Error Statistics for the Survey of Professional Forecasters for Nonfarm Payroll Emp [QA] Release Date: 02/26/2013 Tom Stark Assistant Director and Manager Real-Time Data Research Center Economic Research Department Federal Reserve Bank of Philadelphia 1. OVERVIEW. This document reports error statistics for median projections from the Survey of Professional Forecasters (SPF), conducted since 1990 by the Federal Reserve Bank of Philadelphia. We provide the results in a series of tables and, in the PDF version of this document, a number of charts. The tables show the survey variable forecast and, importantly, the transformation of the data that we used to generate the statistics. (The transformation is usually a quarter-over-quarter growth rate, expressed in annualized percentage points. However, some variables, such as interest rates, the unemployment rate, and housing starts are untransformed and, thus, expressed in their natural units.) The paragraphs below explain the format of the tables and charts and the methods used to compute the statistics. These paragraphs are general. The same discussion applies to all variables in the survey. 2. DESCRIPTION OF TABLES. Tables 1A-1B report error statistics for various forecast horizons, sample periods, and choices of the real-time historical value that we used to assess accuracy. In each quarterly survey, we ask our panelists for their projections for the current quarter and the next four quarters. The current quarter is defined as the quarter in which we conducted the survey. Our tables provide error statistics separately for each quarter of this five-quarter horizon, beginning with the current quarter (denoted H = 1) and ending with the quarter that is four quarters in the future (H = 5). For each horizon, we report the mean forecast error [ME(S)], the mean absolute forecast error [MAE(S)], and the root-mean-square error [RMSE(S)]. All are standard measures of accuracy, though the academic literature generally places the most weight on the latter. We define a forecast error as the difference between the historical value and the forecast. The mean error for each horizon is simply the average of the forecast errors at that horizon, constructed over the sample periods shown in Table 1A. Other things the same, a forecast with a mean error close to zero is better than one with a mean error far from zero. The mean absolute error is the sample average of the absolute value of the errors. Many analysts prefer this measure to the mean error because it does not allow large positive errors to offset large negative errors. In this sense, the mean absolute error gives a cleaner estimate of the size of the errors. Decision makers, however, may care not only about the average size of the errors but also about their variability, as measured by variance. Our last measure of accuracy is one that reflects the influence of the mean error and the variance of the error. The root-mean-square error for the SPF [RMSE(S)], the measure most often used by analysts and academicians, is the square root of the the average squared error. The lower the root-mean-square error, the more accurate the forecast. 2.1. Benchmark Models. The forecast error statistics from the SPF are of interest in their own right. However, it is often more interesting to compare such statistics with those of alternative, or benchmark, forecasts. Tables 1A-1B report four such comparisons. They show the ratio of the root-mean-square error of the SPF forecast to that of four benchmark models. The benchmark models are statistical equations that we estimate on the data. We use the equations to generate projections for the same horizons included in the survey. In effect, we imagine standing back in time at each date when a survey was conducted and generating a separate forecast with each benchmark model. We do this in the same way that a survey panelist would have done using his own model. Table 1A reports the root-mean-square-error ratios using as many observations as possible for each model. The number of observations can differ from model to model. We first compute the RMSE for each model. We then construct the ratio. Table 1B reports RMSE ratios after we adjust the samples to include only the observations common to both models in the pair. Accordingly, the ratios reported in Table 1B may differ slightly from those of Table 1A, depending on the availability of sufficient real-time observations for estimating the benchmark models or for computing the errors of the SPF or benchmark forecasts. Table 1B also reports three two-sided p-values for each ratio. The p-values, corrected for the presence of heteroskedasticity and serial correlation in the time series of differences in squared forecast errors, are those for the test of equality of mean-square error between the SPF and the benchmark. The p-values are those for: (1) The Diebold-Mariano statistic (July 1995, Journal of Business and Economic Statistics), using a uniform lag window with the truncation lag set to the forecast