Oh, Crack!

Today’s EIA supply report showed that gasoline stocks fell to 178.74 Million Barrels, the lowest stock ever observed, due in large part to hurricanes Gustav and Ike. According to the Mineral Management service, over 1.2 Mb/day of refining capacity remain shut down, which is roughly a 70% improvement from the 4 Mb/day that was shut down at the peak shut-in. Since Gustav’s approach the first week of September, nearly 50 Million barrels of refining capacity has been lost.

Image: Total Refining Capacity lost due to Gustav and Ike. Note that Ike (Beginning around 9/12) is repsonsible for the majority of the shut-in.

Four refineries in Port Arthur, owned by ExxonMobil, Motiva, Total Petrochemicals, and Valero, remain shut-in with restarts not slated to potentially begin until the weekend. Several other refineries, including ExxonMobil’s massive 567Kb/day refinery in Baytown, have been operating at reduced rates. Valero did, however, recently announce that their Texas City refinery is up and running, cranking out 200 kb/day of refining capacity.

Several weeks of partial shut-ins remain and I expect the total refining capacity lost during September to approach 55-60 Million barrels by the time all of the refineries are back in action. By the same token, I would not be surprised to see gas stocks fall below 170 Million Barrels before they begin their seasonal climb.
Such news would normally be beneficial to refineries since they are now operating at higher than average margins. And indeed, shares of VLO, WNR, and other refineries were all up between 3-5% today. I own VLO and plan to continue holding, short term, with a price target of $40.

One thing I would like to mention is the “Crack Spread.” Quite simply, the crack spread is the difference in price from the refiner’s input (Crude Oil) and their output (various distillates). It is most simply calculated by adding the value of the front month futures contracts for heating oil and gasoline, and subtracting that for crude oil, and multiplying by a standard factor. Generally, the greater the crack spread, the more profit goes to the refiners. In the industry, the crack spread is used to hedge against possible losses from rising input (crude) costs. The price of crude is the refiner’s greatest uncertainty. In order to protect themselves, they can hedge based on the crack spread, buying crude futures and selling/buying distillate calls/puts. There is now a “crack spread” contract on the NYMEX that allows refiners to do something similiar.

Image: 1 year Crack Spread chart. The spike in early sept is due to Ike. Image Source: Bloomberg

The crack spread today is only $7.94, which, while significantly below recent spikes to 20+, is about 15% higher than at this time last year, and more than 30% above this time in 2006. Late Sept/Early Oct has also marked the bottom of the seasonal crack spread trend.

Elsewhere in the energy world, significant amounts of Gulf oil and NG production remain offline. 66.1% and 61.8% of oil and NG production, respectively, remain shut-in, although the rate of restoration has been increasing in recent days.

Image: Latest Shut-ins for Oil and NG. The “days after initial shut-in” resets upon Ike making landfall.

The shut-ins have not caused significant price action among Oil and NG. As has been oft discussed, the driving force of the commodities has, at least temporarily, turned from supply concerns to falling demand. The EIA announced today that demand for petroleum products fell 5.3% from this time in 2007.

The above discussion leads to me be mildly bullish on the commodity sector. I am holding VLO, UNG, CHK, and DGP at the moment. I do not forsee some sudden increase in demand to spur prices, but rather am anticipating seasonal effects and possibly government intervention to bolster these stocks/ETFs.

With market conditions continuing their erratic and unpredictable behavior, it feels nice to return to a place that is predictable: the weather.

Image: Latest satellite image showing a developing tropical low and a broad, extratropical system.

The complex interaction between two different storm systems that I mentioned in my previous post, has pleasantly evolved quite the way it was predicted to. The low pressure area in the Caribbean that I anticipated to become Kyle was put on hold as it moved inland over Hispaniola, but since emerging this morning has begun to re-organize and I still expect it to become a named storm. Like I said on Sunday, its track has been shifted eastward, with a predicted landfall over Massachusetts of Nova Scotia, probably as a modest tropical storm.

Image: Latest computer models for “proto-Kyle”

At the same time, a very large extra-tropical low has formed off the North Carolina Coastline and will slowly move NW over the next two days. While the two will display a weak Fujiwhara interaction and rotate around eachother, as of right now, they should remain two distinct systems. As such, the Northeastern U.S. will experience two distinct bouts of nasty weather, beginning Thursday and lasting through Sunday. Winds of up to 60 mph will buffet the Carolina and Virginia coastlines with weaker winds to the north as the Nor’easter moves onshore. Hurricane-force winds warnings are up for the coastal waters off the Carolinas and Virginia.

