Oil Market Report: 10 June 2008

Download full reportDownload tables


  • Prices surged to almost $140/bbl on 6 June following comments by an Israeli official that an attack on Iranian nuclear facilities was "inevitable" and came against a tight supply background with no clear sign of the usual second-quarter crude oil stockbuild.  The rally wiped out three weeks of retrenchment as distillate premia eased and non-OECD price subsidies were cut.
  • Global oil product demand is expected to average 86.8 mb/d in 2008, 80 kb/d below last month's estimate, following the reduction of price subsidies in several non-OECD countries.  Global growth is cut even more steeply by 230 kb/d to +0.9% or +800 kb/d when historical upward revisions to 2006 and 2007 data are factored in.
  • Global oil supply rebounded by 490 kb/d in May to average 86.6 mb/d, lifted by higher OPEC crude supply.  The rise however comes after extensive downward revisions to 1Q08 non-OPEC production and lower biofuels and NGLs for the rest of this year.  Despite this, a recovery in non-OPEC output is forecast for the second half of 2008.
  • OPEC May crude supply averaged 32.3 mb/d, 395 kb/d above April, on higher output from Saudi Arabia, Nigeria, Angola and with Iraqi output at a six-year high. Higher output and field commissioning delays push effective spare capacity below 2 mb/d.  The call on OPEC crude and stock change in 2008 is revised up 300 kb/d to 31.6 mb/d.
  • OECD oil stocks fell 8.1 mb in April to 2,562 mb, in stark contrast to the typical build.  An 11 mb draw in US gasoline stocks removed the large 1Q08 surplus while crude and distillate cover tightened in Europe and North America.  Total oil cover remains above average at 53.4 days.
  • Global refinery throughput increased by 0.2 mb/d in May to 73.3 mb/d, as strong US, Russian and Middle Eastern crude runs more than offset the decline in Chinese and European throughputs.  Non-OECD regions could also drive 3Q08 global crude throughputs to 75.7 mb/d, 2.1 mb/d higher than 2Q08 and 1.0 mb/d higher than 3Q07.

Market overview: supply push - demand pull

Oil has surged to almost $140/bbl following comments by an Israeli minister that an attack on Iran is "inevitable" if Iran maintains its enrichment programme, firmly bringing Middle East politics back into the price formation process after a considerable hiatus.  Of course, speculators have been involved in this price jump, but given the subject matter it is probably better defined as being risk management rather than speculation.  Although discussions will focus on whether or not a supply disruption is a likely outcome, from a market perspective, there is also another supply response to consider:  an IEA strategic stock release.  Given that OPEC countries are running close to flat out, the market can take comfort that the IEA is watching developments very closely and is prepared to act quickly if necessary.

But $140/bbl is not just about geopolitical risks - the supply situation remains tight.  This report revises down combined non-OPEC plus OPEC NGL supply growth for the year to just 770 kb/d, with most of the growth taking place in the second half of the year.  On top of that, with higher OPEC supply this month and delays to the Saudi Khursaniyah project, OPEC spare capacity fell below 2 mb/d in May for the first time since 3Q06.  Projecting forward current supplies implies a 350 kb/d global stockbuild in the second quarter, but this would still be considerably smaller than the 900 kb/d seasonal norm seen in the OECD alone.  But come the third and fourth quarters, our projections show an unusual fall in the call on OPEC as non-OPEC supplies increase.  So with the market starting to focus on deliveries for July and beyond, should we expect an easing of the current tight balance?

Third-quarter supply growth comes from a number of areas.  Signs that Canadian upgrader maintenance is concentrated in the first half of this year effectively boost 3Q supply growth.  More output is expected from Lukoil and the Sakhalin 2 project in Russia and small additions are seen from Brazil.  However, there are risks.  Maintenance problems can occur and the hurricane season hits its peak in the third and fourth quarters.  We have an allowance in there of 340 kb/d for peak-month storm disruptions, but cannot ever hope to fully offset the seasonal risks.

This report further lowers demand growth for 2008 to under 1 mb/d, at 800 kb/d, following the decision by several non-OECD countries to reduce subsidies in the face of record oil prices.  There has, however, been little change to absolute demand following historical revisions to non-OECD countries that lift the 2006 baseline.  But could demand be weaker?

Every day there are reports about the impact of high oil prices.  Airlines are cutting flights.  Consumers are rushing to buy more efficient vehicles and car manufacturers are slowing SUV production (as long as SUVs are being produced, they will be discounted to clear in the market place).  Public transport volumes are increasing and vehicle-miles travelled are falling, with supporting evidence from toll receipts.  Consumers are protesting and politicians' statements reflect that mood.  Changes are happening in the OECD, but they will take time to filter through.  However, there are very few signs of slowing demand in non-OECD countries where GDP growth is far more significant than price in determining demand.

In reality, these abnormally high prices are largely explained by fundamentals.  Supply growth so far this year has been poor and higher prices are needed to choke off demand to balance the market (and if so, then absolutely the worst response is to subsidise prices more, or in the case of the OECD, to cut taxes).  If the supply shortfall is in diesel, which is indeed the case, then more targeted technology is also needed in the refining system.  This is a case of supply and demand pulling in opposite directions to push prices higher.  Balances may ease slightly this summer, but it is unlikely that this will prove to be the end to current market tensions.  However, the IEA will seek to provide more answers in early July in its next Medium-Term Oil Market Report - examining in detail supply, demand, refining and biofuel projections though to 2013 and the World Energy Outlook in November through to 2030.



