16 October 2008

Demand Destruction

With oil now hovering around $70/bbl, we've seen a decline of roughly 50 % from the peak this year over only a couple of months. I've stated in the past that I felt the run-up starting in February was largely speculation whereas the push from 2002-2007 was more fundemental. Is this new drop an evidence that 2002-2007 was speculation or is it a fundamental move based on supply and demand? I'm going to argue here that this is, once again, a real move, based largely on demand destruction on-going around the world but in particular inside the United States of America.

Oil consumption in the USA started a downhill roll around December last year. In September, it had a Will-E Coyote moment and rolled off a cliff. It's now in free-fall. Oil demand is highly inelastic in the short-term but their is some phase lag that results in more long-term elasticity. What I mean by that is, it takes awhile for individuals and business to adapt to oil price by drilling for more supply, replacing SUVs with econoboxes, driving slower, etc. But when demand drops fast, the price can drop fast too, because supply is short-term inelastic too.

The latest EIA data estimates US petroleum consumption at 18,865,000 bbl/day. Compared to the same time last year, at 21,024,000 bbl/day, that's a drop of just over 10 % yoy. That's a really big deal. From a GDP-to-oil-consumption relation, it suggests the US is headed into a depression. Even during the 1970s oil shocks the US only experienced drops of 3-4 % a year in consumption.


Figure 1: EIA Weekly Oil Consumption (estimate). The chart shows the story: a gradual downturn and then a sudden drop.

That speculative oil bubble that appeared in February? Gone. Yes, Dorthy, a major recession in the country that consumes a quarter of the world's oil can cause a big drop in price. The USA isn't the whole story, but it is driving this price movement.


The story starts with the OECD countries, and extends to the BRIC (Brazil-Russia-India-China) 'growing' economies. Consumption was pretty flat from 2005-2007, and then it started dropping in 2008. Note that Figure 2 only extends to August, and the US demand dropped over a million barrels a day in September!

Note when looking at these IEA plots, they are all 12-month moving averages, so any moves you see in the curves likely started a few months earlier. Also note that EIA (US DoE) and IEA (France) data are not strictly equivalent, since they use different methods to come up with their estimates.

Oil demand in a number of big economies, e.g. Germany and Japan already fell in response to oil prices last year. They are largely unchanged this year, or indeed up. France is already headed back up, as are a number of smaller countries. In North America, Canada's demand is flat and Mexico is well up.

The big weakness in Eurozone oil consumption is coming from Italy, and the UK. However, their drops are still minor in comparison to the story in the `States. If we look over a longer term view, the IEA consumption data (Figure 5) suggest US demand was flat for 2005-2007, so perhaps this recent drop is just the USA making up for lost ground. Looking at US consumption, and world consumption, the two graphs have very similar shapes.

Figure 5: IEA oil consumption data for USA. Future versions of this graph will show a sharp downturn in September.

The BRIC countries are all less transparent. As a consequence of that, releases of their numbers is even slower than OEDC results we see above that are a couple of months behind. In the same manner that the USA drives OEDC and the other countries are just noise, China drives the BRIC numbers. Unfortunately, we don't have any numbers yet on the Olympic effect in China. Vehicle-kilometers in China was probably way down in August, and that's going to have a short term impact on oil demand. China's skies are once again as smoggy as ever before, so unless they've also changed their stockpiling strategy, this is going to have a short-term impact.

Figure 6: China oil consumption (and projections) beside Vehicle Sales in China. Taken from Malcolm Shealy's (Alacrites Inc.) presentation hosted on IEA website.

So outside of a massive step up in China in August and September (which is unlikely) the only other likely variable to examine is oil supply. Production data (Figure 7) indicates supply is actually up this year. This isn't unexpected. It took a couple of years after oil took off in 2002 for the oil companies to get enough confidence to jump into new projects, but a number of projects started in 2003-2004 are now up and running contributing to supply. While the new oil price may choke off new projects, there's still a number in development to stave off the depletion of existing fields.
Figure 7: World oil plus condensates production, 2005-July 2008 (Data from EIA).

So, in conclusion, I think the supply and demand data show pretty conclusively that supply is up, demand is down, hence the price of oil is dropping like a rock. In fact, we have this event occurring at the same time as a commodity is being popped, so the effect is especially strong. This is a good thing. The world economy and the USA in particular is hurting thanks to the bubblicious real-estate fiasco and low energy prices will help conventional economic activities drag us back out of this big hole. I think the housing bust will be extended — adjustable-rate mortgage resets in the US are only about half-way through the pool of potential defaults — and the world economy is not decoupled from the USA so expect fallout damage to hit exporting countries in manufacturing (China, Japan) and commodities (Canada, Russia, Australia) in waves going forward. The US consumer is going to have to live within his or her means, and hopefully revert back to being a citizen first and consumer second.

Can OPEC squeeze the price by implementing quotas? Yes and no. Production of oil products is up big in 2008. There's slack they can take out of the system. Will they? I don't know. OPEC countries in general have not been diversifying their economies so they will have to do something or face the potential for unrest as revenues come back to earth. Basically it comes down to whether or not Saudi Arabia wants to arrest the fall in price, or if Russia as a non-OPEC producer decides to sacrifice some cash flow or not. If I was the Saudi's, I would be thinking about whether or not OPEC as an organization has outlived its usefulness. A Saudi-Russian oil alliance would probably be more practical, and effective.

Looking into the future, it's likely that demand in the USA will be suppressed for awhile, so as a result, investment in new capacity is going to drop especially in conjunction with higher credit rates. Speculative plays in oil shale, gas-to-liquids, and coal-to-liquids are effectively dead. Remember that in order to develop alternative fossil fuels to oil you need to not just cover the marginal cost of production, but also amortize the capital costs, and the capital costs in non-conventional oil are usually very high. I would also expect new deep water plays (Brazil, Atlantic Canada) to come to a grinding halt, but existing ones should go ahead as the cost of production is not excessive.

The most marginal cost-of-production oil in the world is Alberta bitumen, at around $50/bbl. That's not accounting for capital costs at all, which have been in the range of $150,000 /bbl/day capacity recently. Lower than that, and salaries will have to come down to lower costs. Taking that capacity out of the world's supply would only knock off 1.3 million bbl/day. Alberta's provincial government may be facing the end of yet another oil boom with little to show for it.

1 comment:

Cyril R said...

Figure 2 is extremely misleading, since the vertical axis range only covers around 3 percent of the total absolute range.

Typical IEA graph deception.