| File Name | Description |
| FORCEDFACTOR.SRC | Computes a factorization of a covariance matrix which includes (scale of a) specified columns in the factorization, or, optionally, a scale of specifieds row in the inverse of the factorization. This allows you to either force an orthogonalized component to hit the variables in a specific pattern (done by setting a column of the factorization), or to force that an orthogonalized component be formed from a particular linear combination of innovations (forces a row in the inverse). (Updated in Feb., 2006 to allow multiple rows/columns) |
| By: Estima, 3/21/2007. | |
| Version: 6 or later. Reference: n/a | |
| GIBBSVAR.PRG | This is an example of the use of Gibbs sampling applied to a VAR with a standard Minnesota prior. Different priors can be handled by changing the way that bprior and hprior (the mean and precision of the prior) are created. |
| By: Estima, 7/17/2003. | |
| Version: 5 or later. Reference: |
|
| GaliAER1999.ZIP | Replication file for Jordi Gali, "Technology, Employment and the Business Cycle: Do Technology Shocks Explain Aggregate Fluctuations", American Economic Review, 1999, vol. 89, pp. 249-271. |
| By: Gali, Estima, 3/14/2006. | |
| Version: 6.1 or later. Reference: Gali | |
| KPSWAER1991.ZIP | Replicates results from King, Plosser, Stock and Watson, "Stochastic Trends and Economic Fluctuations", American Economic Review, 1991, vol. 81, pp. 819-840. |
| By: Watson, Estima, 3/14/2006. | |
| Version: 6.2 or later. Reference: King; Plosser; Stock; Watson | |
| MIXVAR.SRC | MIXVAR computes estimates for a single equation using the mixed estimation procedure used by the RATS command ESTIMATE. It can be helpful when you have a variable you wish to forecast which does not fit into a full-blown VAR, because it has no predictive content for the other variables, or when you have a limited amount of data for one variable, so you need to use fewer coefficients in that equation than in others. |
| By: Estima, 7/25/2005. | |
| Version: 5 or later. Reference: |
|
| MODELCOMPANION.SRC | Defines a function "%ModelCompanion(model)" which returns the companion matrix for a model (could be a VAR, but doesn't have to be). |
| By: Estima, 2/15/2006. | |
| Version: 6.1 or later. Reference: |
|
| MONTESUR.PRG | MONTESUR.PRG does Monte Carlo integration for impulse responses for a near-VAR. Implements some key recommendations of Sims and Zha, including the use of fractiles instead of standard deviations in computing error bands, and the use of a degrees of freedom correction on the posterior density for the residual covariance matrix. See the November 2002 RATSLetter for more details. Also, we have also transposed the way the graphs are displayed, so that responses of a single variable to the different shocks are shown in a single row, rather than a single column. |
| By: Estima, 3/16/2004. | |
| Version: 5.03 or later. Reference: Sims; Zha | |
| MONTEVA2.PRG | MONTEVA2.PRG is an updated version of the MONTEVAR.PRG example program for computing error bands for impulse responses for a standard VAR. This implements some key recommendations of Sims and Zha, including the use of fractiles instead of standard deviations in computing error bands, and the use of a degrees of freedom correction on the posterior density for the residual covariance matrix. See the November 2002 RATSLetter for more details. Also, we have also transposed the way the graphs are displayed, so that responses of a single variable to the different shocks are shown in a single row, rather than a single column. |
| By: Estima, 12/2/2002. | |
| Version: 5 or later. Reference: Sims; Zha | |
| MONTEVAR.SRC | This computes and graphs error bands for impulse response functions for a VAR using Monte Carlo simulation. This assumes the model is a symmetric VAR estimated using the ESTIMATE instruction. This uses a Choleski factorization, though it can be modified fairly easily to do any just-identified factorization. |
| By: Estima, 4/5/2007. | |
| Version: 6.3 or later. Reference: |
|
| MONTEZHA.PRG | This implements Sims-Zha's approach for overidentified structural VAR's, using Importance Sampling techniques. See the November 2002 RATSLetter for more details. Also, we have also transposed the way the graphs are displayed, so that responses of a single variable to the different shocks are shown in a single row, rather than a single column. |
| By: Estima, 4/9/2003. | |
| Version: 5 or later. Reference: Sims; Zha | |
| MVBNDECOMP.SRC | Computes a multivariate Beveridge-Nelson decomposition of a set of series via a vector autoregression. Reference: Arino and Newbold, "Computation of the Beveridge-Nelson Decomposition for Multivariate Economic Time Series", Economic Letters, 1998, vol 61, pp 37-42. |
| By: Estima, 7/14/2005. | |
| Version: 6 or later. Reference: Beveridge; Nelson; Arino; Newbold | |
