You Are Being RedirectedThis blog has been moved to You will be redirected in a moment. If you are not redirected in 10 seconds, click here.

0 Feb 23 2012 @ 10:48am by Matt Smith in Crude Oil, Economy, Global Energy, risk management

From Target to Twitter to Tanks

As companies adopt increasingly sophisticated ways to reach customers or identify trends, they are getting super-sleuthishly good at coming up with predictive indicators. I’ve come across a couple of somewhat wacky ones in the last week, and I thought I would share them (especially as regular readers already know that quirky indicators are the way forward), and show how they coaxed me back to considering a current conundrum in Commodityworld™.

The first example relates to how companies – in this case, Target – can analyze your spending habits to target (no pun) you with specific marketing campaigns. As this recent article in the New York Times highlights, companies across a myriad of industries have analytical teams which study consumer spending trends to identify, and exploit, their potential needs and wants. Perhaps the ultimate example of predictive marketing is how Target can identify if someone is pregnant based on seemingly unrelated purchases – which can lead to its own set of problems (as identified in the article).  

The second soothsaying source is Twitter (let’s not even consider the hidden knowledge of Facebook or Google at this juncture, it is too scary). Twitter is a great example of how sentiment analysis can be used to predict events. Tweets are now scoured to gauge stock market sentiment, and their ability to predict winning stocks has become so accurate that CNBC has a daily feature dedicated to it (‘Twicker Watch’). And there are plenty of other ways to use social media to predict events, from terrorist attacks to election winners.

But therein lies the rub as we head back to Commodityworld™, and the kernel of the conundrum of this post: the (gas) tank.  Based on the examples presented above, given the latest spending habits and the latest data trends we should see that gasoline prices should be falling…but they are doing EXACTLY the opposite.

AAA national daily average of retail gasoline prices

Gasoline demand historically is no different from less conventional economic indicators, such as increased Tylenol usage and higher lipstick sales in hard times, in that decreased consumption coincides with lower economic activity, while higher consumption occurs in times of growth. After all, people with a job need to drive to work, and have the money to buy goods, which are generally shipped across the country by road or rail. But although current economic data appears to show improvement, there is a growing divergence between this and gasoline demand.

US gasoline demand

This can partly be explained away by a number of factors: improving fuel efficiency of cars, people buying smaller cars, less teenagers getting driving licenses, increasing use of public transport, and more online retail purchases. But it is also due in part to an elevated unemployment rate, and a change in behavior – people are being more frugal than before ‘The Great Recession’. And finally, it also has a lot to do with high gas prices – people just not being able to afford to drive as much.

The reason that gas prices are so high is because they are being driven (no pun) higher by rising crude prices – which in turn are being driven (ditto) higher by global influences of geopolitical tension and strong demand from emerging markets. And this is where predictive ability and logic falls apart; despite the fundamentals for gasoline telling us that prices should fall, they continue to remain high, and moving higher.  And even more unfortunate than this is the cure which will lead to gas prices falling…that of higher prices in themselves hurting the global economy until a breaking point is reached.

Leave a comment:

» will not be published