I am using a GARCH-DCC MODEL to examine the co-movements of the price variation between the financial markets and the commodity markets. The model output gives me a sum of the ARCH and GARCH coefficients bigger than 1 that would mean that the variance is exponentially increasing over time. Not a desirable phenomenon. i'm getting this in a DCC-GARCH(1,1) but i also get this when i try higher orders. My variables are cointegrated so I made a VECM model and then used the residuals to estimate the DCC-GARCH MODEL.
I have daily data with 700 observations, you can see my output bellow.
What can i do to fix this?
PS: the problem isn't the size of my sample, i also get this result when i use a big sample
Your data are for about two years. It may be, that during these two years, a clear upward trend of the financial market prices together with higher and increasing variations (in comparison to the commodity markets) were observed. In this case, one could expect an exponentially growing relation - a result which is only valid for this period (or perhaps periods of financial upswings).
Hi, thank you for replying.
I thought that could be the problem but when i test the model with my full sample with 4165 observations i get the same problem. I attached the file bellow with the full sample