The difficulty in predicting oil prices

I’m going to stray a bit outside my comfort zone to offer some thoughts on oil prices.

Excluding renewable fuels and transportation, evidence demonstrating a correlation between the price of oil and either renewable energy investment or renewable energy stocks and is somewhat anecdotal. Research demonstrates a link, but others disagree. I imagine it’s a difficult relationship to tease out, given the numerous variables underlying price movement, many unrelated to each other.

[of course, the boom in ethanol, biodiesel and other renewable fuels, as well as the new CAFÉ standards and interest in plug-in and gas/electric hybrid vehicles are all directly tied to the high price of oil. But that’s a different post.]

I’m more interested in the prevalent assumption among the media and general public, that oil prices directly impact interest and/or investment in all renewable energy (for a variety of economic, behavioral and political reasons). I find it an interesting potential paradox that while a long stretch of $100 bbl/oil might have a psychological and economic influence that could ultimately boost renewable energy finance and R&D, there would be considerable negative macro-economic impact in the short-term (which would damage those renewable energy opportunities).

As one of my assumptions for 2008 is that there will be a short-term pause in the “green movement”, I’m specifically interested in the prospect of higher oil prices in 2008 and more importantly, if the outcome can predicted with any certainty.

If the price and volatility of oil drives renewable energy decision-making (as an investor or financier, politician or typical consumer) then having some confidence and clarity about the price of oil would be immensely valuable for a variety of stakeholders.

Thus I found the multiple articles on oil prices in Thursday’s Wall Street Journal to be especially interesting as much for what they said about the staggering impact of high oil prices as for our inability to determine what we might see occurring in the oil industry.

Interestingly the Journal already ran this article two months ago before an oil summit. Back then, I wrote:

If you view $95 oil as the result of speculation, a deflating dollar and a war/risk premium (as opposed to simple supply/demand issues), you can now, according to this article, also add as drivers of high oil prices:

• Resource nationalism
• Limits on oil field access
• Billions of dollars in infrastructure underinvestment
• Declining production levels in existing fields
• Oil field damage due to over-depletion
• Revenue surplus which limits extraction investment
• Labor, construction and equipment bottlenecks
• Booming commodity markets
• Overestimation of near-term potential in alternative supplies

Yesterday’s article took more of a longer-term view, but still offered the same perspective:

Economists, Wall Street commodity traders and even seasoned energy executives were caught flat-footed by oil's dizzying rise. Looking back, several factors came together at the same time to help oil shoot up roughly tenfold in less than a decade and briefly touch $100. Those factors are likely to stick around -- and perhaps push prices up further.

Reading through the article, I was able to add the following to my list of drivers for higher oil prices (responses edited):

• Unexpected surging oil demand in China, the Middle East and other developing countries
• Chronic underinvestment in oil fields in the past decade
• Wall Street penalizing risky extraction efforts
• Insensitivity of U.S. and global oil demand to rising oil prices
  
• Price volatility sparked by 24 hour NYMEX trading, hedge funds and financial traders

In addition, via an email exchange with Lou Grinzo from The Cost of Energy, I confirmed several of these trends and added two other price drivers:

• Rise of national oil companies (NOCs)
• Inability/unwillingness of some exporters to keep pace with demand.

Accordingly, in the process of gathering this information for this post, I ultimately wound up collecting almost two dozen reasons why oil prices are as high as they are today. Which, for me, then begs the question: why has $100 bbl/oil been so hard to predict? And should recent developments really be all that shocking? And in future, can we say with any confidence if these trends and drivers will continue indefinitely?

I feel there is a relatively simple reason why so many investors and analysts have been (and continue to be) caught by surprise with the oil price increases over the past five years. In predicting oil prices over the past twenty years, there hasn’t been much need for innovation in either modeling or research. Oil fluctuated gradually between $20 and $30 (occasionally drifting to $10-15), demand grew consistently and steadily, except in times of recession or the occasional war, and supply was essentially infinite. [as this great chart demonstrates]

To this end, most oil experts and analysts I’ve met tend to be old-school traditional guys, with solid relationships in the Middle-East and a belief that oil supply would forever top demand, and the US would always find a way to smooth pricing through invasion, regime change, sanctions, etc. I have a number of examples which I’m prudently withholding.

Meanwhile, those who predicted peak oil, or that demand could eventually outpace supply have been consistently marginalized, and (for about 20 years) wrong. I would imagine most left the industry. And I would further assume that few young or contrarian individuals ever entered oil trading or analysis, perhaps due to the lack of excitement or change (given the absence of volatility in pricing and need for “physical trading”).

And thus now, as we face a suddenly new and volatile era, we are left with experts who possess a great deal of expertise and experience, only in doing the same basic thing and never really needing to innovate. While I have no doubt as to their economic and mathematical brilliance, we unfortunately need experimentation and non-linear thinking. I imagine it must be paradigm-shattering for some old-timers to have to consider current developments while imagining a world of finite supply and a U.S. unable to “encourage” increasing production. To this end I would expect most oil analysts and investors to have little clue about what to do next, and to continue to be surprised by an oil market that they’ve never really had to consider or model before.

Take for example one senior energy analyst at a large investment bank, who I enjoyed meeting recently. Unlike many of his peers, he's been bullish on oil for several years now. And yet, while he expected high oil prices for 2008, by 2010, he assumes supply constraints will ease considerably, leading to a significant decline of oil prices past 2009.

What struck me in discussing this with him was his reasoning. He ignores most of the 20+ reasons I listed above as to why prices have sparked so high in the past few years and focuses mostly on supply and demand. Moreover, in terms of supply, he thinks OPEC has the capacity to grow up to 3.0 mmbpd, which he believes will easily match average global demand growth of 1.5 mmbpd for the near future.

Beyond this, he has aggressive projections for alternative and non-conventional fuels. He expects ethanol and biodiesel to see incremental growth to almost 0.7 mmbpd by 2010, while Gas to Liquids/Coal to Liquids, oil sands and condensate splitting will add another 1.2 mmbpd by then. This, coupled with additional refinery capacity coming online and an assumption of calming political issues, means that while he expects $80 oil through 2008, after that, he is fairly bearish, and assumes prices will fall significantly.

Unfortunately, to me, this analysis ignores the myriad price drivers described above. And should the OPEC supply growth prove optimistic, the increase in capacity is heavily reliant on the scale up of the alternative fuels mentioned above to match growing oil demand. Meanwhile, forecasting the price of oil, already an immensely difficult exercise given the aforementioned variables in this post, becomes yet more complicated by 2009 if additional oil capacity doesn’t come on line.

In pricing oil, an analyst today needs to consider all of the above challenges and then include a host of new variables associated with the production of alternative fuels (price of corn, tar sand oil extraction, etc). So oil, already incorporating political risk premiums, economic risk and at least two dozen other new variables referenced above, now must factor in variables around the capacity, expense, production and pricing for at least five different unproven fuels, which are necessary to meet growing oil demand.

All of which explains why oil price forecasting is so difficult to get right. And why many "experts" continue to be surprised. And so where does this leave me? Apparently throwing darts and making a blind guess about the average cost of oil in 2008 and beyond.

UPDATE: I only touched on this above, but check out this piece today linking the rising price of oil to a weak dollar, and then some in-depth analysis here.