The end-of-Season (EOS) storage forecast is an important analytical metric that is supposed to help traders understand the underlying supply/demand balance in the market. But does it? From our observations, we have noticed that traders often misunderstand and misinterpret the whole concept of the EOS forecast. This article aims to provide some clarifications.
First of all, what is the EOS forecast and how do analysts arrive at their projections?
The natural gas industry considers two seasons in storage operation: the withdrawal season (from Nov. 1 through March 31) and the injection season (from April 1 through Oct. 31). Therefore, the EOS forecast is essentially a projection on the level of natural gas inventories either for March 31 or for Oct. 31. We refer to the former as the end-of-withdrawal-season storage forecast (EOwS) and to the latter as the end-of-injection-season storage forecast (EOiS).
In both cases, the number is the result of a comprehensive model calculation on the basis of multi-parameter inputs (see diagram below).
This diagram just shows the basics. The reality, however, is more complex. For example, production input will be a derivative of many more factors, such as extraction loss, well spacing, rig efficiency, break-even costs, and much more. Likewise, the exports forecast will have to include structural factors, foreign demand, pipeline nominations, vessels tracking (in case of LNG), etc.
All model inputs have a potential to impact the final result. However, the relative force of their impacts will vary, and therefore different inputs have different weight within the model. For example, the weather forecast, which determines the lion's share of natural gas consumption, is the single most critical input in the model. However, it is also the one most susceptible to changes, which sometimes can be unexpected, swift and dramatic. This is especially true during the winter, when natural gas demand is highly inelastic and supply is relatively fixed (at least in the short term). Therefore, the weather variable can have an overwhelming effect on the EOS storage forecast and lead to price spikes.
On balance, EOS storage forecast is the result of many variables, some of which are exceptionally volatile. Traders, therefore, should not view it as a kind of static figure. In fact, they should not view it as a forecast at all. Instead, they should take it as an indicator, a composite index, and not consider it as a long-term storage projection per se.
The chart below illustrates the evolution of our EOwS storage forecast (for March 31, 2017).
Source: CME Group, Intercontinental Exchange, GeckoiCapital
If the EOS forecast is a composite indicator of the balance between all relevant market variables, then there are two conclusions to make from this chart:
- The rising EOS forecast suggests that market balance is loosening and should (in theory) lead to lower prices. Equally, a declining EOS forecast suggests that market balance is tightening and should (in theory) lead to higher prices.
- The gap between our EOS projection and market expectations represents a potential trading signal. The situation when the gap is positive -- i.e., when our projection is above market expectations -- signifies what we call an "implied bearish surprise" and should (in theory) lead to lower prices. The opposite situation -- i.e., when our projection is below market expectations -- signifies what we call an "implied bullish surprise" and should (in theory) lead to higher prices.
The ultimate aim of the EOS indicator is to help traders understand where the price of natural gas is heading. However, the price of natural gas is, in itself, an endogenous input in the EOS forecast model (unlike the weather forecast, which is an exogenous factor not influenced by other elements within the same model). Price has an impact on consumption and production, which in turn impact storage flows, which then determine EOS storage inventories. Therefore, it is perfectly natural for rising prices to push the EOS indicator higher and for lower prices to pull it lower, ceteris paribus. Thus, the EOS storage forecast model is not a traditional econometric model, but rather a cycle model, which must be updated on a daily basis.
We provide daily updates for both seasons. Our latest projection for the end-of-injection season -- i.e., for Oct. 31, 2017 (issued on Friday morning) -- was 3,457 bcf, which is, of course, not going to happen. That's not because our calculations are wrong, but because prices will remain elevated during the injection season to correct the imbalance.
Remember, the EOS storage forecast is not a forecast. It is a composite index of all market variables.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.