Adding Confidence to Your Trading Decisions

Home Membership Research Track Record Free Content Update Schedule Contacts
Log In

Natural Gas Trading and Analytics Guide


This Trading and Analytics Guide will introduce you to Bluegold Research online content and will help you better understand how our charts and indicators can improve your trading strategy. 


Supply and demand are two most important economic concepts. The price of any commodity is a reflection of supply and demand. In other words, the balance between supply and demand determines the price. Usually, when supply is above demand, then the market balance is loose and the price is low. When supply is below demand, then the market balance is tight and the price is high. Table 1 shows the price of natural gas and key market fundamentals for a number of selected historical periods. (You can download historical production and consumption figures from the Data Archive and also the latest forecasts from the chart pages: Dry Gas Production and National Consumption).   

Table 1. Market Fundamentals
Period Dates Season







NG Price


NG  Price


1 Nov. '10 - Mar. '11 Withdrawal 10,731 12,923 -2,191 4.176 4.240
2 Apr. '11 - Oct. '11 Injection 15,397 13,171 +2,227 4.138 3.524
3 Nov. '11 - Mar. '12 Withdrawal 11,312 12,622 -1,310 2.900 2.191
4 Apr. '12 - Oct. '12 Injection 15,888 14,398 +1,489 2.738 3.471
5 Nov. '12 - Mar. '13 Withdrawal 11,172 13,376 -2,203 3.509 3.976
6 Apr. '13 - Oct. '13 Injection 15,863 13,856 +2,007 3.773 3.496
7 Nov. '13 - Mar. '14 Withdrawal 11,437 14,426 -2,989 4.400 4.584
8 Apr. '14 - Oct. '14 Injection 16,625 13,940 +2,685 4.199 3.728
9 Nov. '14 - Mar. '15 Withdrawal 12,343 14,458 +2,115 3.249 2.590
10 Apr. '15 - Oct. '15 Injection 17,401 14,922 +2,479 2.689 2.033
11 Nov. '15 - Mar. '16 Withdrawal 12,491 13,900 -1,409 2.059 1.903

Total supply = dry gas production + pipeline imports from Canada and Mexico + LNG imports (bcf)

Total demand = national consumption + pipeline exports to Canada and Mexico + LNG exports (bcf)

Implied Balance - the difference between total supply and total demand 

NG Price -  front month futures price (US$ per mmbtu); Z, F, G, H, J for withdrawal season; K, M, N, Q, U, V, X for injection season

EOS - end of season; J close for withdrawal season; X close for injection season

Source: Energy Information Agency, CME Group, Bluegold Research calculations

The above table shows only absolute figures and does not really indicate if the market was “oversupplied” or “undersupplied” in any particular period. Therefore, it is impossible to determine a trading bias. The only way to figure out if the data is bullish or bearish is to look at it in relative terms ― i.e., compare it to a certain benchmark.

Storage figures (stockpiles or inventories) is the single most important piece of fundamental data in natural gas market. It has the strongest impact on the price because it shows the net result of the interaction between supply and demand. Once again, however, it is important to put information into perspective. One of the most common ways to analyze storage data is to compare it to a 5-year average. Table 2 shows changes in the underground storage during injection and withdrawal seasons and their share of a 5-year average change in percentage terms.  

Table 2. Storage Analytics
Period Dates Season


in Storage



% share of

5-year average

NG Price


NG  Price 


1 Nov. '10 - Mar. '11 Withdrawal -2,199 -1,912 115 4.176 4.240
2 Apr. '11 - Oct. '11 Injection +2,171 +2,037 106.6 4.138 3.524
3 Nov. '11 - Mar. '12 Withdrawal -1,321 -2,048 64.5 2.900 2.191
4 Apr. '12 - Oct. '12 Injection +1,435 +2,104 68.2 2.738 3.471
5 Nov. '12 - Mar. '13 Withdrawal -2,220 -1,925 115.3 3.509 3.976
6 Apr. '13 - Oct. '13 Injection +2,092 +1,997 104.8 3.773 3.496
7 Nov. '13 - Mar. '14 Withdrawal -2,955 -1,907 155 4.400 4.584
8 Apr. '14 - Oct. '14 Injection +2,747 +2,018 136.1 4.199 3.728
9 Nov. '14 - Mar. '15 Withdrawal -2,110 -2,181 96.7 3.249 2.590
10 Apr. '15 - Oct. '15 Injection +2,470 2,131 115.9 2.689 2.033
11 Nov. '15 - Mar. '16 Withdrawal -1,463 -2,157 67.8 2.059 1.903