horizon minus unity. When the uniform lag window produces a nonpositive standard error, the Bartlett window is used. (2) The Harvey-Leybourne-Newbold correction (1997, International Journal of Forecasting) to the Diebold-Mariano statistic. (3) The Diebold-Mariano statistic, using a Bartlett lag window with the truncation lag increased four quarters beyond that of (1) and (2). A RMSE ratio below unity indicates that the SPF consensus (median) forecast has a root-mean-square error lower than that of the benchmark. This means the SPF is more accurate. We now describe the benchmark models. The first is perhaps the simplest of all possible benchmarks: A no-change model. In this model, the forecast for quarter T, the one-step-ahead or current-quarter forecast, is simply the historical value for the prior quarter (T - 1). There is, in other words, no change in the forecast compared with the historical value. Moreover, the forecast for the remaining quarters of the horizon is the same as the forecast for the current quarter. We denote the relative RMSE ratio for this benchmark as RMSE(S/NC), using NC to indicate no change. The second and third benchmark models generate projections using one or more historical observations of the the variable forecast, weighted by coefficients estimated from the data. Such autoregressive (AR) models can be formulated in two ways. We can estimate one model to generate the forecasts at all horizons, using an iteration method to generate the projections beyond the current quarter (IAR), or we can directly estimate a new model for each forecast horizon (DAR). The latter formulation has been shown to reduce the bias in a forecast when the underlying model is characterized by certain types of misspecification. The root-mean-square error ratios are denoted RMSE(S/IAR) and RMSE(S/DAR), respectively. The one- through five-step-ahead projections of the benchmark models use information on the quarterly average of the variable forecast. The latest historical observation is for the quarter that is one quarter before the quarter of the first projection in the horizon. In contrast, the panelists generate their projections with the help of additional information. They submit their projections near the middle of each quarter and hence have access to some monthly indicators for the first month of each quarter, when those data are released before the survey deadline. This puts the projections of panelists for some variables at an advantage relative to the corresponding benchmark projections. Moreover, the panelists may also examine the very recent historical values of such monthly indicators in forming their projections for quarterly averages. Such monthly statistical momentum represents an advantage not shared by the benchmark models, which use only quarterly averages. For survey variables whose observations are reported at a monthly frequency, such as interest rates, industrial production, housing starts, and unemployment, we estimate and forecast a fourth benchmark model, the DARM. This model adds recent monthly historical values to the specification of the DAR model. For the projections for unemployment, nonfarm payroll employment, and interest rates, we add the values of monthly observations, beginning with that for the first month of the first quarter of the forecast horizon. These values should be in the information set of the survey panelists at the time they formed their projections. In contrast, for variables such as housing starts and industrial production, we include only lagged values of monthly observations. For such variables, the panelists would not have known the monthly observation for the first month of the first quarter of the forecast horizon. In general, we find that adding monthly observations to the benchmark DAR models improves accuracy. Indeed, for the projections for interest rates and the unemployment rate, the accuracy of the benchmark DARM projections rivals that of the SPF projections. 2.2. Real-Time Data. All benchmark models are estimated on a rolling, fixed window of 60 real-time quarterly observations. Lag lengths, based on either the Akaike information criterion (AIC) or the Schwarz information criterion (SIC), are re-estimated each period. The tables below indicate whether the lag length was was chosen by the AIC or SIC. We would like to make the comparison between the SPF forecast and the forecasts of each benchmark as fair as possible. Therefore, we must subject the benchmark models to the same data environment the survey panelists faced when they made their projections. This is important because macroeconomic data are revised often, and we do not want the benchmark models to use a data set that differs from the one our panelists would have used. We estimate and forecast the benchmark models with real-time data from the Philadelphia Fed real-time data set, using the vintage of data that the survey panelists would have had at the time they generated their own projections. (For more information on the Philadelphia