Market Thoughts
Links | Sitemap | Search:   
  Home  > Commentary  > Archive  > Market Commentary  

Should We Be Worried About the 2007 Hurricane Season?

(June 28, 2007)

Dear Subscribers and Readers,

Just slightly over a year ago, I penned a commentary (see our June 11, 2006 commentary entitled “Oil Shortages this Hurricane Season – Last Thing to Worry about”) discussing the potential for either an oil or natural gas production interruption in the Gulf of Mexico during the 2006 hurricane season.  In that commentary, I concluded: “Again, based on current forecasts as well as history, the chances of similar disruptions in oil and natural gas production in the Gulf of Mexico this year is next to impossible.  In fact, based on current forecasts, this author would be surprised if we get ANY significant disruptions this year.  All this is not lost on the folks at the EIA, as they are currently forecasting a cumulative production loss in the range of only 0 to 35 million barrels of oil and 0 to 206 billion cubic feet due to this year's hurricane season.  For comparison purposes, a cumulative 162 million barrels of oil production and 784 cubic feet of natural gas production have been lost since the 2005 hurricane season began nine months ago.”

While a major reason for our 2006 conclusion came from our analysis of the available NOAA (National Oceanic and Atmospheric Administration) and EIA (Energy Information Administration) data at the time, part of it was also due to sheer luck.  As I mentioned in our 2006 hurricane forecast commentary, no one can predict weather-related events with accuracy – especially weather-related events that are more than two weeks out.  However, over the last few years – and especially in light of the freakish events surrounding the 2005 hurricane season – the forecasting tools that have become available have gotten much better in terms of prediction accuracy.  That being said, part of it was definitely due to sheer luck, as the 2006 ACE (Accumulated Cyclone Energy) Index, which measures the collective intensity and duration of all tropical named storms and hurricanes in the Atlantic, turned out to be a lowly 79 reading, as opposed to the range of 118 to 179 (135 percent to 205 percent of the normal level), that both NOAA and EIA were looking for (which would have corresponded to an 80 percent chance of an above-normal hurricane season in 2006).

At the time, however, I also stated that even if the ACE index turned out to be at the high-end of its range (179), it is still lower than the ACE index exhibited during the 1995 (227.2), 1998 (181.6) and 2004 (224.6) hurricane seasons.  As a matter of fact, an ACE index of 179 is still comparable to the ACE index exhibited during the 1999 (176.1) and 2003 (174.5) hurricane seasons – two hurricane seasons which were not accompanied by any crude oil or natural gas disruptions in the Gulf of Mexico.  Moreover, looking at the history of the ACE index, it was obvious that the 2005 hurricane season was a huge outlier, comparable only to the 1950 hurricane season (even then, only one 60 mph tropical storm and a 110 mph hurricane hit the Gulf of Mexico region).  Following is a chart from the most recent EIA publication (“The 2007 Outlook for Hurricane Impacts on Gulf of Mexico Crude Oil and Natural Gas Production”) showing the North Atlantic ACE Index vs. the number of tropical cyclones for each hurricane season from 1950 to 2006:

Atlantic Basin Hurricane & Tropical Storm Activity

Of course, the question on all our subscribers' minds is now: That is all well and good, but should I now be worried about a potentially active 2007 hurricane season, both from an investment and a physical point of view (especially if you live in Florida)?

The short answer is: Yes, you should be concerned, but again, the 2007 hurricane season is definitely not going to be as disruptive as the 2005 hurricane season.  Of course, anything is possible.  Moreover, history has shown that it has nearly always been better/more profitable to not bet on a disaster or “end of the world” scenario.  But I am a numbers person and am very fond of probability at heart, and the latest NOAA and EIA 2007 hurricane forecasts suggest that I should be concerned, even though any shut-ins of crude oil or natural gas production should not have a lasting impact going forward.

Before I dive deeper into the subject, let me give you a little bit of a background.  Per the EIA, the Gulf of Mexico is a very important source of both crude oil and natural gas for the United States.  During 2006, crude oil production accounted for 27% while natural gas production from the region accounted for 15% of total U.S. production.  That is why so many commentators and analysts are concerned about any adverse impact that severe weather may cause in the region during the 2007 Hurricane Season.  Moreover, according to the EIA: “Atlantic tropical storms and hurricanes that impact the United States usually follow one of three general paths: along the Eastern seaboard, across Florida and striking the Gulf Coast, or through the Caribbean Sea and moving north to strike the Gulf Coast.  About one-third of Atlantic storms eventually pass through the Gulf of Mexico.”  Finally, offshore operators must evacuate personnel and shut down production if severe weather threatens – thus at least temporarily disrupting the flow of both crude oil and natural gas to the mainland.  The following chart shows the paths of major hurricanes in the Gulf of Mexico area from 1995 to 2006 – as well as showing the intensity of each hurricane and at what points:

