Hi, I am looking to do some work on the Power and Natural Gas market's historical time series data. Currently the marginal fuel is Natural Gas and the market moves the price of peak power by approx 7.5. So Power-7.5*Gas = Spark. The spark is variable based on demand, state of the grid, etc. Most people regress Power to gas Gas and come up with a minimum variance hedge ratio which is close to the 7.5 value. Do you think if one believes the spark is a stationary process that I should look in co-integration on the spread?