GardiDevang
New Member
Hello,
I am trying to develop a parametric VaR model for a portfolio of agricultural commodities. The unique aspect of this portfolio is the presence of something know as Basis positions (Cash Price - Nearby Future Prices). In a Basis position, one locks in the margin, which is the Basis, but is still exposed to price fluctuations. The challenge for me is to calculate volatilities for Basis prices which fluctuate from positive to negative and back to positive. For example, below could be the basis prices for five days:
0.2
0.2
-0.8
-0.8
0.2
Given, that the time series moves from positive to negative and back to positive, I cannot calculate the natural log of returns. I calculated simple arithmetic returns, and the average standard deviation (volatility), in some cases, comes to 1000%, which is absurd. Does anyone have an idea of what is the right approach? Would Exponential Weighted Moving Average work in this case?
Thank you
Devang
I am trying to develop a parametric VaR model for a portfolio of agricultural commodities. The unique aspect of this portfolio is the presence of something know as Basis positions (Cash Price - Nearby Future Prices). In a Basis position, one locks in the margin, which is the Basis, but is still exposed to price fluctuations. The challenge for me is to calculate volatilities for Basis prices which fluctuate from positive to negative and back to positive. For example, below could be the basis prices for five days:
0.2
0.2
-0.8
-0.8
0.2
Given, that the time series moves from positive to negative and back to positive, I cannot calculate the natural log of returns. I calculated simple arithmetic returns, and the average standard deviation (volatility), in some cases, comes to 1000%, which is absurd. Does anyone have an idea of what is the right approach? Would Exponential Weighted Moving Average work in this case?
Thank you
Devang