| PSDINITCX.SRC | Contains the user-defined function %PSDINITCX, which implements the calculation of a ergodic variance of a state space model using the diagonalization methods described in Soren Johansen, "A Small Sample Correction for the Test of Cointegrating Rank in the Vector Autoregressive Model", Econometrica, September 2002. This is more efficient than the direct solution of the linear system used in the RATS function %PSDINIT when there are 6 or more states, with the advantage becoming quite noticeable when there are 15 or more. |
| By: Estima, 7/8/2004. | |
| Version: 6 or later. Reference: Johansen | |
| SHORTANDLONGRUN.SRC | This computes a factorization of a covariance matrix with a combination of short and long run restrictions. It can only be applied to just-identified parameterizations where the restrictions are zero restrictions. |
| By: Estima, 3/30/2007. | |
| Version: 6.1 or later. Reference: |
|
| SVAR.SRC | Procedure for estimating the parameters for a structural VAR. By Giannini, Lanzarotti, and Seghelini, based on Giannini's monograph entitled "Topics in Structural VAR Econometrics". An example program using SVAR.SRC and VMA.SRC, called SVAREXAM.PRG, is also available. This requires the data file ITALY.RAT. |
| By: Giannini, Lanzarotti, and Seghelini, 10/18/2005. | |
| Version: 4 or later. Reference: Giannini | |
| TVARSET.SRC | TVARSET sets up a time-varying parameters VAR system for estimation using the Kalman filter. As written, it is based upon the simple symmetric Bayesian VAR prior, with the time-variation proportional to the initial covariance matrix. |
| By: Estima, 7/22/2003. | |
| Version: 5 or later. Reference: |
|
| UhligJME2005.ZIP | Three programs replicating VAR model identification results from Uhlig, "What are the effects of monetary policy on output? Results from an agnostic identification procedure." Journal of Monetary Economics, 2005, 52, pp. 381-419. |
| By: Estima, 3/14/2006. | |
| Version: 6.1 or later. Reference: Uhlig | |
| VAR.SRC | VAR.SRC is a sophisticated, menu-driven procedure for selecting, estimating, and evaluating VAR models. Written by Norman Morin. This procedure makes it easy to graph autocorrelations of your data, test for lag length, do Wald tests on coefficient restrictions and on the Variance/Covariance matrix, test residuals for serial correlation, ARCH, normality (including Jarque-Bera), compute the sum of the Vector Moving Average coefficients, display the VMA representation coefficients, do impulse response analysis using a variety of decompositions, compute forecast error variance decomposition, and more (as if that weren't enough!). The latest update adds bootstrapped standard errors for impulse responses. Other recent updates included the Blanchard + Quah decomposition for IRFs. If you are using RATS Version 4.20 or later, download VAR.SRC. If you are using 4.00 through 4.10, download VAR400.SRC. VAR400 offers most of the same features as VAR, but with a less user-friendly interface due to lack of USERMENU instructions, etc. Also, please note that the 4.0 versions are not being updated, and thus do not include Blanchard-Quah and other recent additions. |
| By: Norman Morin, 12/2/2002. | |
| Version: 4 or later. Reference: Blanchard; Quah; Hamilton; Lutkepohl; Magnus; Doan | |
| VAR400.SRC | Outdated version of VAR.SRC. Use this older version only if you have RATS 4.10 or older. |
| By: Norman Morin, 3/16/1999. | |
| Version: 4 or later. Reference: Hamilton; Lutekepohl; Doan | |
| VARCALC.SRC | VARCalc does a direct calculation of a VAR. |
| By: Estima, 2/15/2006. | |
| Version: 6.1 or later. Reference: |
|
| VARFPE.SRC | Computes minimum FPE representation for the equations in a VAR. |
| By: Estima, 7/25/2005. | |
| Version: 5 or later. Reference: |
|
| VARIRF.SRC | This organizes the graphs of an impulse response function from an already estimated VAR. |
| By: Estima, 7/23/2005. | |
| Version: 6.1 or later. Reference: |
|
| VARLAGSELECT.SRC | VARLagSelect chooses the lag length which minimizes one of the information criteria. |
| By: Estima, 4/1/2007. | |
| Version: 6.1 or later. Reference: |
|
| VARMADLM.SRC | VARMADLM.SRC includes two setup routines for estimating or analyzing a vector ARMA model using the RATS instruction DLM. |
| By: Estima, 1/29/2007. | |
| Version: 6 or later. Reference: |
|
| VMA.SRC | Procedure for computing impulse response functions and forecast error variance decomposition for structural VAR's (requires SVAR.SRC procedure described above). By Giannini, Lanzarotti, and Seghelini. An example program using SVAR.SRC and VMA.SRC, called SVAREXAM.PRG, is also available. This requires the data file ITALY.RAT. |
| By: Giannini, Lanzarotti, and Seghelini, 2/10/1994. | |
| Version: 4 or later. Reference: Giannini | |
| YULEVAR.SRC | This estimates a VAR on stationary data using the Yule-Walker equations. |
| By: Estima, 7/14/2005. | |
| Version: 6.1 or later. Reference: |
|