Image: Projected rainfall totals by the NAM model showing upwards of 5 inches in Ma and Ct.

Depending on how strong the developing tropical low north of Hispaniola becomes, the Northeast could see winds just as strong this weekend. Heavy rain will also be a concern, with New York CIty and Boston set to pick up possibly over 5 inches. Overall, I believe the event will be comparable to Hanna two weeks ago, although drawn out over about 48 hours, making it seem more intense. Should be a blast.

Image Courtesy: Bill Patterson

Comments »

Or, rather, please RSVP ASAP.

If you live in the Carolinas or the Northeast, pay attention. If you want to bank coin on a hurricane play, go back to watching football.

Image: Latest Satellite Image of “Proto-Kyle” hammering Puerto Rico.

The disturbance that I discussed on Friday has pulled itself together and is on the verge of being upgraded to TS Kyle. I give the system an 80% of being Kyle by this time tomorrow evening. The storm is centered just south of Puerto Rico and will cross the island in the next 12-24 hours bringing heavy rain and gusty winds. St. Thomas, located on the Virgin Islands to the NE of Proto-Kyle, recently recorded a wind gust to 40 mph.

Image: Latest radar out of San Juan, PR, showing a large area of rain SE of the island.

The system will then track to the north into the Atlantic east of the Bahamas over the next 2-3 days. Once it is in that area, things get interesting from a meteorological perspective. There are four possible outcomes and I will briefly discuss each of them.

I.

Image: Potential setup with a developing Nor’easter that does not interact with proto-Kyle.

The least interesting of the four. Most of the reliable computer models are predicting that the season’s first Nor’easter will develop off the Carolina coastline by midweek. If this system tracks the northeast, as most nor’easters do, it would open up a weakness in the Bermuda High and drag Proto-Kyle with it. In this scenario, Kyle would likely pass well east of the Eastern Seaboard with only peripheral effects, like high surf and possible beach erosion, affecting the US. This track is favored climatologically, but has been received with less enthusiasm in recent computer model runs.

II.

Image: Potential setup should Proto-Kyle interact with the developing Nor’easter.

Should Proto-Kyle put on a burst of speed, or this alleged Nor’easter develops to the east of where it is currently anticipated, the closer proximity of the two, will introduce the possibility of a rather unusual phenomenon: The Fujiwhara Effect. Under the Fujiwhara, two cyclones rotate counter-clockwise around a common fixed point located approximately between the two.

Image: Two cyclones displaying the Fujiwhara effect

If one storm is stronger than the other, it can draw the other towards it an eventually swallow it. This is what occurred in perhaps the most famous illustration of the Fujiwhara Effect during the 1995 hurricane season. Hurricane Iris would suck in and swallow Tropical Storm Karen in a form of Cyclonic Cannibalism.

The Fujiwhara Effect generally only occurs when storms are within 900 miles of eachother. Should the developing Nor’easter and Proto-Kyle fit this specification, the former will probably be steered the southwest towards Georgia, while Proto-Kyle is steered towards the Mid-Atlantic. The latest GFDL model run, which follows a Fujiwhara-based solution, takes a Category 2 storm up the Chesapeake Bay late in the week.

Image: GFDL model run showing a cat 2 hurricane approaching the Chesapeake Bay.

As someone who has lived in Virginia for about 12 years, this is a very dangerous track and reason for concern. A storm surge pushed into the Bay will be funneled northwards and pile up as it does so leading to greatly exaggerated flooding. It is far too early to discuss it, but there is a threat to NYC. The City has not had a serious hurricane since the 1938 “Long Island Express,” and a direct hit by a hurricane at any strength would be devastating due to unfortunate topography and development on the coastline.

III.

It is also possible that the Nor’easter will simply not develop as fully as the models are calling for. Anticipating cyclogenesis is not something that the computer models do particularly well. Should it not develop, Proto-Kyle would absorb its latent energy, and while not pushed out to sea, would be pulled northward parallel to the eastern seaboard. The latest GFS shows a very large, very strong storm just east of New England next Saturday.

IV.

It is also possible that wind shear from the Nor’easter will rip Kyle apart. Indeed, two storms in proximity is almost always a lose-lose for both parties involved. Each exerts shear on the other hindering development. In no way will the nor’easter enhance proto-Kyle’s development, at least not in a tropical sense.

