Natural Gas Seasonal Spread

Spreads have reduced risk because you are hedged between periods. If the whole curve falls, both legs fall together and the losses offset. Margin is lower thanks to ICE's inter-period offset. You are betting on the relationship between periods, not absolute direction, which makes the seasonal pattern more tradeable.

The best window is May-September. This is before the Winter premium fully builds and while Europe is refilling storage toward the 90%-by-1-November target. Enter when the Winter-Summer spread is below the 5-year average for that date. Avoid entering in December when the premium is already near its peak.

Typically 4-10 weeks. Seasonal patterns take time to develop - these are not day trades. Monitor weekly but do not overtrade. Exit when the target is reached, the stop is hit, or the thesis changes.

Storage is crucial. Normal storage: patterns work normally. Low storage: amplifies the Winter premium (bullish). High storage: dampens it (bearish). Because UK storage is small, also watch import economics and the NBP-TTF relationship. Check GIE AGSI+ daily fill data against the 5-year norm and the 90% target.

Injection season (April-September): Europe injects gas into storage for winter, and prices tend to be lower. Withdrawal season (October-March): storage is drawn down for heating, and prices tend to be higher due to demand. The UK Gas Year runs 1 October to 30 September.

Use the Z-score: (Current Spread - 5-year Average for this date) / Standard Deviation. Z < -1.5 is cheap (potential long), Z > 1.5 is expensive (potential short). Build a database of historical Winter-Summer (and calendar) spreads to calculate the averages by date.

Roll 2-3 weeks before the near period's delivery start - NBP futures stop trading two business days before delivery begins. Close the existing spread and open a new one in the next periods. Calculate the roll cost (the difference between old and new). In contango rolling costs money; in backwardation it can earn.

Cold or negative-NAO forecasts in the pre-winter window accelerate the premium buildup, so enter earlier if a cold winter is signalled. Mild forecasts dampen the pattern - use smaller size or wait. Low-wind spells lift gas-for-power demand, and weather spikes are good for profit-taking.

Limit individual seasonal trades to 2-3% risk of capital. Total seasonal exposure should not exceed 10% of the portfolio. This prevents overexposure to a pattern failure while still allowing meaningful participation. Remember a Season lot is far larger than a Month lot, so size carefully.

Use a time stop. If there is no meaningful progress after 4-6 weeks and the fundamentals (storage, weather, flows) have not changed, consider exiting. The pattern may not develop this year - do not hold indefinitely hoping for it.

Collect 10+ years of daily spread data. Calculate seasonal averages and standard deviations by date. Add regime indicators (storage from AGSI+, NAO/weather, wind). Test statistical significance. Use machine learning for complex patterns. Walk-forward test to validate before trading live.

Compare NBP patterns with TTF and the wider European curve. When all show the same seasonal signal, conviction is higher. Divergence may indicate a UK-specific factor or an NBP-TTF arbitrage. Interconnector flow direction is a real-time balance signal, and global confirmation reduces pattern-failure risk.

Long calls on Winter contracts for defined-risk bullish exposure. Bull call spreads to reduce premium cost. Protective puts on a futures spread for downside protection. Combine a futures spread with long calls for enhanced upside with defined risk. All use options on ICE NBP futures.

Use spread/strip orders when available for guaranteed execution. If legging, enter the illiquid period first (usually the far season or month). Execute in the core London session, away from the late-afternoon settlement window. Scale large positions over multiple sessions to reduce market impact.

Track performance by pattern, regime (storage/weather/wind), and year. Calculate win rate, profit factor and max drawdown for each pattern. Compare to historical patterns. Optimise entry timing, position sizing and exit rules annually based on results.