On the forecasting performance of a small-scale DSGE model

Dynamic stochastic general equilibrium (DSGE) models have recently become standard tools for policy analysis. Nevertheless, their forecasting properties have still barely been explored. In this article, we address this problem by examining the quality of forecasts of the key U.S. economic variables: the three-month Treasury bill yield, the GDP growth rate and GDP price index inflation, from a small-size DSGE model, trivariate vector autoregression (VAR) models and the Philadelphia Fed Survey of Professional Forecasters (SPF). The ex post forecast errors are evaluated on the basis of the data from the period 1994–2006. We apply the Philadelphia Fed “Real-Time Data Set for Macroeconomists” to ensure that the data used in estimating the DSGE and VAR models was comparable to the information available to the SPF.Overall, the results are mixed. When comparing the root mean squared errors for some forecast horizons, it appears that the DSGE model outperforms the other methods in forecasting the GDP growth rate. However, this characteristic turned out to be statistically insignificant. Most of the SPF's forecasts of GDP price index inflation and the short-term interest rate are better than those from the DSGE and VAR models.     

A Note on the Identification of Dynamic Economic Models with Generalized Shock Processes

Dynamic stochastic general equilibrium (DSGE) models with generalized shock processes, such as shock processes which follow a vector autoregression (VAR), have been an active area of research in recent years. Unfortunately, the structural parameters governing DSGE models are not identified when the driving process behind the model follows an unrestricted VAR. This finding implies that parameter estimates derived from recent attempts to estimate DSGE models with generalized driving processes should be treated with caution, and that there always exists a tradeoff between identification and the risk of model misspecification. However, these results also make it easier to address the issue of model misspecification by making it computationally easier to check the validity of cross‐equation restrictions.     

Identifying the true nature of price discovery and cross-market informational flow in the investment grade CDS and equity markets

We examine the comparative efficiency of systematic investment grade credit default swap (CDS) and equity markets using a time-varying coefficient vector autoregression. This modeling framework enables a view of cross-market informational flow along each point in the time-period under investigation by taking into account parameter instability. We obtain smoothing estimates of parameters capturing such flow between CDS and equity markets using daily data from 2004 to 2019, and measure the strength of flow via relative predictive gains. In contrast to prior studies, we find a two-way interactive effect in which certain types of information are captured more efficiently in prices by each market. We also find that the time-varying coefficient vector autoregression results in superior forecasting gains relative to models not accounting for price discovery. These results have implications for systematic investors, arbitrageurs and stakeholders who monitor systematic markets for their informational content.     

Combining VAR and DSGE forecast densities

A popular macroeconomic forecasting strategy utilizes many models to hedge against instabilities of unknown timing; see (among others) Stock and Watson (2004), Clark and McCracken (2010), and Jore et al. (2010). Existing studies of this forecasting strategy exclude dynamic stochastic general equilibrium (DSGE) models, despite the widespread use of these models by monetary policymakers. In this paper, we use the linear opinion pool to combine inflation forecast densities from many vector autoregressions (VARs) and a policymaking DSGE model. The DSGE receives a substantial weight in the pool (at short horizons) provided the VAR components exclude structural breaks. In this case, the inflation forecast densities exhibit calibration failure. Allowing for structural breaks in the VARs reduces the weight on the DSGE considerably, but produces well-calibrated forecast densities for inflation.     

Short-term forecasting of GDP with a DSGE model augmented by monthly indicators

DSGE models are useful tools for evaluating the impact of policy changes, but their use for (short-term) forecasting is still in its infancy. Besides theory-based restrictions, the timeliness of data is an important issue. Since DSGE models are based on quarterly data, they suffer from the publication lag of quarterly national accounts. In this paper we present a framework for the short-term forecasting of GDP based on a medium-scale DSGE model for a small open economy within a currency area. We utilize the information available in monthly indicators based on the approach proposed by Giannone et al. (2009). Using Austrian data, we find that the forecasting performance of the DSGE model can be improved considerably by incorporating monthly indicators, while still maintaining the story-telling capability of the model.     

TIME‐VARYING DYNAMICS OF THE REAL EXCHANGE RATE: AN EMPIRICAL ANALYSIS

We use a time‐varying structural vector autoregression to investigate evolving dynamics of the real exchange rate for the UK, euro area and Canada. We show that demand and nominal shocks have a substantially larger impact on the real exchange rate after the mid 1980s. Real exchange rate volatility, relative to fundamentals, also shows a marked increase after this point in time. However, there is some evidence suggesting a closer business cycle co‐movement of the real exchange rate and fundamentals. Simulations from an open‐economy DSGE model show that these results are consistent with a decline in exchange rate pass‐through. Copyright © 2012 John Wiley & Sons, Ltd.     

Testing the Structural Interpretation of the Price Puzzle with a Cost‐Channel Model*

We estimate a variety of small‐scale new‐Keynesian DSGE models with the cost channel to assess their ability to replicate the ‘price puzzle’, i.e. the inflationary impact of a monetary policy shock typically arising in vector autoregression (VAR) analysis. To correctly identify the monetary policy shock, we distinguish between a standard policy rate shifter and a shock to ‘trend inflation’, i.e. the time‐varying inflation target set by the Fed. Our estimated models predict a negative inflation reaction to a monetary policy tightening. We offer a discussion of the possible sources of mismatch between the VAR evidence and our own.     