Fed real-time data set, go to www.philadelphiafed.org/econ/forecast/real-time-data/.) An open question in the literature on forecasting is: What version or vintage of the data should we use to compute the errors? A closely related question is: What version of the data are professional forecasters trying to predict? Our computations take no strong position on these questions. In Tables 1A - 1B, we evaluate the projections (SPF and benchmark) with five alternative measures of the historical values, all from the Philadelphia Fed real-time data set. These measures range from the initial-release values to the values as we know them today. All together, we compute the forecast error statistics using the following five alternative measures of historical values: (1) The initial or first-release value; (2) The revised value as it appears one quarter after the initial release; (3) The revised value as it appears five quarters after the initial release; (4) The revised value as it appears nine quarters after the initial release; (5) The revised value as it appears today. Each measure of the historical value has advantages and disadvantages. The initial-release value is the first measure released by government statistical agencies. A forecaster might be very interested in this measure because it enables him to evaluate his latest forecast soon after he generated it. However, early releases of the data are often subject to large measurement error. Subsequent releases [(2) - (5)] are more accurate, but they are available much later than the initial release. As we go from the first measure to the fifth, we get more reliability, at the cost of higher delays in availability. The last two columns in Table 1A report the number of observations that we used to compute the error statistics. Some observations are omitted because the data are missing in the real-time data set, such as occurred when federal government statistical agencies closed in late 1995. 2.3. Recent Projections and Realizations. Tables 2 to 7 provide information on recent projections and realizations. They show how we align the data prior to computing the forecast errors that form the backbone of the computations in Tables 1A - 1B. Any error can be written as the equation given by error = realization - forecast. For our computations, we must be more precise because, for each projection (SPF and benchmarks), we have different periods forecast (T) different forecast horizons (h), and several measures of the realization (m). Thus, we can define the forecast error more precisely as error( T, h, m ) = realization( T, m ) - forecast( T, h ). Tables 2 to 7 are organized along these lines. Table 2 shows recent forecasts from the SPF. Each column gives the projection for a different horizon or forecast step (h), beginning with that for the current quarter, defined as the quarter in which we conducted the survey. The dates (T) given in the rows show the periods forecast. These also correspond to the dates that we conducted the survey. Tables 3 to 6 report the recent projections of the four benchmark models. They are organized in the same way as Table 2. Table 7 reports recent values of the five alternative realizations (m) we use to compute the error statistics. 2.4. Qualifications. We note two minor qualifications to the methods discussed above. The first concerns the vintage of data that we used to estimate and forecast the benchmark models for CPI inflation. The second concerns the five measures of realizations used for the unemployment rate, nonfarm payroll employment, and CPI inflation. To estimate and forecast the benchmark models for CPI inflation, we use the vintage of data that would have been available in the middle of each quarter. This postdates by one month the vintage that SPF panelists would have had at their disposal when they formed their projections. To compute the realizations for unemployment, nonfarm payroll employment, and CPI inflation, we use the vintages associated with the middle of each quarter. The measure that we call initial comes from this vintage, even though the initial estimate was available in the vintage dated one month earlier. Thus, for these variables, our initial estimate reflects some revision by government statistical agencies. The effect for unemployment and CPI inflation is likely small. The effect could be somewhat larger for nonfarm payroll employment. 3. DESCRIPTION OF GRAPHS. 