  • Global oil product demand is expected to average 86.8 mb/d in 2008, -80 kb/d below last month's estimate but with annual growth falling to +0.9% or +0.8 mb/d following upward revisions to 2006 and 2007 data.  These revisions are driven by a combination of 2006 baseline adjustments in non-OECD countries, downward revisions to preliminary 1Q08 OECD data, and slightly lower projected growth in Asia following adjustments to administered prices.
  • OECD oil product demand has remained largely unchanged in 2007, but has been lowered by almost 130 kb/d to 48.6 mb/d in 2008 (-0.9% or -0.4 mb/d over 2007).  The revisions result mostly from adjustments to preliminary data in 1Q08.  In OECD Europe, the 1Q08 weakness in oil product demand was partly related to the fact that Easter (when demand normally rises) occurred in April this year, rather than in March.  In OECD Pacific, 1Q08 demand was somewhat distorted by Japan's political tussle over the gasoline tax, which depressed gasoline demand in March (but boosted it in April).  However, Japanese demand for residual fuel oil and direct crude for power generation remained strong given the continued woes of the country's nuclear sector.  In OECD North America, 1Q08 demand was much weaker than expected as a result of the economic slowdown and high prices.

  • Non-OECD oil product demand has been revised up by roughly 160 kb/d to 36.9 mb/d in 2007 (+3.8% or +1.3 mb/d over the previous year) and by about 50 kb/d to 38.1 mb/d in 2008 (+3.4% or 1.2 mb/d).  Although the changes are related to final 2006 baseline reappraisals across all regions, they were essentially driven by India, suggesting oil demand in that country has been stronger than previously thought.  Yet these revisions were only partially carried through to 2008, given the fact that several Asian countries - India, Indonesia, Malaysia, Sri Lanka and Taiwan - have implemented adjustments to their administered price regimes, which should slightly tame oil demand growth in that region.  The lull in oil demand growth, however, may only be temporary as strong economic growth remains the key driving force.  Moreover, in China and the Middle East the prospects of a substantial modification of subsidy schemes look remote.
  • The current oil price rally could impinge upon growth prospects if sustained over time, even though the global economy is more resilient - the 'oil burden' (oil expenditures relative to GDP), albeit rising, has not yet reached its early 1980s peak, while the world's oil intensity has markedly decreased.  Globally, the high oil price is contributing to inflationary pressures.  In the OECD, it may postpone the recovery of the US economy, while in some non-OECD countries the cost of imported oil and/or subsidies is becoming unbearable.  However, whether high prices destroy demand on a large scale will depend mostly on whether China and the Middle East, which account for almost three-quarters of global oil demand growth, substantially modify their administered price regimes.  In the OECD, meanwhile, oil demand is already falling.


According to preliminary data, total OECD inland deliveries (oil products supplied by refineries, pipelines and terminals) rose by 1.3% year-on-year in April, with gains in Europe and the Pacific offsetting losses in North America.  In OECD Europe, demand jumped by 4.8%, largely because Easter holidays boosted distillate deliveries and as temperatures returned to normal.  In OECD Pacific, demand rose by 0.7%, as Japanese gasoline demand surged given the temporary abolition of the gasoline tax, while demand for residual fuel oil and direct crude for power generation remained strong.  Oil product demand in OECD North America (which includes US Territories) fell by 0.6%, weighed down by the economic slowdown and ever higher oil prices.

Furthermore, March data were revised down in all three regions, more significantly in Europe.  As a result, 1Q08 has been lowered by 210 kb/d, thus impacting the 2008 forecast.  OECD demand is now seen averaging 48.6 mb/d in 2008 (-0.9% or -450 kb/d over 2007, and 130 kb/d lower than estimated in last month's report).

North America

According to preliminary April data, oil product demand in North America (including US Territories) contracted by 0.6% year-on-year, given the continued weakness in most product categories in the US.  The contraction in regional deliveries was steep in the case of high-sulphur (heating oil) gasoil (-23.1%, although this is related to reclassification issues in the US, since low-sulphur non-road, locomotive and marine gasoil is now counted as diesel) and naphtha (-22.2%).  As such, OECD North America demand is expected to average 25.0 mb/d in 2008 (-1.9% when compared with 2007).  Given the economic slowdown in the US and the continued rise in international oil prices, transportation fuels demand is expected to decline in 2H08, thus adding to the bleak demand picture.

Adjusted preliminary data indicate that inland deliveries in the continental United States - a proxy of oil product demand - contracted by 1.1% year-on-year in April and by 2.7% in May.  All product categories bar LPG and gasoil registered losses in both months. Demand for gasoline and jet fuel/kerosene use, which are highly sensitive to economic conditions and, increasingly, to price, shrank for the sixth consecutive month.  Although temperatures were somewhat colder (both compared to last year and to the historical average), the overall demand contraction is clearly related to the ongoing economic slowdown amid ever higher oil prices.  In light of these figures, US oil demand is expected to fall by about 2.5% to 20.3 mb/d in 2008.

