Maik H. Wolters, "Evaluating Point and Density Forecasts of DSGE Models", Journal of Applied Econometrics, Vol. 30, No. 1, 2015, pp. 74-96. Overview There are three datasets (each in an excel and a .csv version), and there is also an online appendix, Appendix-MW.pdf. The CSV datasets are ASCII files in DOS format. They are zipped in the three files mw-realtime-data-csv.zip, mw-revised-data-csv.zip, and mw-greenbook-data-csv.zip. Unix/Linux users should use "unzip -a". The Excel datasets, which contain more information (see below), are zipped in the files mw-realtime-data-xls.zip, mw-revised-data-xls.zip, and mw-greenbook-data-xls.zip. Dataset 1: Large real-time dataset from Faust and Wright, 2009, (J. Faust and J.H. Wright, 2009. "Comparing Greenbook and Reduced Form Forecasts using a Large Realtime Dataset", Journal of Business and Economic Statistics, 27(4): 468-479). The dataset has been downloaded from Jon Faust's website: http://e105.org/faustj/progsData.php?d=n and is contained in the folders mw-realtime-data-xls.zip and mw-realtime-data-csv.zip. The files include the extracted transformed time series used for the estimation of the DSGE models. Dataset 2: Data realizations as released two quarters after the quarter the data refers to (include preliminary revisions) from the Philadelphia Fed real-time database. The data are available here: http://www.philadelphiafed.org/research-and-data/real-time-center/real-time-data/. The data are contained in the folders mw-realtime-data-xls.zip and mw-realtime-data-csv.zip. Dataset 3: Greenbook nowcasts and forecasts from the Philadelphia Fed Greenbook datasets. The data are available here: http://www.philadelphiafed.org/research-and-data/real-time-center/greenbook-data/. The data are contained in the folders mw-greenbook-data-xls.zip and mw-greenbook-data-csv.zip. Details on the three datasets: Dataset 1: The Excel version contains all data transformations, while the .csv version only contains the raw data and the final transformed data. Description of the Excel version: Each file contains one vintage of data. The vintage date is indicated by the file name. Each Excel file contains two sheets. - The first sheet "myyyymmdd" contains the original real-time data vintage from Faust and Wright (2009). The first column in each is the date, the second column is the horizon and the remaining columns apart from the last four represent the series extracted from Greenbook archives in level form. the fourth to last column is _snp500. This represents the annualized return on the S&P 500 and does not come from the Greenbook archives. It is of course not subject to data revision. The final three columns represent the growth rate of GNP/GDP, the rate of inflation (as measured by the GNP/GDP deflator) and the rate of inflation (as measured by CPI). The mnemonics are given in the file, mnemonics.xls. The data are used to estimate the large Bayesian VAR and serve also as the basis for the estimation of the DSGE models as described in the following item. The last six rows of each series contain the Greenbook nowcasts and projections as used by Faust and Wright, 2009 (see also the online appendix for information on the Greenbook projections from the Faust and Wright dataset). - The second sheet "DSGEdata" contains transformations of the real-time data and some additional real-time data series from the St. Louis Fed real-time database "Alfred". I apply various transformations to the data and the final time series that are used to estimate the DSGE models can be found in columns M, N, O, Q, T, Y, AA, AQ, AR, AS, AT, AU, AV and AW. They have the following names and content: * xgdp_q_obs: annualized quarter-on-quarter real output growth * pgdp_q_obs: annualized quarter-on-quarter deflator inflation * rff_q_obs: quarterly federal funds rate * pcer_q_obs: quarterly real personal consumption spending growth * fpi_q_obs: quarterly real fixed private investment growth * hours_obs: quarterly detrended hours per person * wage_obs: quarterly real wage growth * pecnn_q_obs: quarterly nominal growth rate of nondurable and service consumption * pecd_q_obs: quarterly nominal growth rate of durable consumption * per_q_obs: quarterly nominal growth rate of residential investment * penr_q_obs: quarterly nominal growth rate of nonresidential investment * paipc_q_obs: quarterly inflation of nondurable and service consumption * paipk_q_obs: quarterly inflation of durable consumption * hoursedo_obs: quarterly detrended hours per person The model by Del Negro & Schorfheide and the model by Fuhrer are estimated on: xgdp_q_obs, pgdp_q_obs, rff_q_obs. The model by Smets & Wouters is estimated on: xgdp_q_obs, pgdp_q_obs, rff_q_obs, pcer_q_obs, fpi_q_obs, hours_obs, wage_obs. The model by Edge, Kiley and Laforte is estimated on: xgdp_q_obs, pgdp_q_obs, rff_q_obs, wage_obs, pecnn_q_obs, pecd_q_obs, per_q_obs, penr_q_obs, paipc_q_obs, paipk_q_obs, hoursedo_obs. Description of the .csv version of the data: There are two files for each data vintage. The files with the names "myyyymmdd.csv" contain the dataset from Faust and Wright. The files with the names "DSGEyyyymmdd.csv" contain the final transformed variables (all variables ending with _obs as described above) used for the estimation of the DSGE models. Dataset 2: Both versions contain one file for each data vintage containing data realizations for output growth, inflation, and the federal funds rate as realeased two quarters after the date they refer to. The Excel files contain not only these preliminary revised time series, but are also linked to the original excel sheets from the Philadelphia Fed real-time database also contained in each Excel file. The .csv file only contain the preliminary revised data series, but not how these have been constructed. Dataset 3: The Excel version contains one file with the Greenbook GNP/GDP and inflation projections and one file with the Greenbook assumptions regarding future values of the federal funds rate. In the csv version, the same information is contained in three files.