My firm generates electricity from wind. Accordingly, most of my generation takes place at night, when prices are low -- and, due to congestion / oversupply, often sharply negative -- so much so that the spread between daytime and nighttime (or, more generally, "peak" and "off-peak") prices itself is often negative.
While my first instinct was to hedge this spread with derivatives (either an spread option using Kirk's approximation or using a bear vertical spread), I noticed that the spread between the daytime and nighttime period is random across days and across months in futures as well as real-time markets, so I'm wondering if such an approach is workable. Alternatively, I am considering an extreme-value theory approach, whereby I would create an insurance product based on the probabilities and expected shortfalls of negative spreads -- akin to a CDS.
Could someone please weigh in on this situation? I don't want to abandon a market-based approach -- but, in what appears to be an either highly inefficient or highly unstable market, it seems the derivatives approach is unusable.