3.1. Root-Mean-Square Errors. For each sample period shown in Table 1, we provide graphs of the root-mean-square error for the SPF forecast. There is one page for each sample period. On each page, we plot (for each forecast horizon) the RMSE on the y-axis. The x-axis shows the measure of the historical value that we used to compute the RMSE. These range from the value on its initial release to the value one quarter later to the value as we know it now (at the time we made the computation). The graphs provide a tremendous amount of information. If we focus on a particular graph, we can see how a change in the measure of the realization (x-axis) affects the root-mean-square-error measure of accuracy. The effect is pronounced for some variables, such as real GDP and some of its components. For others, there is little or no effect. For example, because the historical data on interest rates are not revised, the estimated RMSE is the same in each case. If we compare a particular point on one graph with the same point on another, we see how the forecast horizon affects accuracy. In general, the RMSE rises (accuracy falls) as the forecast horizon lengthens. Finally, if we compare a graph on one page with the corresponding graph on another page, we see how our estimates of accuracy in the SPF change with the sample period. Periods characterized by a high degree of economic turbulence will generally produce large RMSEs. 3.2. Fan Charts. The last chart plots recent historical values and the latest SPF forecast. It also shows confidence intervals for the forecast, based on back-of-the-envelope calculations. The historical values and the SPF forecast are those associated with the latest vintage of data and survey, respectively, available at the time we ran our computer programs. The confidence intervals are constructed under the assumption that the historical forecast errors over the sample (shown in the footnote) follow a normal distribution with a mean of zero and a variance given by the squared root-mean-square error. The latter is estimated over the aforementioned sample, using the measure of history listed in the footnote. --------------------------------------------------------------------------------------- Table 1A. Forecast Error Summary Statistics for SPF Variable: EMP (Nonfarm Payroll Emp [QA]) --------------------------------------------------------------------------------------- Computed Over Various Sample Periods Various Measures of Realizations Transformation: Q/Q Growth Rate Lag Length for IAR(p), DAR(p), and DARM(p) Models: AIC Last Updated: 02/26/2013 16:52 --------------------------------------------------------------------------------------- H ME(S) MAE(S) RMSE(S) RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) Nspf N History: Initial Release 2004:04-2010:03 1 -0.13 0.33 0.50 0.42 0.41 0.41 0.87 24 24 2 -0.56 0.78 1.25 0.68 0.67 0.63 0.72 24 24 3 -0.78 1.00 1.72 0.74 0.69 0.71 0.79 24 24 4 -1.05 1.25 2.12 0.78 0.72 0.80 0.88 24 24 5 -1.12 1.31 2.23 0.76 0.68 0.83 0.86 24 24 H ME(S) MAE(S) RMSE(S) RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) Nspf N History: One Qtr After Initial Release 2004:04-2010:03 1 -0.18 0.39 0.56 0.47 0.46 0.46 0.93 24 24 2 -0.60 0.83 1.28 0.70 0.69 0.65 0.74 24 24 3 -0.82 1.06 1.76 0.75 0.70 0.72 0.80 24 24 4 -1.10 1.29 2.17 0.79 0.73 0.81 0.88 24 24 5 -1.16 1.37 2.29 0.77 0.69 0.83 0.87 24 24 H ME(S) MAE(S) RMSE(S) RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) Nspf N History: Five Qtrs After Initial Release 2004:04-2010:03 1 -0.33 0.58 0.89 0.65 0.63 0.63 0.99 24 24 2 -0.75 0.96 1.56 0.78 0.77 0.73 0.82 24 24 3 -0.97 1.25 2.05 0.82 0.77 0.78 0.86 24 24 4 -1.25 1.52 2.45 0.84 0.78 0.84 0.92 24 24 5 -1.31 1.58 2.57 0.82 0.73 0.86 0.89 24 24 H ME(S) MAE(S) RMSE(S) RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) Nspf N History: Nine Qtrs After Initial Release 2004:04-2010:03 1 -0.34 0.59 0.88 0.65 0.63 0.63 1.01 24 24 2 -0.76 1.03 1.59 0.78 0.77 0.74 0.83 24 24 3 -0.98 1.29 2.08 0.82 0.77 0.78 0.86 24 24 4 -1.26 1.56 2.48 0.85 0.79 0.85 0.92 24 24 5 -1.32 1.62 2.60 0.82 0.74 0.86 0.90 24 24 H ME(S) MAE(S) RMSE(S) RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) Nspf N History: Latest Vintage 2004:04-2010:03 1 -0.33 0.57 0.87 0.65 0.63 0.63 1.01 24 24 2 -0.76 1.02 1.59 0.79 0.77 0.73 0.82 24 24 3 -0.98 1.30 2.09 0.82 0.77 0.79 0.86 24 24 4 -1.26 1.55 2.49 0.86 0.79 0.85 0.93 24 24 5 -1.32 1.61 2.61 0.82 0.74 0.87 0.90 24 24 Notes for Table 1A. (1) The forecast horizon is given by H, where H = 1 is the SPF forecast for the current quarter. (2) The headers ME(S), MAE(S), and RMSE(S) are mean error, mean absolute error, and root-mean-square error for the SPF. (3) The header RMSE(S/NC) is the ratio of the SPF RMSE to that of the no-change (NC) model. (4) The headers RMSE(S/IAR), RMSE(S/DAR) and RMSE(S/DARM) are the ratios of the SPF RMSE to the RMSE of the iterated and direct autoregressive models and the direct autoregressive model augmented with monthly observations, respectively. All models are estimated on a rolling window of 60 observations from the Philadelphia Fed real-time data set. (5) The headers Nspf and N are the number of observations analyzed for the SPF and benchmark models. (6) When the variable forecast is a growth rate or an interest rate, it is expressed in annualized percentage points. When the variable forecast is the unemployment rate, it is expressed in percentage points. (7) Sample periods refer to the dates forecast, not the dates when the forecasts were made. Source: Tom Stark, Research Department, FRB