EVALUATING REAL‐TIME VAR FORECASTS WITH AN INFORMATIVE DEMOCRATIC PRIOR

This paper proposes Bayesian forecasting in a vector autoregression using a democratic prior. This prior is chosen to match the predictions of survey respondents. In particular, the unconditional mean for each series in the vector autoregression is centered around long‐horizon survey forecasts. Heavy shrinkage toward the democratic prior is found to give good real‐time predictions of a range of macroeconomic variables, as these survey projections are good at quickly capturing endpoint shifts. Copyright © 2012 John Wiley & Sons, Ltd.     

Levels,Differences and ECMs – Principles for Improved Econometric Forecasting*

Unit‐root testing can be a preliminary step in model development, an intermediate step, or an end in itself. Some researchers have questioned the value of any unit‐root and cointegration testing, arguing that restrictions based on theory are at least as effective. Such confusion is unsatisfactory. Needed is a set of principles that limit and define the role of the tacit knowledge of the model builders. In a forecasting context, we enumerate the various possible model selection strategies and, based on simulation and empirical evidence, recommend using these tests to improve the specification of an initial general vector autoregression model.     

Macroeconomic Effects of Precautionary Demand for Oil

We evaluate the macroeconomic effects of shocks specific to the oil market, which mainly reflect fluctuations in precautionary demand for oil driven by uncertainty about future supplies. A two‐stage identification procedure is used. First, daily changes in the futures–spot price spread proxy for precautionary demand shocks and the path of oil prices is estimated. This information is then exploited to restrict the oil price response in a vector autoregression. Impulse responses suggest that such shocks reduce output and raise prices. Historical decomposition shows that they contributed significantly to the US recessions in the 1990s and in the early 2000s, but not to the most recent slump. Copyright © 2014 John Wiley & Sons, Ltd.     

Evaluating Point and Density Forecasts of DSGE Models

This paper investigates the accuracy of forecasts from four dynamic stochastic general equilibrium (DSGE) models for inflation, output growth and the federal funds rate using a real‐time dataset synchronized with the Fed's Greenbook projections. Conditioning the model forecasts on the Greenbook nowcasts leads to forecasts that are as accurate as the Greenbook projections for output growth and the federal funds rate. Only for inflation are the model forecasts dominated by the Greenbook projections. A comparison with forecasts from Bayesian vector autoregressions shows that the economic structure of the DSGE models which is useful for the interpretation of forecasts does not lower the accuracy of forecasts. Combining forecasts of several DSGE models increases precision in comparison to individual model forecasts. Comparing density forecasts with the actual distribution of observations shows that DSGE models overestimate uncertainty around point forecasts. Copyright © 2013 John Wiley & Sons, Ltd.     

Mixed random forest,cointegration, and forecasting gasoline prices

One of the most successful forecasting machine learning (ML) procedures is random forest (RF). In this paper, we propose a new mixed RF approach for modeling departures from linearity that helps identify (i) explanatory variables with nonlinear impacts, (ii) threshold values, and (iii) the closest parametric approximation. The methodology is applied to weekly forecasts of gasoline prices, cointegrated with international oil prices and exchange rates. Recent specifications for nonlinear error correction (NEC) models include threshold autoregressive models (TAR) and double-threshold smooth transition autoregressive (STAR) models. We propose a new mixed RF model specification strategy and apply it to the determinants of weekly prices of the Spanish gasoline market from 2010 to 2019. In particular, the mixed RF is able to identify nonlinearities in both the error correction term and the rate of change of oil prices. It provides the best weekly gasoline price forecasting performance and supports the logistic error correction model (ECM) approximation.     

An identification‐robust test for time‐varying parameters in the dynamics of energy prices

We test for the presence of time‐varying parameters (TVP) in the long‐run dynamics of energy prices for oil, natural gas and coal, within a standard class of mean‐reverting models. We also propose residual‐based diagnostic tests and examine out‐of‐sample forecasts. In‐sample LR tests support the TVP model for coal and gas but not for oil, though companion diagnostics suggest that the model is too restrictive to conclusively fit the data. Out‐of‐sample analysis suggests a random‐walk specification for oil price, and TVP models for both real‐time forecasting in the case of gas and long‐run forecasting in the case of coal. Copyright © 2010 John Wiley & Sons, Ltd.     

What (really) accounts for the fall in hours after a technology shock?

The paper asks how state of the art DSGE models that account for the conditional response of hours following a positive neutral technology shock compare in a marginal likelihood race. To that end we construct and estimate several competing small-scale DSGE models that extend the standard real business cycle model. In particular, we identify from the literature six different hypotheses that generate the empirically observed decline in hours worked after a positive technology shock. These models alternatively exhibit (i) sticky prices; (ii) firm entry and exit with time to build; (iii) habit in consumption and costly adjustment of investment; (iv) persistence in the permanent technology shocks; (v) labor market friction with procyclical hiring costs; and (vi) Leontief production function with labor-saving technology shocks. In terms of model posterior probabilities, impulse responses, and autocorrelations, the model favored is the one that exhibits habit formation in consumption and investment adjustment costs. A robustness test shows that the sticky price model becomes as competitive as the habit formation and costly adjustment of investment model when sticky wages are included.     