More interestingly, even if economic conditions were to improve sharply, it is unlikely that US demand - largely driven by transportation fuels, notably gasoline - would rebound sharply.  Indeed, the US seems to be entering a 'post-Hummer' period - the gradual switch away from SUVs and light trucks to smaller, more efficient vehicles, largely prompted by the perception that oil prices will remain high for the foreseeable future.  As such, the US could be following the consumption patterns observed in Europe and Japan, namely stagnating demand for transportation fuels as a result of steady efficiency gains.  Although pickup sales may strengthen as the economy rebounds (given their use as commercial vehicles), the overall fleet is poised to become more efficient over the next decades - and even more so if more stringent legislation regarding fuel efficiency and emissions were to be approved in the next few years.

Anecdotal data seem to support this hypothesis.  In April, almost one in five vehicles sold was 'compact' or 'subcompact' (a decade ago, when SUVs were at their peak, only one in every eight vehicles was small).  By the same token, sales of four-cylinder engines surpassed six-cylinder models for the first time ever.  Moreover, SUV sales were 25% lower in January-April when compared with the same period last year.  More symbolically, in early June General Motors announced the gradual closure over the next two years of four North American plants specialising in SUVs and light trucks production as domestic sales continue to plummet (-27.5% in May).  The car manufacturer also hinted that it may sell its Hummer brand - the embodiment par excellence of large, gas-guzzler SUVs - and confirmed that it will launch its Volt electric car in 2010.


According to preliminary inland delivery data, European oil product demand in April rose by 4.8% year-on-year.  This was 504 kb/d above our previous estimate, as demand in the largest countries turned out to be much higher than expected (+150 kb/d in both France and Germany, +60 kb/d in Poland and +50 kb/d in Italy).  Over two-thirds of this adjustment was related to middle distillates (diesel was up by 150 kb/d and heating oil by 200 kb/d), confirming anecdotal evidence of a tight physical distillate market that pushed distillate cracks to record highs.  Moreover, April temperatures were much colder than last year and slightly below the 10-year average.  By contrast, March estimates were revised down by 205 kb/d following the submission of official oil statistics.  As such, 1Q08 European oil demand is now estimated at 15.1 mb/d, 0.4% lower than the already weak 1Q07.  It should be noted, though, that this year Easter fell in April rather than in March, thus shifting some demand to 2Q08, which has been revised up by 164 kb/d to 15.1 mb/d or +1.4% over 2Q07.

Inland deliveries in Germany were particularly strong in April, posting a 10.9% annual increase, albeit from a low 2007 base.  Diesel and heating oil deliveries recorded the largest gains, growing by 11.6% and 80.8%, respectively, while gasoline and fuel oil continued to contract in line with earlier trends.  Heating oil demand, in particular, was back at its five-year average level for the first time since January (and before that, since October 2006).  Nevertheless, heating oil consumer stocks continued to slide to record lows, averaging 45% of capacity by the end of April, from 46% a month earlier and 53% a year before.  Consumers are expected to start filling their tanks during the summer months, as demand weakens and prices drop, but the exact timing of the purchases depends on price developments.  As in recent years, this could have a large impact on the market since German demand represents more than 20% of Europe's heating oil market.

Demand in France grew by 8.1% year-on-year in April, with all products registering gains (with the exception of gasoline, which contracted by 7.7%). As in Germany, demand for middle distillates increased due to Easter.  Deliveries of heating oil were 44% higher last year, at the upper end of their five-year range, as temperatures were slightly colder than the 10-year norm, while diesel and jet kerosene posted gains of 6.6% and 4.5%, respectively.  Interestingly, despite loud and widespread public protests against ever higher fuel prices from fishermen, farmers and truckers, diesel demand has held up surprisingly well in the first four months of this year (+2.3% over the same period in 2007).  Industrial fuel oil deliveries, meanwhile, were 6.7% higher than last year, in line with stronger economic activity (the industrial production index was up 2.1% in 1Q08).

In Italy, demand rose by 3.5% year-on-year in April, pulled up by diesel (+8.1%) and heating oil (+9.9%) demand.  Residual fuel oil demand rose by 5.1%, supported by higher domestic power generation needs.  According to Terna, the Italian electricity grid operator, Italian electricity demand increased by 3.4% on a yearly basis in April and by 0.6% over March.  Domestic generation jumped by 8.5%, as net imports fell by 23%.

Demand for transportation fuels in the UK was much lower than expected in March (the last month for which data are available), leading to a downward revision of some 120 kb/d; gasoline demand contracted by 19.6% year-on-year, while diesel consumption fell by 8.4%.  Overall, total demand was down by 6.2% in March and by 4.2% in 1Q08.  Demand for all products bar LPG should rebound in April (LPG weakness is likely related to the shutdown of the Grangemouth petrochemical plant).  Finally, Spanish demand rebounded in April (+1.3% year-on-year) due to higher gasoil demand (+5.0%).  By contrast, demand in March was particularly weak (-6.0%), dragged down by gasoline (-8.5%), diesel (-9.3%) and heating oil (-15.1%), due to warmer-than-normal weather.


Preliminary data for April indicate that oil product demand in the Pacific rose by 0.7% year-on-year.  As expected, the rise was largely driven by Japan, where gasoline demand soared following the temporary abolition of the gasoline tax, coupled with growth in fuel oil and direct crude use for power generation.  Given downward revisions to 1Q08 data, OECD Pacific oil demand is marginally adjusted down to 8.3 mb/d in 2008 (+0.6% on a yearly basis).

