I've applied a 4-variable panel data VAR for 19 units for 15 years' data (variables were all normalized to get values between 0 and 1). After checking for cross sectional dependency, appropriate unit root tests were applied for stationarity and where necessary, variables were transformed to make them stationary.
I've used STATA [commands mentioned in a paper by MRM Abrigo and I Love, 2016) to conduct panel VAR. Sabiliity condition was satisfied, VAR shows significant relation between the variables (so does Granger tests). But when I plot the IRF graphs with Monte Carlo simulations, my confidence bands are explosive.
Without the MC simulation, the IRF graphs look good.
How can IRF graphs be inconsistent with my data that is stationary and results that are stable?
Is it something to do with using normalized values? Or a problem with the instrumental lags selected?
Thanks a lot in advance!
https://www.researchgate.net/post/Problem-in-my-panel-VAR-impulse-response-functions,u can follow this link.
https://www.diw.de/documents/publikationen/73/diw_01.c.601220.de/dp1762.pdf,this link may be useful.
Your confidence bands do not look unusual at all. All confidence bands for forecasts of any sort widen as they move further into the future, and tend to do so at an increasing rate, as forecasting something something further and further into the future becomes more and more uncertain.
You need not go into any literature. David's answer is simple and correct. A forecast for the next period, in general, includes an addition of or multiplication by a random variable (the "error term"). The mean is zero, but it has a positive variance (standard deviation). In general, the deviation coes not change from period to period, but if the random variable is added for, say n periods, the accumulated variance will be n times the variance for one period. The standard deviation, therefore, increases with the root of the number of periods until the end of the forecast.
- Badges
- Science topic
- Similar topics
- Correction