October 22nd, 2013, 10:00 pm
As outrun explained well : you can add a number of parameters and consider a number of different contracts to include in your model (Spread Intraday/DayAhead , WeekAhead/MonthAhead...) , on a single gas year you have many spreads but not much tradable contracts in the market with liquidity, and how do you value your vol curve on these contracts where there is no options trading or at least no that you can see (OTC or embedded in other complex products).What i suggest you to do, as it is often really close to the pricing found by way more complex models use in industry and remains doable by a single person is to select a certain number of contract ID, DA,WEA, WA,MA (Intraday,day ahead, week end ahead...) you take the historical data for each of this contract for the last 5 Gas Years and built a backward tree of what would have been the Optimal Way for the storage to maximise your PNL considering all the Injection and Withdrawal Rates, costs and racheting curve (Variation of Injection/Withdrawal rates depending on the volume in the storage) . It is the same as locking the intrinsic value in the forward market and then looking at the historical PNL you'd have realized by arbitraging with spreads, that is equivalent of buying your flexibility at 0 and doing the gamma, so your PNL is the historical value of your storage flexibility. You'll see that Intraday is almost never used and majority of what you could have done is DayAhead /WeekAhead or WeekAhead/MonthAhead. Note that even this kind of model required time to develop and implement and that your data integrity is important because fake data can allow a certain path on the tree which would never have been reachable otherwise. I dont work in this industry anymore but looking at the gas market, i'd be very surprise that the flex got much market value left because of the structure of the market. IF you have a really fast storage , for example one you can fill up in 10 days then yes, you have huge flex, but for a classic 90days storage on TTF , i dont think you will get a better bid than 0.10cts of Eur / MWh over intrinsic.Maybe outrun could confirm , precise or correct what I just said.
Last edited by
jige on October 22nd, 2013, 10:00 pm, edited 1 time in total.