Philadelphia. ---------------------------------------------------------------------------------------------- Table 1B. Ratios of Root-Mean-Square Errors for SPF Variable: EMP (Nonfarm Payroll Emp [QA]) Alternative P-Values in Parentheses ---------------------------------------------------------------------------------------------- Computed Over Various Sample Periods Various Measures of Realizations Transformation: Q/Q Growth Rate Lag Length for IAR(p), DAR(p), and DARM(p) Models: AIC Last Updated: 02/26/2013 16:52 ---------------------------------------------------------------------------------------------- History: Initial Release 2004:04-2010:03 H RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) N1 N2 N3 N4 1 0.419 0.411 0.411 0.867 24 24 24 24 (0.001) (0.009) (0.009) (0.042) (0.003) (0.018) (0.018) (0.058) (0.047) (0.046) (0.046) (0.061) 2 0.676 0.670 0.626 0.723 24 24 24 24 (0.071) (0.057) (0.069) (0.019) (0.104) (0.087) (0.101) (0.039) (0.143) (0.126) (0.125) (0.076) 3 0.736 0.693 0.714 0.791 24 24 24 24 (0.307) (0.234) (0.165) (0.218) (0.370) (0.297) (0.226) (0.282) (0.242) (0.190) (0.148) (0.173) 4 0.782 0.719 0.800 0.881 24 24 24 24 (0.373) (0.263) (0.176) (0.258) (0.455) (0.349) (0.259) (0.344) (0.316) (0.242) (0.162) (0.236) 5 0.762 0.677 0.828 0.864 24 24 24 24 (0.254) (0.215) (0.081) (0.104) (0.363) (0.324) (0.169) (0.199) (0.262) (0.227) (0.112) (0.137) History: One Qtr After Initial Release 2004:04-2010:03 H RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) N1 N2 N3 N4 1 0.470 0.461 0.461 0.929 24 24 24 24 (0.001) (0.015) (0.015) (0.234) (0.004) (0.025) (0.025) (0.256) (0.045) (0.047) (0.047) (0.058) 2 0.697 0.694 0.648 0.744 24 24 24 24 (0.078) (0.052) (0.065) (0.017) (0.112) (0.081) (0.097) (0.035) (0.145) (0.121) (0.121) (0.069) 3 0.746 0.701 0.721 0.803 24 24 24 24 (0.309) (0.230) (0.161) (0.229) (0.372) (0.294) (0.222) (0.293) (0.241) (0.187) (0.143) (0.179) 4 0.791 0.728 0.806 0.884 24 24 24 24 (0.375) (0.261) (0.170) (0.236) (0.456) (0.347) (0.253) (0.322) (0.319) (0.240) (0.158) (0.220) 5 0.771 0.686 0.833 0.869 24 24 24 24 (0.253) (0.213) (0.076) (0.099) (0.363) (0.322) (0.162) (0.193) (0.265) (0.226) (0.108) (0.135) History: Five Qtrs After Initial Release 2004:04-2010:03 H RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) N1 N2 N3 N4 1 0.649 0.628 0.628 0.987 24 24 24 24 (0.005) (0.021) (0.021) (0.526) (0.011) (0.033) (0.033) (0.541) (0.037) (0.045) (0.045) (0.409) 2 0.782 0.772 0.732 0.818 24 24 24 24 (0.046) (0.030) (0.047) (0.002) (0.074) (0.053) (0.076) (0.008) (0.124) (0.102) (0.104) (0.036) 3 0.817 0.770 0.784 0.860 24 24 24 24 (0.328) (0.213) (0.128) (0.206) (0.390) (0.276) (0.186) (0.269) (0.251) (0.167) (0.115) (0.155) 4 0.845 0.782 0.844 0.924 24 24 24 24 (0.408) (0.250) (0.137) (0.263) (0.487) (0.337) (0.217) (0.349) (0.342) (0.228) (0.127) (0.237) 5 0.817 0.731 0.861 0.895 24 24 24 24 (0.259) (0.198) (0.048) (0.049) (0.369) (0.307) (0.122) (0.124) (0.275) (0.216) (0.086) (0.096) History: Nine Qtrs After Initial Release 2004:04-2010:03 H RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) N1 N2 N3 N4 1 0.655 0.635 0.635 1.011 24 24 24 24 (0.009) (0.028) (0.028) (0.714) (0.017) (0.042) (0.042) (0.723) (0.038) (0.045) (0.045) (0.726) 2 0.784 0.773 0.735 0.826 24 24 24 24 (0.044) (0.026) (0.050) (0.003) (0.071) (0.048) (0.079) (0.010) (0.121) (0.093) (0.107) (0.043) 3 0.818 0.770 0.784 0.864 24 24 24 24 (0.319) (0.207) (0.120) (0.219) (0.381) (0.270) (0.178) (0.283) (0.241) (0.161) (0.108) (0.168) 4 0.852 0.789 0.849 0.925 24 24 24 24 (0.423) (0.257) (0.140) (0.259) (0.501) (0.343) (0.220) (0.345) (0.355) (0.234) (0.128) (0.229) 5 0.822 0.736 0.864 0.897 24 24 24 24 (0.258) (0.199) (0.049) (0.047) (0.368) (0.307) (0.124) (0.121) (0.277) (0.217) (0.089) (0.093) History: Latest Vintage 2004:04-2010:03 H RMSE(S/NC) RMSE(S/IAR) RMSE(S/DAR) RMSE(S/DARM) N1 N2 N3 N4 1 0.651 0.630 0.630 1.015 24 24 24 24 (0.008) (0.026) (0.026) (0.649) (0.016) (0.040) (0.040) (0.660) (0.032) (0.040) (0.040) (0.661) 2 0.785 0.772 0.733 0.824 24 24 24 24 (0.042) (0.026) (0.045) (0.002) (0.069) (0.048) (0.073) (0.008) (0.119) (0.091) (0.098) (0.034) 3 0.820 0.772 0.786 0.863 24 24 24 24 (0.320) (0.209) (0.121) (0.206) (0.382) (0.272) (0.178) (0.269) (0.242) (0.163) (0.108) (0.156) 4 0.855 0.792 0.851 0.926 24 24 24 24 (0.431) (0.261) (0.139) (0.252) (0.508) (0.347) (0.219) (0.338) (0.362) (0.237) (0.127) (0.221) 5 0.825 0.739 0.866 0.899 24 24 24 24 (0.259) (0.197) (0.045) (0.045) (0.369) (0.306) (0.117) (0.117) (0.279) (0.217) (0.084) (0.091) Notes for Table 1B. (1) The forecast horizon is given by H, where H = 1 is the SPF forecast for the current quarter. (2) The headers RMSE(S/NC), RMSE(S/IAR), RMSE(S/DAR), and RMSE(S/DARM) are the ratios of the SPF root-mean-square error to that of the benchmark models: No-change (NC), indirect autoregression (IAR), direct autoregession (DAR), and direct autoregression augmented with monthly information (DARM). These statistics may differ slightly from those reported in Table 1A because they incorporate only those observations common to