According to preliminary data, oil product demand in Japan - which accounts for about two-thirds of OECD Pacific's total - rose by 3.4% year-on-year in April.  Although relatively mild temperatures depressed demand for heating (jet fuel/kerosene shrank by 15.1%), overall demand was supported by a sharp increase in gasoline deliveries (+17.3%) a result of the temporary abolition of the gasoline tax in late March, and by continued demand for fuel oil (+19.8%) and direct crude for power generation (included in 'other products', which rose by 38.6%).  Oil demand is thus expected to reach 5.0 mb/d (+0.4%) in 2008.

However, the structural decline observed in the past several years should resume.  On the one hand, the reinstatement of the gasoline tax suggests that gasoline demand should resume its downward trend in 2008, for the fourth year in a row.  On the other hand, if the nuclear outages are solved in 2009, oil-fired power generation should contract significantly.  In addition, kerosene should also fall as Japanese consumers switch in greater numbers to electricity for heating purposes.

Towards Widespread Demand Destruction?

Over the past month, oil prices have reached record highs, both in nominal and real terms.  In May, spot WTI averaged about $125/bbl - roughly $23/bbl more than the previous inflation-adjusted peak observed in April 1980.  (Admittedly, since there was not a well-established global crude oil spot market in the early 1980s, the choice of both WTI and the US monthly consumer price index from the US Bureau of Labour Statistics as a deflator is somewhat arbitrary, but it is arguably a reasonable and well-known proxy of the average market price of oil.)  In Europe, although the price rise has been tempered by exchange rate appreciation since 2003, prices have also reached record highs.

In the light of previous oil shocks, does the current, steady rise of oil prices augur badly upon global economic growth and overall oil demand?  Although high prices are undoubtedly having an effect, this needs to be qualified.  A starting point is the question of how expensive has oil really become.

  • Global oil expenditures as a share of global GDP (a proxy for the cost of importing oil) remain lower than in the 1980s.  The so-called oil burden stood at around 4.2% in 2007, compared with over 7.3% in 1980.  It may sharply increase to as much as 6.0% in 2008 (assuming that prices remain at current levels for the rest of the year and based upon the IMF's most recent economic forecasts), but it would still be lower than in 1980.  Stripping out of this calculation the GDP of both the FSU and the Middle East - which, given their status of net oil exporters, do not pay international market prices for domestically consumed oil - leads to very similar results (6.8% in 1980, 3.9% in 2007 and 5.6% projected in 2008).

  • The reason why the oil burden has not reached the levels recorded in the early 1980s is due to the fact that real global GDP grew at a faster pace than real oil prices during most of the 1980s and 1990s, with the exception of a few odd years.  As such, income gains more than offset price variations.  Since the late 1990s, however, the opposite has occurred:  real oil prices have risen much more rapidly than global economic activity.
  • Oil intensity (the volume of oil required to produce a unit of GDP) halved from roughly 1.3 b/d per $1M (2007 base) in the early 1970s to less than 0.6 b/d per $1M today as the global economy became more oil-efficient.  This improvement, however, has been more marked in OECD countries; most non-OECD economies, particularly in the Middle East, have a much higher intensity despite their lower GDP per capita, compared with the 'mature' demand patterns observed in the developed world (in purchasing power parity terms).  This is partly due to the fact that many non-OECD countries are at (or reaching) the stage, in terms of GDP per capita, where oil demand typically accelerates, and also to the shift of heavy industry to these countries.

At first glance, therefore, it would appear that the global economy is less vulnerable than in the recent past, even though oil prices have reached historically high levels.  Yet to conclude that the current oil price rally is harmless would be misleading.  Indeed, current oil prices will arguably have damaging and long-lasting economic consequences.

  • In the OECD, high prices are fostering inflation, and more crucially, stoking inflationary expectations just as the world's largest economy - the US - is facing growing concerns regarding the effects of the unfolding housing and credit turmoil.  This complicates the task of central banks and creates serious distortions among advanced economies.  For example, the Federal Reserve has sought to stave off the threat of a sharp economic slowdown by significantly loosening monetary policy over the past few months.  By contrast, the European Central Bank continues to focus on inflation, thus keeping its high interest rates.  The end result has been a dramatic exchange rate movement:  the much weaker dollar is contributing to feed US inflation, while the stronger euro may choke off growth in Europe.  More importantly, although it seems unlikely at this point that sustained high oil prices will trigger a recession across the OECD, they may delay the recovery of the US economy well beyond 2010.
  • In most non-OECD countries, meanwhile, high oil prices are becoming unbearable and are also fuelling inflation.  On the one hand, the cost of imported oil is becoming prohibitive.  Only a few oil-importing countries have to a certain extent been able to cushion the price rise:  those that export other, higher-priced commodities or that have seen their exchange rates appreciate (several central and western African currencies, for example, are indexed to the euro).  But for the vast majority - generally the poorest countries - higher oil prices will in effect have a dramatic effect upon income and development levels.  On the other hand, many countries where end-user prices are capped or subsidised are being obliged to make significant adjustments to their price regimes as these become unaffordable, with all the economic, political and social consequences that this entails.

