I prefer my own analysis of the raw numbers. If you use the EIA's weekly millions of barrels supplied(demand) numbers for gasoline going back to 2000,you find that November ranks 7th out of the twelve months. August rating a 1, September a 9, and October a 5. This is done by grouping every week that ends in that particular month and averaging them.

Now if you compare the average for the month of November against that of November the year previous, this is what you find. Between 2000 and 2004, the year-on year average for November increased in the range of 1.27% in 2003 to 4.10% in 2001. The first three weeks of November 2005 show an increase of only 0.13%.  If one were to look at the trend from the first and the second weeks of the month this year - it is a distinct possibility that year-on-year demand this month will be negative.

It was down in October. And it was down a whopping 3.52% in September. The September number appears to be the largest monthly drop in the last 5 years.

So what do these numbers show? That demand is increasing slowly from the point of the Hurricane problems, that this demand is relatively low by historical standards of at least the last 5 years, but that the numbers for November are completely within the range of normality.

Demand typically drops off a cliff in September, rebounds in October, and then dips a little in November.

In other words, absolutely nothing to be excited about - for any reason. I would hold off and try to look at some longer-range trends.