both the SPF and the benchmark model. The previous statistics make use of all available observations for each model. (3) All models are estimated on a rolling window of 60 observations from the Philadelphia Fed real-time data set. (4) A set of three two-sided p-values (in parentheses) accompanies each statistic. These are the p-values for the test of the equality of mean-square-error. The first is for the Diebold-Mariano (1995, JBES) statistic, using a uniform lag window with the trunction lag set to the forecast horizon minus one. (The tables report the p-values using a Bartlett window when the uniform window produces a negative standard error.) The second is for the Harvey-Leybourne-Newbold (1997, IJF) correction to the Diebold-Mariano statistic. The third is for the Diebold-Mariano statistic, using a Bartlett lag window with the truncation lag increased four quarters. (5) The headers N1, N2, N3, and N4 show the number of observations used in constructing each ratio of root-mean-square errors. (6) Sample periods refer to the dates forecast, not the dates when the forecasts were made. Source: Tom Stark, Research Department, FRB Philadelphia. -------------------------------------------------------------------- Table 2. Recent SPF Forecasts - Dated at the Quarter Forecast -------------------------------------------------------------------- Variable: EMP (Nonfarm Payroll Emp [QA]) By Forecast Step (1 to 5) Transformation: Q/Q Growth Rate Last Updated: 02/26/2013 16:52 -------------------------------------------------------------------- Qtr Forecast Step 1 Step 2 Step 3 Step 4 Step 5 2006:02 1.523 1.599 1.647 1.498 1.713 2006:03 1.118 1.476 1.406 1.629 1.319 2006:04 1.092 1.132 1.300 1.250 1.307 2007:01 1.215 1.039 1.050 1.321 1.339 2007:02 1.000 1.026 0.878 1.118 1.098 2007:03 1.076 1.030 1.056 1.048 1.202 2007:04 1.003 0.997 0.919 1.122 1.039 2008:01 0.342 0.893 0.989 1.122 1.182 2008:02 -0.391 0.395 0.724 0.993 1.122 2008:03 -0.404 -0.042 0.692 0.990 1.047 2008:04 -1.928 -0.396 0.231 0.805 1.158 2009:01 -4.751 -1.906 -0.260 0.484 0.738 2009:02 -4.600 -2.750 -0.952 0.507 0.692 2009:03 -2.458 -2.541 -1.805 -0.064 0.552 2009:04 -1.451 -0.739 -0.954 -0.388 0.175 2010:01 0.006 -0.321 0.473 0.182 0.351 2010:02 1.933 1.082 0.530 0.563 0.949 2010:03 0.074 1.115 0.635 1.463 0.832 2010:04 0.802 1.054 1.416 1.120 1.307 2011:01 1.191 1.056 1.469 1.972 1.312 2011:02 1.769 1.736 1.339 1.755 1.994 2011:03 0.975 1.792 1.846 1.279 1.740 2011:04 1.059 1.359 1.594 1.949 1.464 2012:01 1.467 1.108 1.653 2.006 1.832 2012:02 1.500 1.294 1.154 1.258 1.654 2012:03 1.133 1.541 1.320 1.389 1.702 2012:04 1.344 1.223 1.559 1.466 1.146 2013:01 1.490 1.147 1.367 1.531 1.408 2013:02 NA 1.384 1.313 1.254 1.665 2013:03 NA NA 1.539 1.525 1.334 2013:04 NA NA NA 1.609 1.593 2014:01 NA NA NA NA 1.523 Notes for Table 2. (1) Each column gives the sequence of SPF projections for a given forecast step. The forecast steps range from one (the forecast for the quarter in which the survey was conducted) to four quarters in the future (step 5). (2) The dates listed in the rows are the dates forecast, not the dates when the forecasts were made, with the exception of the forecast at step one, for which the two dates coincide. Source: Tom Stark, Research Department, FRB Philadelphia. ------------------------------------------------------------------------------------ Table 3. Recent Benchmark Model 1 IAR Forecasts - Dated at the Quarter Forecast ------------------------------------------------------------------------------------ Variable: EMP (Nonfarm Payroll Emp [QA]) By Forecast Step (1 to 5) Transformation: Q/Q Growth Rate Lag Length for IAR(p): AIC Last Updated: 02/26/2013 16:52 ------------------------------------------------------------------------------------ Qtr Forecast Step 1 Step 2 Step 3 Step 4 Step 5 2006:02 1.590 1.248 1.622 1.641 1.354 2006:03 1.432 1.742 1.329 1.591 1.574 2006:04 1.234 1.334 1.696 1.360 1.598 2007:01 1.515 1.319 1.460 1.741 1.425 2007:02 1.477 1.510 1.321 1.465 1.721 2007:03 1.317 1.497 1.538 1.378 1.536 2007:04 0.903 1.300 1.531 1.550 1.402 2008:01 0.871 0.883 1.354 1.559 1.571 2008:02 -0.052 0.976 0.989 1.379 1.589 2008:03 -0.706 -0.240 1.054 1.055 1.424 2008:04 -0.807 -0.567 0.022 1.141 1.151 2009:01 -3.481 -0.701 -0.416 0.092 1.208 2009:02 -6.434 -3.387 -0.469 -0.207 0.295 2009:03 -4.420 -6.891 -3.297 -0.278 -0.020 2009:04 -2.461 -4.342 -7.367 -3.211 -0.071 2010:01 -0.116 -2.275 -4.267 -7.862 -3.128 2010:02 1.203 0.761 -2.103 -4.195 -8.376 2010:03 2.941 1.836 1.306 -1.946 -4.126 2010:04 -0.013 3.464 2.117 1.585 -1.803 2011:01 0.205 -0.433 3.474 2.157 1.672 2011:02 1.547 0.468 -0.309 3.246 2.029 2011:03 1.470 1.579 0.376 -0.235 2.829 2011:04 0.742 1.361 1.493 0.473 -0.055 2012:01 1.099 0.552 1.215 1.360 0.479 2012:02 2.112 1.089 0.490 1.060 1.214 2012:03 0.913 2.034 0.934 0.441 0.921 2012:04 0.892 0.555 1.778 0.849 0.431 2013:01 1.429 0.843 0.469 1.526 0.752 2013:02 NA 1.325 0.689 0.372 1.276 2013:03 NA NA 1.106 0.626 0.354 2013:04 NA NA NA 0.932 0.560 2014:01 NA NA