Estimating and accounting for the output gap with large Bayesian vector autoregressions

We consider how to estimate the trend and cycle of a time series, such as real gross domestic product, given a large information set. Our approach makes use of the Beveridge–Nelson decomposition based on a vector autoregression, but with two practical considerations. First, we show how to determine which conditioning variables span the relevant information by directly accounting for the Beveridge–Nelson trend and cycle in terms of contributions from different forecast errors. Second, we employ Bayesian shrinkage to avoid overfitting in finite samples when estimating models that are large enough to include many possible sources of information. An empirical application with up to 138 variables covering various aspects of the US economy reveals that the unemployment rate, inflation, and, to a lesser extent, housing starts, aggregate consumption, stock prices, real money balances, and the federal funds rate contain relevant information beyond that in output growth for estimating the output gap, with estimates largely robust to substituting some of these variables or incorporating additional variables.     

Adaptive learning forecasting,with applications in forecasting agricultural prices

We introduce a new forecasting methodology, referred to as adaptive learning forecasting, that allows for both forecast averaging and forecast error learning. We analyze its theoretical properties and demonstrate that it provides a priori MSE improvements under certain conditions. The learning rate based on past forecast errors is shown to be non-linear. This methodology is of wide applicability and can provide MSE improvements even for the simplest benchmark models. We illustrate the method’s application using data on agricultural prices for several agricultural products, as well as on real GDP growth for several of the corresponding countries. The time series of agricultural prices are short and show an irregular cyclicality that can be linked to economic performance and productivity, and we consider a variety of forecasting models, both univariate and bivariate, that are linked to output and productivity. Our results support both the efficacy of the new method and the forecastability of agricultural prices.     

Speculation in the Oil Market

The run‐up in oil prices since 2004 coincided with growing investment in commodity markets and increased price co‐movement among different commodities. We assess whether speculation in the oil market played a role in driving this salient empirical pattern. We identify oil shocks from a large dataset using a dynamic factor model. This method is motivated by the fact that a small‐scale vector autoregression is not informationally sufficient to identify the shocks. The main results are as follows. (i) While global demand shocks account for the largest share of oil price fluctuations, speculative shocks are the second most important driver. (ii) The increase in oil prices over the last decade is mainly driven by the strength of global demand. However, speculation played a significant role in the oil price increase between 2004 and 2008 and its subsequent collapse. (iii) The co‐movement between oil prices and the prices of other commodities is mainly explained by global demand shocks. Our results support the view that the recent oil price increase is mainly driven by the strength of global demand but that the financialization process of commodity markets also played a role. Copyright © 2014 John Wiley & Sons, Ltd.     

Modeling the price side of econometric models: An analysis of the underlying hypotheses

Using a non-linear vector autoregression, this paper investigates the dynamic interactions between a set of disaggregated price series. A hypothesis, positing that all sectoral prices are a function of a ‘major price index’, is tested and the way this hypothesis is used in econometric models is discussed. The model is based upon the theory of covariance stationary time series. Non-linear estimation procedures are used.     

DSGE forecasts of the lost recovery

The years following the Great Recession were challenging for forecasters. Unlike other deep downturns, this recession was not followed by a swift recovery, but instead generated a sizable and persistent output gap that was not accompanied by deflation as a traditional Phillips curve relationship would have predicted. Moreover, the zero lower bound and unconventional monetary policy generated an unprecedented policy environment. We document the actual real-time forecasting performance of the New York Fed dynamic stochastic general equilibrium (DSGE) model during this period and explain the results using the pseudo real-time forecasting performance results from a battery of DSGE models. We find the New York Fed DSGE model’s forecasting accuracy to be comparable to that of private forecasters, and notably better for output growth than the median forecasts from the FOMC’s Summary of Economic Projections. The model’s financial frictions were key in obtaining these results, as they implied a slow recovery following the financial crisis.     

Oil prices and exchange rates in oil-exporting countries: evidence from TAR and M-TAR models

The paper studies the long-run relation and short-run dynamics between real oil prices and real exchange rates in a sample of 13 oil-exporting countries. The purpose of the study is to examine the possibility of Dutch disease in these countries. Tests of cointegration using threshold and momentum-threshold autoregressive (TAR and M-TAR) models suggest the possibility of the disease in 3-out-of 13 countries??Bolivia, Mexico and Norway. For these countries, we also find that (a) oil prices have a long-run effect on the exchange rates; and (b) exchange rates adjust faster to positive deviations from the equilibrium; and (c) there is no evidence of short-run causality between real exchange rates and real oil prices in either direction. Over all, these findings suggest a weak link between oil prices and real exchange rates and thus limited evidence in favor of the Dutch disease.