Will high prices destroy oil demand on a large scale?  Here again, the answer must be qualified.  In those regions where consumers are largely exposed to international oil prices - namely OECD countries and most of the world's poorest nations - oil demand is already stagnating or even falling, according to available data.  In particular, as noted in this report, demand in the US - the traditional driver of oil demand growth in the OECD - is poised to contract markedly in 2008 as a result of the double squeeze provided by the slowing economy and higher prices.  In the short term, discretionary driving is likely to contract further, while the use of public transportation (whenever it is available) should continue to increase.  More significantly, in the medium to long term, as consumers realise that high prices are not due to temporary spikes and are therefore bound to remain at high levels, the structure of the vehicle fleet will arguably change gradually in favour of smaller cars and away from SUVs and light trucks.  Such a trend would be further compounded by more stringent federal mandates on fuel efficiency and by the adoption of diesel-fuelled passenger cars, unthinkable in the recent past but now gaining attention given technological advances.

As such, short-term oil demand growth is now expected to come only from countries where GDP growth is poised to remain strong and where consumers are shielded from the vagaries of the international oil market by virtue of administered price regimes - essentially oil producing countries (in the Middle East, plus Nigeria and Venezuela), China and several large Latin American countries (namely Argentina and Mexico, and to a lesser extent Colombia and Chile).  Buoyant economic performance, combined with end-user subsidies, are indeed feeding a 'demand shock' - as opposed to the supply shock of the 1970s and 1980s - which largely explains why oil prices remain firm.  The question is then whether this group of countries will eventually abandon or adjust capped prices, and whether this will have an effect upon economic growth and ultimately upon oil demand.  Even though it is highly difficult to predict policy changes, this report is based on the following assumptions:

  • Oil-exporting countries will not alter their price regimes in the short term, as they can arguably afford subsidies.  This includes Middle Eastern countries, Nigeria, Venezuela and Mexico.  Moreover, as OPEC countries feature the world's cheapest fuel prices (in Venezuela, for example, gasoline costs ¢5/litre), demand growth will continue to be particularly buoyant there.  The only exception is Iran, which has successfully reduced gasoline consumption through a rationing scheme, and may now extend that policy to gasoil use.
  • Other countries will probably only make token adjustments, given domestic considerations that prevent more fundamental changes.  For many governments, inflation is a key concern; in addition, each country features specific deterring conditions - for example, the Olympics in China or mounting social tensions in Argentina.  However, only China has arguably the financial might to sustain subsidies, and could thus be able delay any price adjustment even if international prices continue to rise.  By contrast, the Indian case is illustrative of the limits of administered regimes:  the government was unable to further postpone a retail price increase - the cost of subsidies was simply becoming too high - and capped prices threatened financial hardship for state-owned companies.
  • A retail price adjustment in China would have to be very large in order to effectively curb domestic, and more generally, Asian oil demand growth.  Despite all the hype generated by the recent adjustments of administered price regimes in the region, only a large price adjustment in China has the potential to significantly alter the demand picture.  As long as economic activity - the primary driver of oil demand growth - remains strong, oil demand will shrug off moderate price hikes and will continue to be buoyant.  Moreover, oil demand growth in China could even accelerate if a retail price hike improved the supply picture:  the continued fuel shortages that have beset the country since 2007 suggest that pent-up demand remains considerable.

Ultimately, while some of the effects of high oil prices upon inflation, consumer spending and growth are plain to see, the interactions are actually very complex.  The IEA will be conducting further in-depth study on these issues in the months ahead and will publish more insights in the forthcoming World Energy Outlook 2008.



Preliminary data indicate that China's apparent demand (refinery output plus net oil product imports, adjusted for fuel oil and direct crude burning and stock changes) rose by an estimated 2.9% year-on-year in April.  This relatively modest rise (compared with +11.9% in March) was mostly related to greater-than-expected weakness in residual fuel oil demand - the feedstock of choice for 'teapot' refineries - which contracted by 21.0%.  Demand for transportation fuels, by contrast, was actually stronger than previously thought, with gasoline soaring by 20.8%, gasoil by 8.2% and jet fuel/kerosene by 16.0%.  As in recent months, demand growth has continued to be met by strong oil product imports (+105 kb/d versus the previous month), as refinery maintenance curbed runs by 2.7% from March.  Chinese demand is now expected to average almost 8.0 mb/d (+5.5% over 2007), slightly higher than estimated in our last report, on the assumption of reconstruction work following May's damaging earthquake in Sichuan.

Despite the array of measures decreed by the government last month, anecdotal evidence suggests that fuel shortages, notably of gasoil, have not only failed to recede but also probably worsened in several provinces.  Yunnan, Guizhou, Guangdong and Fujian have reportedly been the worst hit (Sichuan's earthquake may be to blame in the case of neighbouring Guizhou and Yunnan, but less so regarding Guangdong and Fujian).  Assuming that state-owned companies have no reason to shun the domestic market since they are now subsidised on a monthly basis, there are two plausible explanations for this state of affairs.  One is that the government may be more interested in building stocks of crude and oil products ahead of the Olympics than meeting current pent-up demand.  This would be supported by the strong import figures.  From this perspective, shortages would vanish once the Olympics begin and the world's attention is focused on China, which would be eager to prove it is a modern nation where markets are well supplied.

Alternatively, state-owned PetroChina and Sinopec may be only building minimal stocks, if at all, and are stretching their capabilities as much as possible to meet domestic demand (which they claim to be doing).  As such, logistical bottlenecks and localised supply disruptions would be the main culprits of the current wave of shortages.  Indeed, many independent refineries (which overall represent at least some 12% of total refinery production) have reportedly shut down or are operating at very low capacity, given the mismatch between domestic and international prices.  This would mean that the state-owned giants must not only supply their regular customers but also other sectors such as farmers or fishermen, who normally use off-spec. products from independent refineries.  In this case, oil product stocks could actually be falling, as some reports have suggested.  Consequently, it is not impossible that our apparent demand numbers, even though they are significantly higher than official Chinese demand numbers from the Joint Oil Data Initiative (JODI), may understate actual demand growth.