NA NA 0.767 Notes for Table 3. (1) Each column gives the sequence of benchmark IAR projections for a given forecast step. The forecast steps range from one to five. The first step corresponds to the forecast that SPF panelists make for the quarter in which the survey is conducted. (2) The dates listed in the rows are the dates forecast, not the dates when the forecasts were made, with the exception of the forecast at step one, for which the two dates coincide. (3) The IAR benchmark model is estimated on a fixed 60-quarter rolling window. Its forecasts are computed with the indirect method. Estimation uses data from the Philadelphia Fed real-time data set. Source: Tom Stark, Research Department, FRB Philadelphia. --------------------------------------------------------------------------------------- Table 4. Recent Benchmark Model 2 No-Change Forecasts - Dated at the Quarter Forecast --------------------------------------------------------------------------------------- Variable: EMP (Nonfarm Payroll Emp [QA]) By Forecast Step (1 to 5) Transformation: Q/Q Growth Rate Last Updated: 02/26/2013 16:52 --------------------------------------------------------------------------------------- Qtr Forecast Step 1 Step 2 Step 3 Step 4 Step 5 2006:02 1.706 1.232 1.603 1.858 1.581 2006:03 1.203 1.706 1.232 1.603 1.858 2006:04 1.334 1.203 1.706 1.232 1.603 2007:01 1.481 1.334 1.203 1.706 1.232 2007:02 1.460 1.481 1.334 1.203 1.706 2007:03 1.275 1.460 1.481 1.334 1.203 2007:04 0.861 1.275 1.460 1.481 1.334 2008:01 0.833 0.861 1.275 1.460 1.481 2008:02 -0.321 0.833 0.861 1.275 1.460 2008:03 -0.584 -0.321 0.833 0.861 1.275 2008:04 -0.951 -0.584 -0.321 0.833 0.861 2009:01 -3.579 -0.951 -0.584 -0.321 0.833 2009:02 -5.993 -3.579 -0.951 -0.584 -0.321 2009:03 -4.502 -5.993 -3.579 -0.951 -0.584 2009:04 -2.665 -4.502 -5.993 -3.579 -0.951 2010:01 -1.382 -2.665 -4.502 -5.993 -3.579 2010:02 0.202 -1.382 -2.665 -4.502 -5.993 2010:03 2.111 0.202 -1.382 -2.665 -4.502 2010:04 -0.162 2.111 0.202 -1.382 -2.665 2011:01 0.724 -0.162 2.111 0.202 -1.382 2011:02 1.330 0.724 -0.162 2.111 0.202 2011:03 1.464 1.330 0.724 -0.162 2.111 2011:04 0.858 1.464 1.330 0.724 -0.162 2012:01 1.367 0.858 1.464 1.330 0.724 2012:02 2.138 1.367 0.858 1.464 1.330 2012:03 0.996 2.138 1.367 0.858 1.464 2012:04 1.261 0.996 2.138 1.367 0.858 2013:01 1.614 1.261 0.996 2.138 1.367 2013:02 NA 1.614 1.261 0.996 2.138 2013:03 NA NA 1.614 1.261 0.996 2013:04 NA NA NA 1.614 1.261 2014:01 NA NA NA NA 1.614 Notes for Table 4. (1) Each column gives the sequence of benchmark no-change projections for a given forecast step. The forecast steps range from one to five. The first step corresponds to the forecast that SPF panelists make for the quarter in which the survey is conducted. (2) The dates listed in the rows are the dates forecast, not the dates when the forecasts were made, with the exception of the forecast at step one, for which the two dates coincide. (3) The projections use data from the Philadelphia Fed real-time data set. Source: Tom Stark, Research Department, FRB Philadelphia. ------------------------------------------------------------------------------------ Table 5. Recent Benchmark Model 3 DAR Forecasts - Dated at the Quarter Forecast ------------------------------------------------------------------------------------ Variable: EMP (Nonfarm Payroll Emp [QA]) By Forecast Step (1 to 5) Transformation: Q/Q Growth Rate Lag Length for DAR(p): AIC Last Updated: 02/26/2013 16:52 ------------------------------------------------------------------------------------ Qtr Forecast Step 1 Step 2 Step 3 Step 4 Step 5 2006:02 1.590 1.546 1.520 1.744 1.598 2006:03 1.432 1.512 1.344 1.476 1.496 2006:04 1.234 1.467 1.701 1.347 1.520 2007:01 1.515 1.198 1.409 1.646 1.263 2007:02 1.477 1.365 1.484 1.443 1.609 2007:03 1.317 1.649 1.587 1.514 1.455 2007:04 0.903 0.982 1.562 1.618 1.529 2008:01 0.871 0.818 1.431 1.600 1.633 2008:02 -0.052 0.919 1.143 1.483 1.627 2008:03 -0.706 -0.371 1.073 1.257 1.533 2008:04 -0.807 -0.637 0.230 1.196 1.333 2009:01 -3.481 -0.675 -0.035 0.485 1.275 2009:02 -6.434 -3.250 -0.381 0.260 0.684 2009:03 -4.420 -7.196 -2.665 -0.067 0.452 2009:04 -2.461 -5.051 -5.519 -1.914 0.165 2010:01 -0.116 -2.407 -4.185 -4.220 -1.604 2010:02 1.203 -0.162 -2.647 -3.694 -3.335 2010:03 2.941 1.821 -1.035 -2.369 -2.904 2010:04 -0.013 3.407 1.181 -1.120 -1.041 2011:01 0.205 -0.507 2.750 0.244 1.344 2011:02 1.547 0.468 -1.176 2.028 0.145 2011:03 1.470 1.687 1.288 -0.558 1.101 2011:04 0.742 1.433 1.323 1.069 0.344 2012:01 1.099 0.576 1.147 1.167 0.631 2012:02 2.112 1.015 0.445 0.989 0.835 2012:03 0.913 2.045 1.196 0.459 0.910 2012:04 0.892 0.653 1.735 1.060 0.656 2013:01 1.429 0.711 0.229 1.447 0.789 2013:02 NA 1.311 0.688 0.218 1.180 2013:03 NA NA 1.143 0.898 0.258 2013:04 NA NA NA 1.041 0.676 2014:01 NA NA NA NA 0.383 Notes for Table 5. (1) Each column gives the sequence of benchmark DAR projections for a given forecast step. The forecast steps range from one to five. The first step corresponds to the forecast that SPF panelists