In terms of which track is most likely to occur, I would like to introduce the concept of “Occam’s Razor,” which says that of all the solutions, that which is simplest is most likely to occur. Before 2008, I have found that the tropics obey this principle rather well. In this case, the “simplest solution” would be the first one, in which the Nor’easter opens up a weakness in the ridge and Kyle passes out to sea. However, with the very unusual tracks taken by Fay and Ike during the 2008 season, serious consideration must be given to all tracks. Right now, the track most favored by the computer models is a cross between (2) and (3) in which Proto-Kyle tracks northward followed by a bend to the northwest towards the coastline.

Image: Latest computer model guidance for Proto-Kyle.

Overall, I give Kyle a 50% of affecting the US coastline at some point in its lifetime. Personally, I think the track out to sea makes the most sense, and I would not be surprised to see the models shift east in upcoming days. However, until/if such a shift occurs, the east coast should keep a very close eye on the system.

As for its intensity, passage over PR should not inhibit develop significantly. Once it emerges into the Atlantic, the waters are warm enough for steady intensification, and several models are calling for it to reach hurricane strength. Once it moves north of about 30N, the waters cool and shear from the developing Nor’easter will begin to weaken the storm. I give the storm a 60% chance of reaching hurricane strength, but only a 25% chance of making landfall at said intensity.

Elsewhere in the Tropics, many of the models are calling for a tropical storm (Laura) to form between the Lesser Antilles and the African Coastline. Additionally, several are forecasting a tropical low to develop in the Gulf of Tehuanapec in the Pacific, track northward, cross Mexico and enter the Bay of Campeche and develop into a tropical storm.

Note: Today’s Patriot’s team was a fluke. If anybody says anything about Evil Empires and Karma, I will send Kyle right up your weedy, overgrown front lawn. Laugh it up. I dare you.

Image Courtesy: Bill Patterson

Comments »

Two weeks ago, the tropical Atlantic was in a period of unprecented activity. Four storms—Gustav, Hanna, Ike, and Josephine—were active simultaneously. And going back to mid-august, a record 6 consecutive tropical storms or hurricanes made landfall in the United States. Like DiMaggio’s 56, or Chamberlain’s 100, this is a streak that will probably not be challenged for many years.

1) Dolly struck south Texas as a 110 mph Cat 2.

2) Edouard hit east Texas as a strong TS

3) Fay apparently didn’t take too kindly to Florida, striking the state a record four times, all as a strong TS. This is probably another record that will stand the test of time. Just from a mechanics standpoint, it seems nearly impossible for a single state to be struck more than four times.

4) Gustav struck eastern Lousiana as a 110 mph Cat 2, doing significant damage to coastal communities, but sparing NO.

5) Hanna struck the Bahamas, the Carolinas, and the Northeast US as a strong TS, doing little damage, but bringing significant wind and rain.

6) Hurricane Ike made a direct hit on Galveston, Tx, doing over $27B in damage, making it the third costliest hurricane in US history, behind Katrina and Andrew.

Obviously, it was a pretty rough 6 weeks to live in Texas, with two hurricanes and 1 TS making landfall in the state.

However, as quickly as things heated up, the tropics have rapidly quieted back down and there are currently no named storms in the Atlantic, or the Pacific, for that matter.

 

Image: Comparison between the performance of the S&P 500 and the number of threatening hurricanes in the Atlantic showing a (probably coincidental) inverse relationship between the two.

It makes one wonder if there is some sort of tag-team tournament going on between the hurricane demons and the stock market demons. The former gets August and the first half of September to do their thing, and then, tag, the latter get to mop up what is left. Imagine working at a Houston-based fund this past month. Ouch.

 

 

Image: Hurricane Frequency, showing a climatological peak on Sept. 10. Note also a secondary peak in mid-Oct. Image Credit: NOAA

The climatological peak of the hurricane season occurred on September 10^th, a week ago yesterday. This specific date doesn’t really mean too much, but rather that, historically, the SSTs are as warm as they will get and wind shear is at climatological lows. Soon, the former will decrease as autumn begins and the latter will increase as the Jet Stream begins its migration back to the south.

 

Image: Current Atlantic Satellite showing very disorganized convection.