Other Non-OECD

According to Indian preliminary data, oil product sales - a proxy of demand - soared by 6.5% year-on-year in April.  Transportation fuels - gasoline, jet fuel/kerosene and gasoil sales - have once again been the main contributors to demand growth, rising collectively by roughly 11.5% year-on-year.  It should be noted that the baseline has been sharply revised up following the submission of 2006 data.  Total demand was thus some 170 kb/d higher than expected, growing by +9.5% in 2006 and +5.1% in 2007 (to 2.8 mb/d and 3.0 mb/d, respectively).  In 2008, demand is poised to average almost 3.1 mb/d (+3.9% over 2007, slightly down versus our last report).

This forecast attempts to account for the effects of the retail fuel price rise decreed by the government in early June (+11% for gasoline, +9.5% for gasoil and 17% for LPG).  The Finance Minister had argued in favour of the hike on the basis that, at current international oil prices, state-owned companies could well face a liquidity crisis in only a few months.  Yet the increase was reflective of a political compromise - some government members had pushed for a much higher increase, in the order of 16-22%.  India's government has thus emulated China's (which raised end-user prices on gasoline and gasoil by 9% last November), hoping to a) curb runaway demand, and b) bring fiscal relief to both the country's federal budget and to state-owned companies.  However, there are reasons to doubt that the price hike will prompt significant changes in both areas.

On the basis of India's estimated price elasticity (-0.4), transportation fuel demand should fall by about 4% as a result of the price rise.  Yet the income elasticity is almost twice as high (+0.7); in other words, demand has been driven primarily by income growth, rather than by artificially low prices.  In fact, gasoline is now more expensive in India than in the US, for example, yet Indian income per capita is significantly lower.  As long as India's buoyant economic expansion continues, the price hike will at best slow down the pace of demand growth, rather than trigger a fall in absolute terms.  A significant portion of Indian motorists - those belonging to the country's emerging middle classes, who are supporting the brisk expansion of the automotive market - will arguably be able to absorb the increase.  For poorer drivers, meanwhile, gasoil adulteration - a serious problem in India - will continue to be the backup solution, since the price of kerosene has been left untouched.  It should also be noted that gasoline demand, at roughly 280 kb/d (prior to the hike) is dwarfed by gasoil demand (1.1 mb/d).

Another issue is whether the price hike will help to reduce the gap between domestic retail prices and the cost of imported crude and products, which are heavily taxed.  Although retail prices are now roughly in line with international levels, the cost faced by refiners is much higher.  Although changes in customs duties - abolished for crude oil and reduced from 7.5% to 2.5% in the case of oil products - and a small cut in excise taxes will contribute to reduce the gap, the government's subsidy burden and the financial losses incurred by state-owned companies are not poised to recede to more manageable levels.  According to the government's own estimates, the so-called 'under-recoveries' (losses) of state-owned companies should diminish by about 25% to some $115 million per day (that is, over $40 billion per year) - hardly a dramatic improvement.  The government intends to issue another set of sovereign 'oil bonds' (used to partially compensate the state-owned oil companies), but this will just shuffle the burden around, rather than further curbing it.  The government has bought some time, but the spectre of a liquidity crisis hitting state-owned companies - and widespread shortages - will cast an ominous shadow upon India's downstream sector as long as international oil prices remain high.  In fact, one large company - Indian Oil Company (IOC) - is reportedly considering mortgaging part of its assets in order to avert a cash crunch.

As part of our ongoing assessment of FSU demand, Russian naphtha demand has been revised up following the submission of official data for this product (for the first time ever) and the reappraisal of the 'other products' category.  The new data suggest that some naphtha volumes - primarily used by the petrochemical industry - had been incorrectly reported in the latter category.  Overall demand, however, remains largely unchanged, seen averaging 2.9 mb/d in 2008 (+2.1% over 2007).  Naphtha is now estimated at 238 kb/d, almost three times as high as our previous assessment.  By the same token, 'other products' are revised down by some 150 kb/d to 497 kb/d.

Dealing with High Prices:  A Sequel

Following on last month's discussion on how non-OECD countries are dealing with ever higher oil prices, we provide an update on the most recent developments.  These vary greatly from one country to the next.  In general terms, several Asian countries - with the exception of Thailand - have indeed decided to bite the bullet and increase domestic prices in a seemingly domino-like fashion.  By contrast, most Latin American countries - with the exception of Brazil - are actually reinforcing subsidies.


