Winter Heating Oil, Nat. Gas and Crude Oil Preview 3 comments
-
Font Size:
-
Print
- TweetThis
As we proceed into the winter season, as far as refiners are concerned, the focus will shift from the unleaded gasoline inventories to the residual fuel or heating oil [HO] inventories. When the winter is expected to be a warm one, sort of like what we saw the last few years, HO demand is less than the norm and prices for this product generally lag the price of the crude barrel (crack spread favors crude). When the winter is expected to be a cold one, refiners prepare for the rise in demand of this product. So when things are expected to be warm, lower demand is expected....cold, higher demand is expected. Simple, right?
Well, what happens when the expected does not occur? Two things. First, when the cold sweeps through and is unexpected, more heating oil is used, driving inventories lower until the refiners bring the level back on-line to where supply and demand meet. This drives the price higher rather quickly. Over the past few years with the warmer than expected winters, the price of heat has been leveling off before the winter really picks up. Basically, the months of October and November have determined the action for the rest of the winter for the price in heating oil.
So here is where the real fun begins. I use two weather models in my forecasts for the winter. I don't actually trade off them directly but I use them to better understand why people are buying and selling on the weather during different times of the year. For example, my reports had Gustav coming into the US at about a cat 1 or 2 whereas the market was looking for much larger and greater damage. That did not occur - crude and natural gas dived as a result! Now, if I had only sold short every crude and natural gas contract on the exchange!!! Anyhow I did partake on the dive (as I mentioned in my crude call a few weeks ago).
Ok, moving on to the winter, basically, my two models are indicating it will be colder than normal. There does not seem to be a consensus on the weather this winter online though most services require some sort of payment so getting the whole story is somewhat difficult. Anyhow, one of my sources indicates temps of around 3 to 5 degrees higher than normal for October/November. Interestingly, my trading models have heating oil perhaps finding support around that time period. Furthermore, in terms of residual fuel inventory levels for this time of the year, the level currently resides at its lowest level since 2005. Essentially refiners are not moving into this market aggressively, with the mindset that perhaps we are not seeing a cold winter coming (on top of cascading prices right now not helping things either).
click to enlarge
This all comes together and sets the market up for a surprise. Essentially if we get colder than expected winter weather, demand will be higher for heat. This will put stress on the already low inventory levels and push prices upwards, perhaps putting a floor on crude prices through the winter. Interestingly, year over year inventory levels for natural gas are also the lowest since 2005 at this time of the year - another market perhaps not looking for a colder winter (though electricity demand is playing a roll here... 8% lower y/y from what I have recently read). A cold weather demand push would give both of these products increased demand and thus higher prices.
Disclosure: Trading crude contracts on a daily basis.
Related Articles
|


























This article has 3 comments:
With the decline in crude prices, refining margins have slowed their seasonal decline and actually buoyed from an early August low of 5.78% (crack spread equivalent: $6.92 per barrel).
The crack spread, and other petroleum complex data, are updated every Wednesday in Hard Assets Investors' daily column, "Brad's Desktop." The most recent update is here: www.hardassetsinvestor....
Also, do you know if there is an exchange out there that lists a spread between Natural Gas and heating oil? That ratio is around 45% lower than the hostroical norm and with 85% historic correlation, should revert to the mean over time...