Given that forecasts—particularly for the next 2 weeks—have trended warmer over the past few days, the “fair price” has indeed drifted downwards and is now sitting at $3.34 (from $3.44 at the time of submission of this article) if all five-weeks of projections are considered, which likely explains the impressive decline over the last 2 days. Of course, five-week temperature forecasts have a lot of built-in uncertainty, which is why I also provide the higher-confidence 3-week “fair price” which is now $3.26 (versus $3.30). Based on these numbers, there is no fundamental reason for natural gas to rally significantly from here, as long as there is no sudden shift in the forecast, which looks rather unlikely at this juncture. I would not be surprised to see a bump over the next 2-3 days due to a bullish number (>200 BCF) on Thursday or technical fatigue. Not to spam the thread, but if you are interested in noting how the estimates change on a day-by-day basis, I do update all of my stats on my website (address in article or direct link in my profile) every morning at between 6am-7am.

As for your second question, should we finish at 2 tcf, my model indicates a fair price of $3.24 based on multiple iterations of the past 18 months of pricing and storage data. This number may change slightly based on the next couple of weeks. For comparison, the same model projected last year’s end-of-season fair value to be $2.20 versus a realized price of ~$1.90 suggesting that the commodity may fall under the fair price in extreme cases (which would mark a great buying opportunity). However, 2 tcf should inspire much less panic selling than did 2.4 tcf and will unlikely afford such an opportunity.

Thanks for your interest.]]>

Given that forecasts—particularly for the next 2 weeks—have trended warmer over the past few days, the “fair price” has indeed drifted downwards and is now sitting at $3.34 (from $3.44 at the time of submission of this article) if all five-weeks of projections are considered, which likely explains the impressive decline over the last 2 days. Of course, five-week temperature forecasts have a lot of built-in uncertainty, which is why I also provide the higher-confidence 3-week “fair price” which is now $3.26 (versus $3.30). Based on these numbers, there is no fundamental reason for natural gas to rally significantly from here, as long as there is no sudden shift in the forecast, which looks rather unlikely at this juncture. I would not be surprised to see a bump over the next 2-3 days due to a bullish number (>200 BCF) on Thursday or technical fatigue. Not to spam the thread, but if you are interested in noting how the estimates change on a day-by-day basis, I do update all of my stats on my website (address in article or direct link in my profile) every morning at between 6am-7am.

As for your second question, should we finish at 2 tcf, my model indicates a fair price of $3.24 based on multiple iterations of the past 18 months of pricing and storage data. This number may change slightly based on the next couple of weeks. For comparison, the same model projected last year’s end-of-season fair value to be $2.20 versus a realized price of ~$1.90 suggesting that the commodity may fall under the fair price in extreme cases (which would mark a great buying opportunity). However, 2 tcf should inspire much less panic selling than did 2.4 tcf and will unlikely afford such an opportunity.

Thanks for your interest.]]>

If I understand your first question correctly, I have not accounted for the calendar spread between contracts in my 5-week price projections, which generate the 21 cent spread. This was a lapse on my part when creating the model that I have not seen fit to alter since contracts for the next 2 months had been trading within a cent or two of eachother for the past month or so. Now that we’re rolling over into the March contract, I will adapt some of my formulas to take calendar spread into account, especially since the spreads are starting to widen up again.

For your second question, without going into too many details on some of the more proprietary elements of my model, one element of the model runs regressions on multiple time frames using past pricing and storage data to generate a series of “fair price” values. These are then weighted (i.e. multiplied by a fraction of 1) based on their actual historical ability to predict future prices as well as their “theoretical” ability to do so, based on r squared and other correlation coefficients (i.e. "fit," perhaps a misnomer). The number you see listed as a “fair price” is a weighted average of a series of estimates. “Fair Price” numbers are updated daily on my website. I hope I answered at least part of your question.]]>

If I understand your first question correctly, I have not accounted for the calendar spread between contracts in my 5-week price projections, which generate the 21 cent spread. This was a lapse on my part when creating the model that I have not seen fit to alter since contracts for the next 2 months had been trading within a cent or two of eachother for the past month or so. Now that we’re rolling over into the March contract, I will adapt some of my formulas to take calendar spread into account, especially since the spreads are starting to widen up again.

For your second question, without going into too many details on some of the more proprietary elements of my model, one element of the model runs regressions on multiple time frames using past pricing and storage data to generate a series of “fair price” values. These are then weighted (i.e. multiplied by a fraction of 1) based on their actual historical ability to predict future prices as well as their “theoretical” ability to do so, based on r squared and other correlation coefficients (i.e. "fit," perhaps a misnomer). The number you see listed as a “fair price” is a weighted average of a series of estimates. “Fair Price” numbers are updated daily on my website. I hope I answered at least part of your question.]]>