  • Bangladesh.  State-run Bangladesh Petroleum Corporation proposed a 37-80% hike in fuel prices.  A government decision is expected soon.
  • India.  As discussed extensively above, the government upped retail fuel prices on 4 June (+11% for gasoline, +9.5% for gasoil and 17% for LPG).  The price hike, however, is contentious:  the opposition, and more prominently, the government's nominal Communist allies are intent on fuelling discontent and have called for street protests and strikes (the rural poor, most affected by inflation, form the majority of the electorate).  Inflation is already running at almost 8%, the highest rate in over three years, and could increase by as much as one percentage point following the price increase.
  • Indonesia.  Despite a wave of protests across the archipelago, the government announced a sharp fuel price increase.  From 24 May, gasoline rose by 33%, and gasoil by 28%.  The move was bold, given the country's history of price-related social unrest, but was probably unavoidable as the weight of subsidies ballooned (the government now expects to save some $3.7 billion).  Attempting to cushion the blow, however, the administration will transfer some $1.5 billion in cash to the poorest Indonesians (19 million households).
  • Malaysia.  Despite having announced last month that fuel price caps would be maintained, the government backtracked and implemented a very steep price increase on 4 June.  Gasoline prices rose by 41%, while gasoil prices soared by 63%.  Before the move, the government had imposed a ban on gasoline sales to foreign vehicles within 50 km of the country's border with Thailand and Singapore.  This, however, was unlikely to tame the soaring cost of subsidies (the sales ban was indeed lifted following the price hike).  At almost 5% of GDP, Malaysia's budget deficit was becoming unpalatable; subsidies accounted for around one-third of the budget.  To cushion the blow, the government will hand out cash to motorcyclists and small car owners.  More interestingly, the government appears to have concluded that since any hike would entail a political cost, it was better to implement a significant change.  This echoes developments in Indonesia, but contrasts sharply with the much more timid increases in other Asian countries, notably India.  It should be noted that, before the hike, Indonesia and Malaysia featured the cheapest fuel prices in the region.
  • Sri Lanka.  Retail prices of gasoline and gasoil rose by 31% and 38%, respectively, on 25 May.
  • Taiwan.  The new administration partially removed the price controls that had been imposed in November 2007.  From 1 June, the price of gasoline and diesel rose by 13% and 16%, respectively.  The increase was less steep than previously indicated (around 20%), suggesting that the newly government has been forced to acknowledge widespread public opposition and concerns upon inflationary pressures, already compounded by rising food prices, and upon economic growth prospects.  Moreover, the government will subsidise public transport providers and taxi drivers, as well as fishermen and farmers.  Nevertheless, starting in July, fuel prices will be adjusted monthly under a floating price mechanism.
  • Thailand.  Following the reintroduction of subsidies in mid-March to moderate the impact of high prices on consumers, the government has asked several refiners to absorb the full impact of a further price adjustment decreed in 26 May (the gasoil ex-refinery price was lowered by ¢9/litre), on the grounds that refining margins have been strong.  The targeted refiners, all affiliates of PTT (Thaioil, PTT Aromatics and Refining, IRPC, and Bangchak Petroleum), produce the equivalent of 65% of total gasoil demand.  The lower-priced gasoil will be sold exclusively to private bus operators in Bangkok, fishermen and farmers (collectively some 8% of demand); all other users will continue to pay market prices.

Latin America

  • Brazil.  In early May, state-owned Petrobras, which controls 98% of the country's refining sector, announced an ex-refinery price increase for both gasoline (+10%) and gasoil (+15%) for the first time since 2005, in a bid to stem mounting downstream losses.  Although prices are technically liberalised (the domestic market was deregulated in January 2002), the government has pressured the state-owned company to maintain low prices.  Concerned about inflationary pressures, the government lowered gasoline excise taxes in order to maintain pump prices constant.  In the case of gasoil, though, the tax reduction was marginal, and retail prices rose by about 9%.  It should be noted that Brazil's flex-fuel vehicle fleet has also contributed to keep gasoline prices relatively low- drivers can simply switch to cheaper ethanol if gasoline becomes more expensive.  In fact, ethanol consumption surpassed gasoline demand for the first time ever in April.
  • Chile.  In early June, the government added $1 billion to the country's Fuel Price Stabilisation Fund, which is intended to cushion price fluctuations.  With such a move, the government hopes to engineer a price fall of around 10%.  In order to prompt truckers to end a nationwide strike that seriously disrupted the country's for several days, the government has also agreed to cut the excise tax on diesel by 80% from 1 July 2008 to 30 June 2009.
  • Colombia.  The government postponed plans to eliminate fuel subsidies over 2009-2010 in an effort to keep inflation under control.  However, in order to reduce the fiscal impact, upstream oil companies operating in the country will be required to provide an additional 'contribution' (yet to be defined).  Nonetheless, the government is set to remove subsidies for gasoline and diesel in June 2010 and December 2011, respectively (a year later than originally intended).  As opposed to most other countries, state-owned Ecopetrol does not sell refined products directly to consumers; as such, the government absorbs the losses resulting from sales at below-market prices.
  • Finally, three countries stand as unwilling to remove their administered price regimes:  Argentina, Mexico and Venezuela.  Argentina is struggling with recurrent shortages, as private companies are reluctant to supply the domestic market, despite heavy-handed pressure from the government, which is struggling to cap inflation (officially at around 8%, although most independent observers reckon it is as high as 25%).  Meanwhile, Mexico and Venezuela can afford to maintain subsidies given their status as oil exporters.  The cost, however, is high:  about 2% of GDP in Mexico (around $19 billion, mostly because the country is forced to import large volumes of oil products, notably gasoline) and perhaps as much as 7% of GDP in Venezuela (the country does not publish precise figures on this).