EOS - end of season; J close for withdrawal season; X close for injection season

Source: Energy Information Agency, CME Group, Bluegold Research calculations

Now it is easier to see if the market was “oversupplied” or “undersupplied” in a particular period.

Withdrawal Season Example

In Period 3, draws were very low compared to the norm (just 64.5% of the 5-year average), which resulted in a much lower price by the end of the season ($2.191 per mmbtu). Likewise, draws in Period 7 were above the 5-year norm (155% share), which not only means that demand was stronger than supply (it is to be expected during the withdrawal season), but also that the subsequent misbalance in the market was quite large compared to history. Therefore, the price by the end of the season was relatively high ($4.584 per mmbtu). 

Injection Season Example 

In Period 4, injections were very weak if compared to history (just 68.2% of the 5-year average), so the price recovered from $2.191 to $3.471 per mmbtu. Following the same logic, in Period 8, the price went down (from $4.584 to $3.728 per mmbtu) due to strong injections (136% share of the 5-year average). 


We publish four separate charts to help you understand the underlying picture in the market balance: Injection, WithdrawalComparative Analysis, and Flows.

Withdrawal and Injection charts show the level of working gas in the underground storage in comparison with previous years. Comparative Analysis chart plots the level of working gas in the underground storage against a five-year average, maximum and minimum stocks. All three charts also display our long-term storage forecast. As a rule, the higher our storage projectios curve is, the lower should be the price and vice versa.

For example, this is how our Comparative Analysis chart looked on January 6, 2016 (when natural gas was trading in 2.267-2.345 range).

Figure 1. Comparative Analysis Chart (print screen as of January 6, 2016)

At that time, we said that we still saw high EOS inventories by the end of March (link to Tweet). Indeed, we expected withdrawal season to finish with around 2,200 bcf in storage, which is well above the 5-year average and not much below a 5-year maximum (see Figure 1 above). Therefore, the projection was bearish, meaning that short positions had a higher probability of success. The subsequent rally on (January 7 and 8) proved to be short-lived and by January 19, natural gas lost 15% of its value. 

Flows chart shows the actual change in the underground storage compared with previous years. It also displays our long-term forecast. This chart may be more useful for short-term traders or day-traders. The key here is to look for differences between projected flows and a 5-year norm and between projected flows and market expectations.

For example, this is how our Flows chart looked on January 14, 2016 (when natural gas was trading in 2.139-2.274 range).

Figure 2. Natural Gas Storage Flows chart (print screen as of January 14, 2016)

At that time, we were expecting the next three reports (for weeks ending Jan 15, Jan 22, and Jan 29) to show relatively strong draws ― i.e., below the 5-year average curve (see Figure 2 above; link to Tweet). On January 14, the market closed at 2.139. Over the next few weeks, as bullish reports were on the horizon, the price slowly recovered and by January 29, the market closed at 2.298 (+7.43% from January 14 close).

However, it was already clear even on January 14, that any rallies would be hard to sustain because draws for February looked less bullish ― i.e., above the 5-year average curve (see Figure 2). Predictably, March contract lost 6% on February 1.

All storage charts are updated on a daily basis


Our Gravity Models were specifically designed to provide a quick overview of the fundamental trend in the market. Days of Supply Model 1 (DoS-M1) and Market Balance Model 2 (MB-M2) show the same thing ― the likely direction of the front month futures price. The only difference between them is that DoS-M1 generates its trends based on domestic factors (national consumption and national production), whereas MB-M2 also employs external factors (exports and imports). DoS-M1 may not provide a full picture of the situation in the marketplace, but it still captures the lion share of fundamental developments. That is because the weight of external factors remains negligible. Imports make up just 10% of total US dry gas production, while exports make up 7% of total US national gas consumption. Imports and exports, therefore, essentially balance each other. However, the importance of external factors ― especially, that of exports ― is projected to increase in the coming years. Therefore, we decided to keep track of them separately in the MB-M2.  

