twacowfca Posted October 11, 2009 Posted October 11, 2009 In Aug 2006, we bought FFH Oct 2006 calls on the theory that the hurricane season would continue to be benign because hurricanes follow a fractal pattern rather than the normal distribution that guru forcasters assume. When the stock ran up, we cashed out and used the profits to buy leaps. Result: a 30 bagger :) We are not tempted to try this with SHLD. Unlike FFH, SHLD appears to be stuck in a toaster. Walmart put them into Cpt. 11. Next episode: Cpt 22
Munger_Disciple Posted October 12, 2009 Posted October 12, 2009 Can you explain how the fractal pattern theory predict the severity of hurricanes?
twacowfca Posted October 12, 2009 Author Posted October 12, 2009 Hurricanes and stock markets have a distribution that isn't normal or bell shaped. There are periods of low incidence or volatility punctuated by shorter periods when everything goes bonkers. It's hard to predict when these "regime changes" will occur, but the important thing to keep in mind is that whatever phase you're in is likely to continue. There isn't a smooth transition and regression to the mean is jerky. In fact the very idea of a mean is flawed. It's better to think of each phase as having it's own mean. The most important observation is to realize that whawever phase you're in is likely to continue for a time that may often be estimated by reference to historicle records. Therefore, the volatility or severity of that period is highly correlated with recent events and very little with the events of the previous regime immediately before it.
twacowfca Posted October 12, 2009 Author Posted October 12, 2009 Here's more about how this idea of the fractal pattern of hurricane distribution in the Gulf of Mexico and western Atlantic has worked very profitably. 2005 was the worst hurricane season ever: 6 major hurricanes making landfall, including Katrina. Property rates in exposed areas and natural disaster retro rates increased by a factor of 6. We wanted a piece of this but we didn't want to get burned,so we waited to see how the 2006 season developed. By late August, the season was a zero, nothing! Every storm that started to develop or come close to the US had it's top blown off or was blown away from the US by contrary winds. This was the type of pattern of low risk associated with an El Nino but the hurricane forcasters had not detected an El Nino and were still predicting a very bad season in ignorance of the implication that a fractal regime change had likely occurred. We decided to act. We bought large amounts of Lancashire Holdings because their new CEO, Richard Brindle, had had by far the best record at Lloyds of London returning @ 26 percent per annum to the Names who invested in the syndicates for which he was chief underwriter in the 80's and 90's. His forte is assessing catastrophe risk/reward. We also bought FFH calls as detailed in the last post. These were dirt cheap because of the short attack against them. The shorts thought that another bad hurricane season was likely and that this would finish them. We thought this unlikely. The rest is history! PS: We did the same thing this year too,for the same reason, adding even more to these two great companies that now amount to about two thirds of our portfolio. :) a bad hurricane would finish .
Munger_Disciple Posted October 12, 2009 Posted October 12, 2009 Given that "regime changes" in volatility/severity are almost impossible to detect, it seems that it is unlikely that one can successfully predict reliably with this fractal theory.
twacowfca Posted October 12, 2009 Author Posted October 12, 2009 Yes and no. It can't be predicted with very high probability that a regime change has occurred until you can look back in time and see the demarcations between the regimes. However, it makes sense to recalibrate probabilities when a regime change MAY have happened. With hurricanes in the Gulf & W Atlantic, a sudden shift from high incidence in one season to zero or ultra low incidence in the first half of the next season is the hallmark of a dramatic event, an El Nino effect. Therefore, in 2005, by mid season, there were two possibilities that were becoming increasingly probable with each passing day with no hurricanes threatening US shores : a regime change to a more benign period of several years ( typically 12 to 17 years) and or an El Nino effect that may last for no more than one year. When the 2005 season was almost over, the forcasters finally lowered their probabilities after they had firm evidence that there had been an early arrival of an El Nino.
twacowfca Posted October 13, 2009 Author Posted October 13, 2009 Correction! In my previous post when elaborating on fractal patterns and the El Nino effect I referred to the year 2005. This was in error. 2006 was the correct year when the hurricane count plunged. My bad.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now