Skip to main content
3 votes

What am I missing in the vector case?

I have not seen any packages or functions that automate the whole process. So, you may have to create your own procedures. First, we need to ensure whether the spreadsheet has the displacement records ...
DrJerryTAO's user avatar
  • 1,636
3 votes

Normality assumption n=48

From my understanding, more important than normally distributed variables (prior to modelling) is the normality of model residuals (evaluated from a Q-Q plot, for example) after your model has been ...
sylvesterspot's user avatar
2 votes

VAR regression between I(1) and I(0)

If you are fairly sure that $Y_1$ is I(1) and $Y_2$ is I(0), take the first difference of $Y_1$ and model $(\Delta Y_1,Y_2)$ using a VAR. There is no point in using a VECM, as there is no ...
Richard Hardy's user avatar
2 votes

Intercept or trend in a VAR model for stationary percentage changes

You want the model specification that approximates the data generating process well enough. If your percentage changes have nonzero means (which is plausible in contexts with nonzero average growth), ...
Richard Hardy's user avatar
1 vote
Accepted

Estimating VAR of differences of potentially cointegrated variables

If the variables $Y$ are cointegrated, VECM is an appropriate model. A VAR for $\Delta Y$ (first-differences of $Y$) is simply the VECM without the error correction term (ECT). Thus, using a VAR for $\...
Richard Hardy's user avatar
1 vote

Decomposition of VAR(1) coefficient matrix

Are you sure you want to reduce the degrees of freedom? I think what you mean is you want to reduce dimensionality. You can use e.g. LASSO to reduce the number of parameters to be estimated. See e.g. ...
Pavel Filip's user avatar
1 vote
Accepted

Can I compute a VAR Model and then work on only one OLS equation?

Yes. Working with a single equation form a VAR model is fine. On its own, it would be called autoregressive distributed lag (ARDL) model. (In a way, ARDL is more flexible than a VAR, as you can choose ...
Richard Hardy's user avatar
1 vote

How to choose the best one model from ARIMAX, ARCH/GARCH and VAR?

I would suggest looking at forecast accuracy / cross validation. By testing the model on part of the dataset that was not used to fit the model. In other settings such as multivariable regression, it ...
user1029384756's user avatar
1 vote

Selection of best VARX model using VAR() in R

If you are interested in FED interest rate, I would look at how well the different models are able to predict it out of sample, namely, what is the expected loss from the (one-step-ahead) forecast ...
Richard Hardy's user avatar
1 vote

VARMAX does not improve over ARIMA

Challenges of using time series forecasting to predict inflation The main issue with using exogenous variables to forecast inflation is that inflation is impacted by a myriad of different factors - ...
Michael Grogan's user avatar
1 vote
Accepted

What am I missing in the vector case?

If your time series is a random walk like a type of Brownian motion, then the distance from the starting point is not a stationary time series, and an autocorrelation function has little meaning. In ...
Sextus Empiricus's user avatar
1 vote

Interpreting the VECM: which variable corrects towards which one?

If $\alpha_Y\neq 0$, $Y_t$ adjust to the equilibrium between $Y$ and $X$ ($Y$ "error-crrects"). If $\alpha_X\neq 0$, $X_t$ adjust to the equilibrium between $Y$ and $X$ ($X$ "error-...
Richard Hardy's user avatar
1 vote
Accepted

BVAR model: Draws and Burn-In?

In a Bayesian model, we are often mainly interested in the posterior distribution, as it describes our knowledge about the parameters of interest given our priors and after having seen the data. Now, ...
Christoph Hanck's user avatar
1 vote

In VAR model, can I include not-granger-causing variables in impulse response anaysis?

Regarding plotting the IRF, you can do it for all dependent variables in the model. So if you think a variable belongs in the model, you can plot its IRF. Also, I find your terminology confusing. An ...
Richard Hardy's user avatar
1 vote
Accepted

Eigenvalues of VAR(1) coefficient matrix

The eigenvalues of $\Phi^\intercal \Phi$ are the eigenvalues of $\Phi$ squared, and so also have moduli less than $1$. Proof here For square $\Phi$, $\begin{Vmatrix} \cdot \end{Vmatrix}_2$ is a ...
Taylor's user avatar
  • 20.8k
1 vote

Interpreting Impulse Response Function after first differences of logarithm transformation

Your impulse response function graph shows time evolution of how s2 responds to a shock that originated in s1. A one-unit (one-standard-deviation of s1) change (shock) in your variable s1 at time t ...
Pavel Filip's user avatar

Only top scored, non community-wiki answers of a minimum length are eligible