Figure 3 and Figure 4 (below) show the evolution of DoS-M1 in December 2015.   

Figure 3. Days of Supply Model 1 chart (print screen as of December 15, 2015)  

On December 15, 2015 (when natural gas was trading in 1.822-1.900 range and a trading week average price (TWA) was around $1.85 per mmbtu), our DoS-M1 was clearly indicating the upcoming upward movement in price (link to Tweet). The market closed at 1.767 on Friday, December 18 and then rallied more than 35% to close at 2.372 on December 29. 

However, on December 29, 2015, it was already clear from the DoS-M1 that this rally will be short-lived (see Figure 4 below) and that from mid-January the downtrend will resume (link to Tweet). Indeed, the market closed at $2.472 per mmbtu on January 8 and then declined by more than 20%.  

Figure 4. Days of Supply Model 1 chart (print screen as of December 29, 2015)  

Please note, that we have since renamed our curves:

DoS → Inclination 1

Trend 1 → Inclination 2

Trend 2 → TREND

Also, note that DoS-M1 and MB-M2 show the average price (TWA), not the latest price. The price curve moves further to the right every Friday when a new forecast week is added.

The general market theory implies that the price level depends not only on today’s supply-demand situation, but also on the balance expected in the future. Therefore, it is very important to keep in mind the commodity forward curve when making trading decisions. If you see that DoS-M1 or MB-M2 curves are going up, do not rush to go long as this trend may be already (at least partially) reflected in price due to the contango in the market. 


We publish three seasonal charts: Natural Gas Contract, Natural Gas Calendar and EP Sector Consumption.

NG Contract Seasonality and Calendar Seasonality charts show the expected level of return for 12 natural gas contracts (in case of contract seasonality) and for 12 months (in case of calendar seasonality). You can use these charts to look for significant deviations between recent price performance and its historical performance. 

For example, on January 29, 2016 the market closed at 2.298, which at that time represented a 7.13% growth for H (March) contract. This kind of performance contrasted sharply with the historical returns for H contract (see Figure 5 below). On this basis, it seemed logical to go short. Indeed, it would have been a profitable trade, as natural gas lost some 14% over the next four days.

Figure 5. Natural Gas Contract Seasonality chart (print screen as of January 29, 2016)  

Seasonals are certainly not the only basis for a trade and they can be skewed by one-off events or coincidences. Overall, they are just one among many tools and are something to keep in mind every time the calendar turns.

EP Sector Consumption chart shows the consumption of natural gas in the Electric Power sector (as a percentage of total gas consumption) in each of the calendar months. This information allows assessing the strength of the effect of NG-Coal spreads on natural gas consumption in a particular period. For example, if natgas-coal spread is low in August (when gas consumption in the Electric Power sector is at its peak), then the effect on coal-to-gas switching will be very strong with a possible bullish effect on injections. Likewise, natgas-coal spreads (either low or high) play a much smaller role during winter months.  


We know that many of our clients are day-traders. However, we are not big fans of day trading. Of course, if you’re a short-term trader and you know what you are doing, we do not want to discourage you. In fact, our Market Fundamentals section was created specifically for day-traders, following their request. However, we do want to stress that natural gas market is exceptionally volatile. 'Day-trading' in this market is very risky. We would like to recommend a different, more conservative approach. Our recommendation is very simple:    

1. Look at the fundamental trends (Storage Charts, Gravity Models)

2. Decide what the fundamentals trend is (bullish or bearish)

3. If the fundamental trend is bullish - consider only long positions

4. If the fundamental trends is bearish - consider only short positions

Once you have decided what positions you want to consider, then look for deviations (Gravity Models, Fair Value Estimator, Weather Anomaly Indicator). Use the latest data in Market Fundamentals section to confirm your decision, but be wary of the risks. 

