidence that inflation forecasts are in line with forecast rationality. When we pool the data and study sectoral inflation forecasts of financial analysts, trade unions and the business sector, we find evidence for asymmetry in the loss function and against forecast rationality. Upon comparing the micro-level results with those for pooled and sectoral data, we conclude that forecast rationality should be assessed based on micro-level data, and that freer access to this data would allow a more rigorous analysis and discussion of the information content of the surveys.

Some observers have conjectured that oil supply shocks in the United States and in other countries are behind the plunge in the price of oil since June 2014. Others have suggested that a major shock to oil price expectations occurred when in late November 2014 OPEC announced that it would maintain current production levels despite the steady increase in non-OPEC oil production. Both conjectures are perfectly reasonable ex ante, yet we provide quantitative evidence that neither explanation appears supported by the data. We show that more than half of the decline in the price of oil was predictable in real time as of June 2014 and therefore must have reflected the cumulative effects of earlier oil demand and supply shocks. Among the shocks that occurred after June 2014, the most influential shock resembles a negative shock to the demand for oil associated with a weakening economy in December 2014. In contrast, there is no evidence of any large positive oil supply shocks between June and December. We conclude that the difference in the evolution of the price of oil, which declined by 44% over this period, compared with other commodity prices, which on average only declined by about 5%-15%, reflects oil-market specific developments that took place prior to June 2014.

We construct an empirical heterogeneous agent model which optimally combines forecasts from fundamentalist and chartist agents and evaluates its out-of-sample forecast performance using daily data covering an overall period from January 1999 to June 2014 for six of the most widely traded currencies. We use daily financial data such as level, slope and curvature yield curve factors, equity prices, as well as risk aversion and global trade activity measures in the fundamentalist agent’s predictor set to obtain a proxy for the market’s view on the state of the macroeconomy. Chartist agents rely upon standard momentum, moving average and relative strength index technical indicators in their predictor set. Individual agent specific forecasts are constructed using a flexible dynamic model averaging framework and are then aggregated into a model combined forecast using a forecast combination regression. We show that our empirical heterogeneous agent model produces statistically significant and sizable forecast improvements over a random walk benchmark, reaching out-of-sample R2 values of 1.41, 1.07, 0.99 and 0.74 percent at the daily one-step ahead horizon for 4 out of the 6 currencies that we consider. Forecast gains remain significant for horizons up to threedays ahead. The forecast improvements are largely realised before and around the time of the Lehman Brothers collapse. We show further that our model combined forecasts produce economic value to a mean variance investor, yielding annualized Sharpe ratios of around 0.89 and annualized performance fees in excess of 460 basis points.

Real-time assessment of quarterly GDP growth rates is crucial for evaluation of economy’s current perspectives given the fact that respective data is normally subject to substantial publication delays by national statistical agencies. Large information sets of real-time indicators which could be used to approximate GDP growth rates in the quarter of interest are in practice characterized by unbalanced data, mixed frequencies, systematic data revisions, as well as a more general curse of dimensionality problem. The latter issues could, however, be practically resolved by means of dynamic factor modeling that has recently been recognized as a helpful tool to evaluate current economic conditions by means of higher frequency indicators. Our major results show that the performance of dynamic factor models in predicting Russian GDP dynamics appears to be superior as compared to other common alternative specifications. At the same time, we empirically show that the arrival of new data seems to consistently improve DFM’s predictive accuracy throughout sequential nowcast vintages. We also introduce the analysis of nowcast evolution resulting from the gradual expansion of the dataset of explanatory variables, as well as the framework for estimating contributions of different blocks of predictors into now-casts of Russian GDP.

This paper offers a theoretical explanation for the stylized fact that forecast combinations with estimated optimal weights often perform poorly in applications. The properties of the forecast combination are typically derived under the assumption that the weights are fixed, while in practice they need to be estimated. If the fact that the weights are random rather than fixed is taken into account during the optimality derivation, then the forecast combination will be biased (even when the original forecasts are unbiased) and its variance is larger than in the fixed-weights case. In particular, there is no guarantee that the ‘optimal’ forecast combination will be better than the equal-weights case or even improve on the original forecasts. We provide the underlying theory, some special cases, and a numerical illustration. publisher: Elsevier articletitle: The forecast combination puzzle: A simple theoretical explanation journaltitle: International Journal of Forecasting articlelink: http://dx.doi.org/10.1016/j.ijforecast.2015.12.005 content_type: article copyright: Crown Copyright © 2016 Published by Elsevier B.V. on behalf of International Institute of Forecasters. All rights reserved. ispartof: International Journal of Forecasting vol:32 issue:3 pages:754-762 status: published