Week 3 Flashcards
Give an informal description of Granger-causality.
Give a formal description of Granger-causality.
What are the two hypotheses tested, when testing for Granger-causality?
What test do we use for Granger-causality? With which parameters?
F-test, with N, and KT-K^2p-K degrees of freedom.
- N: number of restrictions
- K: number of variables
- p: from VAR(p)
- T: number of datapoints
Give a general description of the derivation of Granger-causality.
What are some limitations of Granger-causality?
Which two questions do response analyses answer?
Descripe the idea behind impulse response analysis.
How do you find the response of an impulse?
How is the accumulated impulse response calculated? What is the long-run response?
Describe the steps of the Bootstrap method.
What is the idea behind orthogonal impulse responses?
What is the MA representation of responses to orthogonal shocks? What is a disadvantage?
How to calculate a 1-step point forecast? Why is this valid?
Of a known VAR(p) process
What is the general formula for a h-step forecast?
Of a known VAR(p) process
What is the forecast error for a 1-step, 2-step and general h-step prediction?
Of a known VAR(p) process.
What is the mean and variance of a forecast?
Of a known VAR(p) process
Describe how the interval of a forecast is derived.
Of a known VAR(p) process
What is the forecast error of an estimatated VAR(p) process?
What is the MSE (matrix) of a forecast?
Of an estimated VAR(p) process.
How to create an interval of an estimated processes?
Estimated VAR(p) process.
How can we determine the “true” order of a VAR(p) process?
Why do we often not care about the “true” order of a VAR(p) model?
What is the 1-step-ahead forecast mean squared error matrix?
What is the FPE?
What are the three criteria for model selection?
