London, 9 August 2010

Prices for crude oil and products may have been boosted by the existence of a "hurricane premium," according to some analysts, helping keep the market buoyant despite rising reported inventories of crude oil and refined products in the United States. The market is still haunted by the memory of the unusually active and damaging 2005 season, when hurricanes Katrina (August 2005) and Rita (September 2005) caused extensive damage to offshore oil and gas installations and onshore refineries on the Gulf Coast.

Long positions in crude and gasoline contracts expiring between September and December 2010 are essentially call options on the possibility of a hurricane taking out offshore platforms and pipelines or onshore refineries, triggering a substantial loss of production and a price spike. Even the fear of a hurricane tracking into the danger area might be enough to propel prices higher and create substantial profits.

ACTIVE SEASON FORECAST

The U.S. National Oceanic and Atmospheric Administration (NOAA) has predicted an unusually active hurricane season this year, which has made explicit or implicit optionality in these positions more valuable than normal. NOAA is currently forecasting 14-20 named storms this year (with top winds of 39 miles per hour or more); including 8-12 at hurricane force (74 miles per hour or higher); of which 4-6 could be the most destructive type of Category 3-5 Major Hurricanes (at least 111 miles per hour). The forecast includes the three storms that have already occurred (Alex, Bonnie and Colin).

The number of storms has been cut slightly from NOAA's previous forecast issued in May. At the time NOAA Administrator Jane Lubchenko warned it could be one of the most active on record. But it is still more active than usual. On average, the season has seen just 11 named storms, 6 reaching hurricane force, and only 2 major hurricanes.

LIMITED TIME TO EXPIRY

The main Atlantic hurricane season runs from June to November, with activity peaking between late August and early October. Like any option, the hurricane season calls have an effective expiry (October) and become less valuable the closer to expiry they come without a major hurricane forming and heading in towards the Gulf Coast.

If a hurricane premium has propped up crude oil and products prices in recent weeks, the effect will start to reverse and put downward pressure on prices from mid-September onward as the call option approaches towards expiry, unless a major hurricane strikes or threatens the coastal areas in the meantime.

The hurricane premium should lose a small amount of " time value" each day between now and the end of the season a major storm does not strike. While the daily loss of time value will be quite small at present, it will accelerate in the second half of September as the peak of the season passes if no hurricane has hit.

MAJOR STORMS, ON TARGET

The second key issue is whether the market is accurately assessing the probability of a damaging hurricane strike. While tropical storms and hurricanes are common, only a few enter the central and western parts of the U.S. Gulf. Only a very few of those reach the devastating intensity of a major hurricane (Categories 3-5) that poses severe danger to the petroleum complex.

The NOAA forecasts are for all tropical storms and hurricanes forming in the Atlantic Basin. Only a few of these will make landfall in the United States, and only a minority of those will strike or threaten the central and western U.S. Gulf areas and refineries along the Texas and Louisiana coast.

The charts (download below) are taken from NOAA's National Hurricane Center and show the tracks of major hurricanes making landfall in the United States in 1981-1990 (Chart 1), 1991-2000 (Chart 2) and 2001-2005 (Chart 3). Chart 3 includes the devastating hurricanes Katrina and Rita in 2005. But note how infrequently major hurricanes hit the relevant part of the coast. In the 25 years up to and including 2005, only 17 major hurricanes made landfall in the United States. Of those only 8 menaced the central/western U.S. Gulf: Alica (1983); Elena (1985); Andrew (1992); possibly Opal (1995); Ivan (2004); and finally Dennis, Katrina and Rita (all 2005).

We should add to the total major hurricanes Gustav and Ike (August and September 2008) (Chart 4). The point remains that hurricanes at Category 3 or above and hitting the right area to damage oil and refining infrastructure are quite infrequent.

Of course the probability of a major hurricane hitting the relevant part of the U.S. Gulf is higher in an active season than a quiet one. Even so, the market may be giving too much weight to the experience of 2005 and 2008, which were unusual years, rather than the other 27 years in the 1981-2009 distribution.

Is this another example of investors and some traders focusing too much on the tails of the distribution rather than the central tendency? If a hurricane did form in the right area and track towards the central or western Gulf, oil prices would spike higher. Because hurricane tracks are subject to so much uncertainty more than 24 hours ahead, even a distant hurricane can have an impact. But any rally might not last long and be quickly reversed if there are a lot of options already built into the market and eager to take profits before the storm passes.

HARDENED INFRASTRUCTURE

One final issue is the impact any major hurricane would have on oil and gas infrastructure in the event of a direct or even tangential hit. If it was at major risk of failure, it is reasonable to assume much of the older and weaker infrastructure would already have failed during the 2005 and 2008 storms. What remains should be sturdier. It has already been tested in the balance.

Moreover, after the impact the 2005 and 2008 hurricanes, offshore operators and refiners have been hardening their installations to make them less liable to fail. Short-term precautionary shut-ins may become more frequent as producers and refiners become more cautious, but the impact of that on prices, given high stocks and surplus production capacity, is likely to be limited, even if it adds to short-term volatility. Long unintended shut-ins of the sort that happened in 2005 may be less likely to occur.

But, if investors are hoping a hurricane will take a significant amount of crude production and refining capacity offline, to help work down the massive build up of crude and refined products being held in commercial storage in the United States, they could be disappointed.

A long futures position in crude or gasoline, or an explicit call option, is in effect a deep-out-of-the money call on a hurricane strike on the central and western U.S. Gulf. Such options for low probability high consequence events are notoriously hard to value properly. The prevalence of current talk about a "hurricane premium" indicates the market may be assigning a relatively high value in practice. But the premium will start diminishing rapidly if no hurricane has hit by the middle of next month.

Ends --

By John Kemp, Reuters market analyst - for Commodities Now.

The views expressed are his own.

Attachments:

2008atl.jpg	504 Kb
DC_14_1981-1990.pdf	209 Kb
DC_15_1991-2000.pdf	120 Kb
DC_16_2001-2005.pdf	135 Kb