make for the quarter in which the survey is conducted. (2) The dates listed in the rows are the dates forecast, not the dates when the forecasts were made, with the exception of the forecast at step one, for which the two dates coincide. (3) The DAR benchmark model is estimated on a fixed 60-quarter rolling window. Its forecasts are computed with the direct method. Estimation uses data from the Philadelphia Fed real-time data set. Source: Tom Stark, Research Department, FRB Philadelphia. ------------------------------------------------------------------------------------ Table 6. Recent Benchmark Model 4 DARM Forecasts - Dated at the Quarter Forecast ------------------------------------------------------------------------------------ Variable: EMP (Nonfarm Payroll Emp [QA]) By Forecast Step (1 to 5) Transformation: Q/Q Growth Rate Lag Length for DARM(p): AIC Last Updated: 02/26/2013 16:52 ------------------------------------------------------------------------------------ Qtr Forecast Step 1 Step 2 Step 3 Step 4 Step 5 2006:02 1.588 1.584 0.385 1.584 2.028 2006:03 1.081 1.525 1.514 0.760 1.496 2006:04 1.261 1.321 1.469 1.804 1.217 2007:01 1.465 1.105 1.265 1.458 1.645 2007:02 1.153 1.442 1.231 1.302 1.399 2007:03 1.136 1.435 1.311 1.362 1.352 2007:04 1.075 0.903 1.136 1.426 1.437 2008:01 0.404 1.137 1.144 1.188 1.436 2008:02 -0.324 0.679 1.363 1.274 1.286 2008:03 -0.496 -0.131 0.470 1.430 1.391 2008:04 -1.847 -0.333 0.113 0.609 1.489 2009:01 -4.806 -1.291 -0.106 0.336 0.798 2009:02 -5.438 -4.058 -1.245 0.202 0.604 2009:03 -3.238 -5.983 -4.035 -0.827 0.412 2009:04 -1.656 -3.782 -5.753 -3.259 -0.248 2010:01 -0.335 -1.899 -2.375 -3.960 -2.471 2010:02 2.111 0.331 -1.993 -2.324 -3.015 2010:03 -0.481 2.425 0.083 -0.964 -1.678 2010:04 0.555 0.809 2.743 -0.048 -1.439 2011:01 0.719 0.003 0.298 1.842 -0.143 2011:02 2.187 0.844 0.321 -0.260 0.909 2011:03 0.667 1.999 0.853 1.306 0.618 2011:04 0.999 0.806 1.971 0.526 1.078 2012:01 1.921 0.567 0.888 1.803 0.493 2012:02 1.304 1.767 0.391 0.430 1.402 2012:03 1.004 1.374 1.757 0.596 0.587 2012:04 1.478 0.727 1.050 1.446 0.549 2013:01 1.646 1.341 0.665 0.647 1.166 2013:02 NA 1.450 1.195 0.698 0.586 2013:03 NA NA 1.104 1.162 0.663 2013:04 NA NA NA 1.011 0.916 2014:01 NA NA NA NA 0.773 Notes for Table 6. (1) Each column gives the sequence of benchmark DARM projections for a given forecast step. The forecast steps range from one to five. The first step corresponds to the forecast that SPF panelists make for the quarter in which the survey is conducted. (2) The dates listed in the rows are the dates forecast, not the dates when the forecasts were made, with the exception of the forecast at step one, for which the two dates coincide. (3) The DARM benchmark model is estimated on a fixed 60-quarter rolling window. Its forecasts are computed with the direct method and incorporate recent monthly values of the dependent variable. Estimation uses data from the Philadelphia Fed real-time data set. Source: Tom Stark, Research Department, FRB Philadelphia. ------------------------------------------------------------ Table 7. Recent Realizations (Various Measures) Source: Philadelphia Fed Real-Time Data Set ------------------------------------------------------------ Variable: EMP (Nonfarm Payroll Emp [QA]) Transformation: Q/Q Growth Rate Last Updated: 02/26/2013 16:52 1- Initial Release 2- One Qtr After Initial Release 3- Five Qtrs After Initial Release 4- Nine Qtrs After Initial Release 5- Latest Vintage ------------------------------------------------------------- Obs. Date (1) (2) (3) (4) (5) 2006:02 1.203 1.213 1.545 1.356 1.526 2006:03 1.334 1.566 1.523 1.150 1.309 2006:04 1.481 1.501 1.280 1.406 1.047 2007:01 1.460 1.458 1.160 1.465 1.580 2007:02 1.275 1.219 0.925 0.853 1.026 2007:03 0.861 0.755 0.110 0.106 0.218 2007:04 0.833 0.793 1.185 0.599 0.682 2008:01 -0.321 -0.328 -0.165 0.102 0.087 2008:02 -0.584 -0.631 -1.254 -1.173 -1.374 2008:03 -0.951 -1.460 -2.281 -2.165 -2.089 2008:04 -3.579 -3.677 -4.827 -4.681 -4.653 2009:01 -5.993 -5.949 -6.375 -6.497 -6.529 2009:02 -4.502 -4.521 -5.027 -5.478 -5.543 2009:03 -2.665 -3.050 -3.008 -3.023 -3.020 2009:04 -1.382 -1.341 -1.638 -1.520 -1.477 2010:01 0.202 0.179 -0.319 -0.396 -0.438 2010:02 2.111 2.159 1.978 2.000 1.938 2010:03 -0.162 -0.226 -0.124 -0.117 -0.117 2010:04 0.724 0.756 0.944 NA 0.972 2011:01 1.330 1.302 1.417 NA 1.170 2011:02 1.464 1.438 1.701 NA 2.012 2011:03 0.858 0.899 1.281 NA 1.281 2011:04 1.367 1.388 NA NA 1.689 2012:01 2.138 2.127 NA NA 2.353 2012:02 0.996 0.976 NA NA 1.364 2012:03 1.261 1.230 NA NA 1.230 2012:04 1.614 NA NA NA 1.614 2013:01 NA NA NA NA NA 2013:02 NA NA NA NA NA 2013:03 NA NA NA NA NA 2013:04 NA NA NA NA NA 2014:01 NA NA NA NA NA Notes for Table 7. (1) Each column reports a sequence of realizations from the Philadelphia Fed real-time data set. (2) The date listed in each row is the observation date. (3) Moving across a particular row shows how the observation is revised in subsequent releases. Source: Tom Stark, Research Department, FRB Philadelphia.