Major Hurricanes in the Gulf of Mexico, 1995-2006

According to NOAA, the Atlantic Basic is projected to experience above-normal hurricane activity this year, with 13 to 17 named storms, including 7 to 10 hurricanes – 3 to 5 of those which will be “intense” (i.e. Category 3 or above).  This corresponds to a 75% probability of an above-average hurricane season in the Atlantic Basic, which is also likely to correspond to above-average activity in the Gulf of Mexico.  Based on the EIA's Monte Carlo hurricane outage simulations, the EIA expects “a total of about 13.2 million barrels (bbls) of crude oil and 86.5 billion cubic feet (bcf) of natural gas to be shut in during the 2007 hurricane season.”  Moreover, the EIA's simulations “indicate a 1.3-percent probability of more than 100 million bbls of crude oil and/or 600 bcf of natural gas being shut-in during the 2007 hurricane season, similar to the cumulative impact of Hurricanes Katrina and Rita in 2005. Conversely, we project a 2.2-percent chance that offshore Gulf of Mexico production will be unaffected this year as it was during last year's hurricane season.”

But Henry, like you had mentioned before, isn't the weather almost impossible to predict?

Yes, and this author also flinches whenever someone starts talking about doing Monte Carlo simulations for forecasting purposes – especially from a weather or financial market perspective.  That being said – even though the current tools are relatively crude – these are the only ones we have.  For example, the EIA estimated the number of hurricanes in the Gulf of Mexico this year by taking the 1950 to 2006 average and applying a ratio – calculated by taking this year's predicted number of hurricanes in the Atlantic Basic divided by its 1950 to 2006 average.  As a result, the corresponding estimate number of hurricanes in the Gulf of Mexico this year is: 1.2 intense hurricanes, 1.2 moderate hurricanes, and 2.7 tropical storms.  As for the uncertainty of these forecasts, this is being partially reflected by the large standard deviations of estimated shut-in production as calculated by the EIA under different scenarios, as shown in the following table:

Estimated Shut-in Production by Type of Weather System, 1995-2006

Meanwhile, the following table shows the EIA's various projected shut-in scenarios and the probabilities that are assigned to each scenario, for both crude oil and natural gas:

Simulated Cumulative Shut-in Production for 2007 Hurricane Season

Assuming the worst case scenario (95th percentile, again, readers should take any simulation that is done by utilizing a normal distribution with a grain of salt), the estimated cumulative shut-in production due to the 2007 hurricane season of 72 million barrels of crude oil and 431 bcf of natural gas still compares favorably to the 2005 hurricane season, when the Minerals Management Service reported a cumulative shut-in production of 109 million barrels and 561 bcf through the end of 2005, respectively.

At this point, you may say that this is basically a crapshoot.  And basically, it is, but it will be wrong to assume that this year's hurricane season will be like last year, due to the following two important factors:

  1. We know that because of the 2005 hurricane season, many oil and natural gas drillers have rebuilt their platforms and pipelines with stronger infrastructure, which should make them better able to withstand any upcoming hurricanes and storm-related events.  What we do not know, however, is how these structures will fare in Category 4 or 5 winds – even though we know that the engineers' models have “told them” that these new structures should stand up to much tougher winds compared to the old infrastructure.

  2. Since 1995, the Atlantic Basic has been experiencing above-average hurricane activity – a pattern which hurricane researchers refer to as the “multidecadal cycle.”  This is a major reason why forecasters are expecting above-average hurricane activity to continue for another five to ten years.  More importantly, unlike last year's hurricane season, this year's hurricane season is expected to coincide with La Niña conditions.  As discussed by NOAA: “The combination of an active Atlantic era and La Niña is known to produce very active Atlantic hurricane seasons. The high likelihood of a La Niña-like influence on the 2007 Atlantic hurricane season is in complete contrast to last year, when a rapidly developing El Niño during August-September contributed to a shut-down in hurricane activity during October and November (Bell et al. 2007).”

Point number 2 is especially important, as it signals that any surprises from the NOAA and EIA's median forecast should be on the upside.  If anything, any higher-than-expected activity in the Gulf of Mexico during this year's hurricane season should be a good test for the new drilling infrastructure that has been installed since the 2005 hurricane season.  While some of this newer equipment should no doubt be able to withstand much stronger winds then ever before, this is most probably not true for those folks who may have homes in the hurricanes' paths this year.  For folks who reside in parts of Florida or the Gulf Coast that are especially susceptible to hurricanes, I strongly urge you to be careful this year.

Signing off,

Henry K. To, CFA

Article Tools

Subscribe to this FREE commentary

Discuss this page

E-mail this page to your friends

Printer-friendly version of this page

  Copyright © 2010 MarketThoughts LLC. | Privacy Policy | Terms & Conditions