Once again, we personally do not think that day trading in natural gas is a good idea. We believe that it is always safer to initiate positions with a long-term and mid-term view in mind. For example, on May 31, 2016 it was clear that storage injection would be bullish (see Figure 6 below). Therefore, there was a bullish trading bias in the market. Bullish trading bias at that time meant that injections were projected to stay below 5-year average. Therefore, the market balance was projected to get tighter. Therefore, there were likely to be more bullish days ahead, than bearish days. In this situation, a conservative trader would only look for opportunities to go long. Even if Market Fundamentals issue a short-term bearish signal, a conservative trader would rather wait for price to decline and then go long and not play the possible decline itself. Or a trader can simply ignore a short-term signal and still go long based on the fundamental trend. By the way, since May 31, prices increased by more than 20%.   

Figure 6. Natural Gas Storage Flows chart (print screen as of May 31, 2016)  

The single, most important fundamental indicator is (and has always been) storage forecast. Some traders focus exclusively on storage figures and do not use any other indicators. This makes sense because storage forecast (especially end-of-season storage forecast) is a single figure, which is clear, well defined, but most importantly, it is closely tracked by market players on a daily basis. Conversely, Gravity Models are unique and complex systems, which employ many variables, each of which can be incorrect. Still, Gravity Models have an advantage of showing the underlying mid-term trend in the marketplace. They can also help spot important deviations. For example, one such deviation occurred on May 19, 2016 (see Figure 7 below).    

Figure 7. Market Balance Model 2 chart (print screen as of May 19, 2016)  

Generally, we recommend being open-minded and conservative. Go short if you think the price is too high and needs to correct. Go long if you think otherwise. Open smaller positions when deviations are not significant, open bigger positions when deviations are large. Always stay on the side of the fundamental trend, but do not rush in. Psychologically and technically going short in contango markets is easier. Going long is more tricky. Your criteria for going either short or long should be the same, but with long positions try to be more conservative. When price goes up, it usually goes up simultaneously across the whole of the forward curve, so bullish forecasts get reflected way into the future.


Market Fundamentals section shows the latest forecasts for four key market variables: consumptionproduction, exports and imports. Subsequently, these four variables yield forecasts for three important indicators: external trade balancetotal balance and storage

Negative changes in the balances indicate that the supply/demand equilibrium is getting tighter. Theoretically, a tight market balance should lead to price increases. Therefore, negative changes in the balance are viewed as a bullish signal.

Positive changes in the balance reveal the opposite situation – i.e., softer demand and stronger supply. Therefore, positive changes are viewed as a bearish signal.

Normally, the signal is strongest when changes for all analyzed periods are on the same side – either negative or positive. Obviously, the greater the overall (cumulative) change in balance, the stronger should be the signal. It has been observed, however, that the relative importance of the near-term outlook diminishes with time. So, for example, in mid-September, we will still be releasing latest forecasts and changes for monthly totals in the Long-Term chart and for averages over a 15 day period (in this case for September 30) in the Mid-Term Chart.  However, the relative importance of near-term forecast will be lower than for periods that are further away. We believe that the impact of the fundamental changes for a particular period are starting to lose in significance once the end of that period is less than 10 days away. That is why, we would normally add a new analyzed period around that time.

Traders would focus on absolute values (measured either in mmcf per day or bcf), as well as on changes from the previous forecasts. It is assumed that absolute values determine the price regime, while the changes determine the price direction (at least this is what economic theory teaches us). It is therefore further assumed that absolute values are already reflected in the current prices, whereas changes are not and thus a trader should pay a special attention to the actual changes in the forecast if he or she wants to anticipate the price direction.

Total Balance and Storage can be viewed as "umbrella indicators", because they essentially summarize all the other developments in the markeplace, such as production, consumption, exports and imports. However, it is still important to monitor the actual components of the market balance. For example, changes in consumption and production have a greater influence than changes in exports or imports. As a rule, consumption tends to be more volatile, while production is relatively stable. However, if the market decides that negative changes in production will be structural and long-term, prices could rise even against the backdrop of declining consumption.

Copyright 2015 - 2019 © Bluegold Research     [ Terms of Use ]   [ Contacts ]
Site by Smartum IT