Tuesday, July 28, 2009

Hurricane Watch: Worldwide Accumulated Cyclone Energy Index Approaching 50-Year Low. And: A Risky Insurance Bet.

As promised yesterday in "Bill Gates takes on hurricanes. And: A Quiet Tornado Season":
Two months into a very quiet hurricane season, and following a quiet tornado season*, we are coming into the two most dangerous months for hurricanes. Until we get, at minimum, a tropical storm, I'll be posting on ancillary topics....
Don't go running out to buy the property/casaulty insurers now. That opportunity has passed.
I'm thinking of some sort of pair trade or maybe a straight directional bet, against.
There are three insurance ETF's, unfortunately none are pure-play p/c, each has health/life exposure.
The SPDR KBW Insurance (KIE) has the most life/health, the iShares Dow Jones US Insurance (IAK) the least, with the PowerShares Dynamic Insurance (PIC) in between. Here's the IAK vs. the S&P 500 over the last year. Down harder (think AIG), up faster:
Chart for iShares Dow Jones US Insurance (IAK)

Here's the story that got me thinking about such things, from the personal Florida State University webpage of Ryan Maue:
May - June - July Northern Hemisphere Tropical Cyclone Activity: With 5-days left in the month of July, the three month ACE sum for 2009 is in the lead for the lowest since at least 1970.

Sorted monthly data: Text File

and worldwide:

Global Tropical Cyclone Energy nearing 50-year lows

Figure: Using a longer-database of hurricane tracks for the globe, the recent downturn in global TC energy is nearing record low levels of inactivity - the lowest in 50-years. Full details forthcoming. [updated June 26, 2009]
So why am I thinking of betting against? The Atlantic Multi-Decadal Oscillation has gone positive again after five months of negative anomalies. The wild card is, of course, El Nino which has two effects, wind shear which stops hurricanes from forming but which, if they do form, raises the chances of an East Cost vs. Gulf Coast hit by a couple percent.

Last Thursday the Houston Chronical's SciGuy (Eric Berger) pointed out in "Hurricane season start is slow, but we're not off hook":

...El Niños'intensity

And while El Niños may suppress overall activity, such years can still produce savage storms. One of the three most-intense storms at a U.S. landfall, Hurricane Andrew, developed during an El Niño in 1992.

So have some of the most famed storms ever to strike Texas and Louisiana: Alicia (1983), Betsy (1965) and the great storm of 1900, which came during a severe El Niño, said Jill Hasling, president of Houston's Weather Research Center.

“There might be fewer storms during an El Niño,” she said. “But it only takes one.”

During an average Atlantic season, 10 tropical storms or hurricanes develop, but since 1995 the Atlantic has seen an upswing in activity that most scientists attribute to a long-term natural pattern.

Given this season's slow start and the onset of El Niño, most seasonal forecasters now say about 10 named storms will form, one of the lowest totals of the past 15 years....

The article has this table:


• 1900: Galveston hurricane, severe El Niño

• 1932: Texas hurricane, severe El Niño

• 1943: Galveston hurricane, moderate El Niño

• 1957: Hurricane Audrey, Texas/Louisiana, severe El Niño

• 1965: Hurricane Betsy, New Orleans, moderate El Niño

• 1983: Hurricane Alicia, severe El Niño

• 1986: Hurricane Bonnie, High Island, moderate El Niño

Source: Houston Weather Research Center

That 1900 storm killed 8000 people.

See also his blog post "Let's hope history doesn't repeat itself: The 1932 hurricane"