GARCH method refers to conditional risk and requires Engle's test, however, convergence issues and some risks are unconditional. In dairy industry, I do not expect greater volatility so a nonlinear moving average std deviation may suffice. I know young researchers like new medicines because they think Aspirin no longer work. However, Hendry and several others have never used the new medicines. GARCH  method is not bullet-proof see Pagan and Ullah (1988) see pp99-103 Journal of Applied Econometrics. Why not use rolling or recursive standrd deviation? OR see IMF staff Papers by Anthony Spencer 1996pp469-470 OR see RESTAT May 2008 by Ghossal and Loungai

Thank you very much for your response. Let me put it this way, what if the government fixed rates of milk are barely moving in last decade but the input prices like feed and off course volatile market of oil results in greater volatility of input prices. Eventually, putting farmers into a definite risk in the market. I want to quantiy the extent of risk imperiled to farmers in future. 

I am new to this method and I have been looking into the literature to find if this model fits my problem or not.

I believe you are talking about expected risk from oil price fluctuations. Is that what you mean. One correction data are, not "is" is what you want to write. Muhammad in what type of modelling equation are you going to input expected risk? Have you attempted to go over some of the papers I referred to in my earlier answer? 

Yes dear professor, I am referring to input price fluctuations and oil price is one example of that. What I'm going to do is, I will use the expected risk values as a measure of severity. And, the next step is knowing the risk perception of dairy farmers followed by the study of factors affecting the adoption of risk management tools. 

I will use the risk severity values as a basis for my study. What else do you suggest? 

I tried to locate those papers but, unfortunately didn't find them. 

I appreciate your respose.