The Federal Reserve Bank of Minneapolis is no longer maintaining a MF-VAR model. The June 3, 2016 forecast is the final update. Model documentation, files, and forecasts are available below.
The Minneapolis Mixed Frequency Vector Autoregression (MF-VAR) model of the U.S. economy is the most recent in a line of statistical forecasting models that began with research conducted at the Federal Reserve Bank of Minneapolis and the University of Minnesota in the 1970s and early 1980s. Rising computer power led to resurgent interest in these models in the 1990s, which in turn brought about improvements along several dimensions. The current model reflects many of these advances. A noteworthy feature of the model is that it combines data measured at both monthly and quarterly frequencies so that forecasts reflect up-to-date monthly information as it becomes available. Please note that the forecasts produced by the MF-VAR model do not represent official forecasts of the Minneapolis Fed, its president, the Federal Reserve System, or the FOMC.
Creator: Beauchemin, Kenneth Ronald Abstract:
This memo describes a revision to the mixed-frequency vector autoregression (MF-VAR) model originally constructed by Schorfheide and Song (2012) and subsequently revised by Beauchemin (2013). In this most recent version, the 14-variable model is expanded to include nonfarm payroll employment. The forecast performance of the augmented model is compared with that of its predecessor.
Monthly forecast charts from November 2013 to June 2016. Charts include Real GDP Growth, PCE Inflation, Core PCE Inflation, and Unemployment Rate.
Creator: Guvenen, Fatih and Kuruscu, Burhanettin Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 427 Abstract:
In this paper, we construct a parsimonious overlapping-generations model of human capital accumulation and study its quantitative implications for the evolution of the U.S. wage distribution from 1970 to 2000. A key feature of the model is that individuals differ in their ability to accumulate human capital, which is the main source of wage inequality in this model. We examine the response of this model to skill-biased technical change (SBTC), which is modeled as an increase in the trend growth rate of the price of human capital starting in the early 1970s. The model displays behavior that is consistent with several important trends observed in the US data, including the rise in overall wage inequality; the fall and subsequent rise in the college premium, as well as the fact that this behavior was most pronounced for younger workers; the rise in within-group inequality; the stagnation in median wage growth; and the small rise in consumption inequality despite the large rise in wage inequality. We consider different scenarios regarding how individuals’ expectations evolve during SBTC. Specifically, we study the case where individuals immediately realize the advent of SBTC (perfect foresight), and the case where they initially underestimate the future growth of the price of human capital (pessimistic priors), but learn the truth in a Bayesian fashion over time. Lack of perfect foresight appears to have little effect on the main results of the paper. Overall, the model shows promise for explaining a diverse set of wage distribution trends observed since the 1970s in a unifying human capital framework.
Subject (JEL): J31 - Wage Level and Structure; Wage Differentials, E25 - Aggregate Factor Income Distribution, J24 - Human Capital; Skills; Occupational Choice; Labor Productivity, and E21 - Macroeconomics: Consumption; Saving; Wealth
Creator: Beauchemin, Kenneth Ronald Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 493 Abstract:
This paper describes recent modifications to the mixed-frequency model vector autoregression (MF-VAR) constructed by Schorfheide and Song (2012). The changes to the model are restricted solely to the set of variables included in the model; all other aspects of the model remain unchanged. Forecast evaluations are conducted to gauge the accuracy of the revised model to standard benchmarks and the original model.
Keyword: Forecasting and Bayesian Vector Autoregression Subject (JEL): C53 - Forecasting Models; Simulation Methods, C11 - Bayesian Analysis: General, and C32 - Multiple or Simultaneous Equation Models: Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models; Diffusion Processes; State Space Models
Creator: Schorfheide, Frank and Song, Dongho Series: Working paper (Federal Reserve Bank of Minneapolis. Research Department) Number: 701 Abstract:
This paper develops a vector autoregression (VAR) for macroeconomic time series which are observed at mixed frequencies – quarterly and monthly. The mixed-frequency VAR is cast in state-space form and estimated with Bayesian methods under a Minnesota-style prior. Using a real-time data set, we generate and evaluate forecasts from the mixed-frequency VAR and compare them to forecasts from a VAR that is estimated based on data time-aggregated to quarterly frequency. We document how information that becomes available within the quarter improves the forecasts in real time.
Keyword: Vector autoregressions, Real-time data, Bayesian methods, and Macroeconomic forecasting Subject (JEL): C11 - Bayesian Analysis: General, C32 - Multiple or Simultaneous Equation Models: Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models; Diffusion Processes; State Space Models, and C53 - Forecasting Models; Simulation Methods
Creator: Miller, Preston J. and Roberds, William Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 109 Abstract:
Doan, Litterman, and Sims (DLS) have suggested using conditional forecasts to do policy analysis with Bayesian vector autoregression (BVAR) models. Their method seems to violate the Lucas critique, which implies that coefficients of a BVAR model will change when there is a change in policy rules. In this paper we construct a BVAR macro model and attempt to determine whether the Lucas critique is important quantitatively. We find evidence following two candidate policy rule changes of significant coefficient instability and of a deterioration in the performance of the DLS method.
Keyword: Coefficient instability, Bayesian vector autoregression, and Conditional forecasts
Creator: McGrattan, Ellen R., Miyachi, Kazuaki, and Peralta-Alva, Adrian Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 586 Abstract:
Japan is facing the problem of how to finance retirement, health care, and long-term care expenditures as the population ages. This paper analyzes the impact of policy options intended to address this problem by employing a dynamic general equilibrium overlapping generations model, specifically parameterized to match both the macro- and microeconomic level data of Japan. We find that financing the costs of aging through gradual increases in the consumption tax rate delivers better macroeconomic performance and higher welfare for most individuals relative to other financing options, including raising social security contributions, debt financing, and a uniform increase in health care and long-term care copayments.
Keyword: Japan, Aging, Retirement, Health care, and Taxation Subject (JEL): I13 - Health Insurance, Public and Private, E62 - Fiscal Policy, H51 - National Government Expenditures and Health, and H55 - Social Security and Public Pensions
Creator: Conesa, Juan Carlos, Kehoe, Timothy Jerome, 1953-, Nygaard, Vegard M., and Raveendranathan, Gajendran Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 583 Abstract:
We develop and calibrate an overlapping generations general equilibrium model of the U.S. economy with heterogeneous consumers who face idiosyncratic earnings and health risk to study the implications of exogenous trends in increasing college attainment, decreasing fertility, and increasing longevity between 2005 and 2100. While all three trends contribute to a higher old age dependency ratio, increasing college attainment has different macroeconomic implications because it increases labor productivity. Decreasing fertility and increasing longevity require the government to increase the average labor tax rate from 32.0 to 44.4 percent. Increasing college attainment lowers the required tax increase by 10.1 percentage points. The required tax increase is higher under general equilibrium than in a small open economy with a constant interest rate because the reduction in the interest rate lowers capital income tax revenues.
Keyword: College attainment, Overlapping generations, Aging, Health care, and Taxation Subject (JEL): I13 - Health Insurance, Public and Private, J11 - Demographic Trends, Macroeconomic Effects, and Forecasts, H51 - National Government Expenditures and Health, H20 - Taxation, Subsidies, and Revenue: General, and H55 - Social Security and Public Pensions
Creator: Supel, Thomas M. Series: Staff report (Federal Reserve Bank of Minneapolis. Research Department) Number: 012 Abstract:
No abstract available.