There are no threat areas currently in the Atlantic. A large area of cloudiness extends from the eastern Caribbean into the Atlantic just east and north of the Leeward Islands, but it is diffuse and pressures are not falling. The NOGAPs model suggests that something might spin up in the Caribbean early next week, but none of the other models are following suit. I give this system a 30% chance of developing by next Wednesday as it makes its way into the Caribbean. Just an observation: this large, showery environment is a similar sort of pattern as the one which spawned Hurricane Wilma in 2005.

Image: TOP: Satellite from 10-16-05, showing a similiar pattern of diffuse cloudiness. BELOW: Satellite from 10-21-05 showing Hurricane Wilma, the most intense Atlantic hurricane on record.

 

 

Even in death, hurricane Gustav continues to taunt commodities traders. Today’s NG supply report, which allegedly contained Gustav-instigated “disruptions,” showed an increase in stocks of 67 mmcf/d, well above the 50 anticipated by analysts, causing NG to tank 4%. This data weighed on my energy holdings, giving me a sub-par 3% daily gain.

Image: Latest EIA NG storage report.

Beginning next week, interruptions from Ike will be present and I expect that we will see a much smaller increase in stock.

 

Hurricane Ike—The Epilogue.

I am not going to rehash the meteorological specifics of Ike, given that any news station worth their salt has pretty much got that part down.

In terms of energy interruptions, Ike did considerably more damage than Gustav, destroying 28 of 3800 (0.75%) offshore rigs, and disabling many more. The Mineral Management Service estimates that these rigs are only responsible for 11,000 barrels of Crude/Day and 82 MMcf/Day of NG, so in the long term, this will not be terribly significant. However, over 90% of Oil Production and 70% of NG Production remain offline, are are lagging the restarts of Gustav.

Image: Oil Shut-Ins due to Ike and Gustav. While both had similiar max shut-ins, restarts after Ike have been much slower.

 

 

Image: NG Shut-Ins due to Ike and Gustav. While both had similiar max shut-ins, restarts after Ike have been much slower. However, both are an improvement over oil shut-ins (above).

 

Power outages remain for 2.3M customers, 1.6M of whom are in Texas. While refineries were left largely undamaged by Ike, these power outages are preventing some facilities from restarting, and 3.0 million barrels/day of refining potential remains shut-down.

The death toll from Ike remains at a very uncertain 63. Most of these come from areas well north of Ike’s landfall. Authorities fear that those who perished in the small communities on Galveston Island were washed out to sea.

I will have a detailed discussion of the current commodities situation over the weekend.

Note: Now that I am back on campus, expect to see a statistically significant jump in iBC traffic due to my exquisite marketing skills.

 

 

Image Courtesy: Bill Patterson

Comments »

Image: Kilauea’s Cauldera, Halemaumau, active. Source: Meghan Purdy

Image: Lava flow from Kilauea entering the sea. Pretty awesome. Source: Meghan Purdy

I just got back to Boston after a weeklong field trip in Hawaii and, holy shit, what a terrible week to miss. FRE, FNM, and LEH all go belly-up, both Galveston and commodities get flattened by Ike, and Tom Brady blows out his knee. Crap.

Fortunately, I cleaned out and hedged my portfolio before I left and did not suffer too egregiously, although I am still holding VLO, CHK, and UNG. Covered my LM short this morning at 36.25 for a tidy 20%. Hat-tip: Fly.

The tropics are very quiet right now, but a pattern shift will occur early next week that will make tropical development more likely. I will have a complete discussion of this, as well as a lengthy commentary on the current commodities situation, especially as it compares to 2005, sometime tomorrow.

Image Source: Bill Patterson

Comments »

Played probably 3-4 times per night during the NBA finals last June, played on the soundtrack to The Departed, and on the Simpsons..

#2. “Shipping Up to Boston,” Dropkick Murphys

[youtube:http://www.youtube.com/watch?v=x-64CaD8GXw 450 300]

Image Courtesy: Bill Patterson

Comments »

I am currently in Hawaii on a department-sponsored until Friday, Sept. 12. In celebration of returning to Boston for the year, I am going to be a “countdown” of sorts of Boston-themed music. This is good stuff.

While this is not a song about Boston strictly speaking, it is sung after every Red Sox victory and is as a part of Boston culture as is any Song. Neil Diamond announced last fall that JFK’s daughter, Caroline, who was only 11 years old at the time, was the inspiration for the song.

#3, “Sweet Caroline,” Neil Diamond

[youtube:http://www.youtube.com/watch?v=WzsUOmqpaeg 450 300]

Image Courtesy: Bill Patterson

Comments »