  • Global oil supply rebounded by 490 kb/d in May to average 86.6 mb/d after extensive downward revisions to 1Q non-OPEC production.  May's increase was driven by higher OPEC crude supply and an assumed recovery in both FSU and Chinese output.  OPEC continues to dominate global supply growth, as non-OPEC production has languished at or below levels of a year ago for the past three quarters.  Nonetheless, strong recovery from non-OPEC is expected in the second half of 2008.
  • May non-OPEC production rose 95 kb/d versus April to 49.3 mb/d.  Increases from Russia, the Caspian and China countered lower Canadian and North Sea output.  Downward revisions for other biofuels, Africa and Australia cut 225 kb/d from 1Q08 supply, with 325 kb/d removed for 2008 overall, after a 100 kb/d downward revision for 2007.  Non-OPEC supply averages 49.3 mb/d in 2Q, potentially rising to 51.2 mb/d by end-year as seasonal outages end, and assuming new projects start-up on schedule.
  • Non-OPEC supply growth in 2008 now averages 455 kb/d (close to 2006/2007 levels), and is concentrated towards end-year.  The FSU, Brazil, Australasia, biofuels, the US GOM and Canadian oil sands are key contributors, while Mexico, the North Sea, other onshore North America and the Middle East see continued decline. An incremental 325 kb/d is expected from OPEC NGL in 2008, to 5.1 mb/d.
  • OPEC May crude supply averaged 32.3 mb/d, a gain of 395 kb/d versus April.  Increases were widespread, with Saudi Arabia, Iraq, Nigeria and Angola prominent.  Pipeline outages and marketing constraints continued to curb Nigerian and Iranian supplies respectively.  Iraqi output hit a six and a half year high at 2.5 mb/d, as Basrah exports increased further. Saudi Arabia signalled its intent to boost supplies in response to higher refiner demand.
  • OPEC effective spare capacity dipped below 2 mb/d for the first time since 3Q06. Delays in commissioning new upstream capacity, plus higher physical output, were responsible for the tighter level of upstream flexibility. However, OPEC producers should add a net 600 kb/d to capacity by end-2008, reaching 35.6 mb/d on the strength of additions from Saudi Arabia, Nigeria, Angola and Algeria. Further increase could come from Iraq depending on contract and security issues.
  • The call on OPEC crude and stock change is raised by 0.2-0.3 mb/d for 2006 and 2007 on higher assessed non-OECD demand. The reduction applied to 2008 demand lags that applied to non-OPEC and OPEC NGL output, leading to a 0.3 mb/d upward revision to the 2008 'call'.  Averaging 31.9 mb/d for 2Q08, the 'call' then drops to 31.4 mb/d, as non-OPEC supply growth rebounds.  But further non-OPEC capacity slippage would correspondingly raise the 'call' in the second half of the year.

Note:  Random events present downside risk to the non-OPEC production forecast contained in this report.  These events can include accidents, unplanned or unannounced maintenance, technical problems, labour strikes, political unrest, guerrilla activity, wars and weather-related supply losses.  Specific allowance has been made in the forecast for scheduled maintenance in all regions and for typical seasonal supply outages (including hurricane-related stoppages) in North America.  In addition, from July 2007, a nationally allocated (but not field-specific) reliability adjustment has also been applied for the non-OPEC forecast to reflect a historical tendency for unexpected events to reduce actual supply compared with the initial forecast.  This totals ?410 kb/d for non-OPEC as a whole, with downward adjustments focused in the OECD.


OPEC May crude supply averaged 32.3 mb/d, a gain of 395 kb/d versus April.  April's output estimate was revised up marginally, largely due to a higher assessment for April Iraqi exports.  Increases in May were widespread, with Saudi Arabia, Iraq, Nigeria and Angola prominent amongst those seeing higher supply.  However, pipeline outages and limited markets for heavy, sour crude continued to constrain Nigerian and Iranian supply respectively.  Iraqi supply hit a six and a half year high at 2.5 mb/d, as Basrah exports increased and northern exports via Turkey were sustained. Saudi Arabia reportedly began raising output around 10 May, signalling its intention to boost supplies by as much as 300 kb/d by June in the face of higher refiner demand. But delays in commissioning new upstream capacity (not least in Saudi Arabia), allied to higher physical output, pushed our estimate of effective OPEC spare capacity below 2 mb/d for the first time since 3Q06.

June OPEC supplies may increase further, subject to both higher expectations for Saudi output and Iraqi exports being realised.  Moreover, June is expected to see the partial release of substantial crude volumes held in floating storage off Iran.  Conflicting reports in May suggested that between 15 and 30 million barrels has been stored aboard VLCCs anchored near Kharg Island.  In contrast, loading schedules suggest that Nigerian supply could remain constrained in June.

Saudi Arabian supply is estimated to have increased by some 150 kb/d in May to average 9.2 mb/d.  Saudi Oil Minister Ali al-Naimi announced in mid-May that the Kingdom would boost output by 300 kb/d to around 9.45 mb/d by June to meet higher demand.  Reportedly, incremental output will come from the Ghawar and Safaniyah fields, producing Arab Light and Arab Heavy crude respectively.  Despite seasonally rising eastbound volumes, Asian buyers have reportedly been showing some resistance to what they see as over-priced lighter Saudi barrels, so it will be interesting to see how formula prices evolve in coming months.

OPEC Capacity Growth Revisited

Current installed OPEC crude production capacity is assessed at just shy of 35 mb/d.  This is based on oil that could theoretically be produced at the wellhead within 30 days and sustained at that level for 90 days.  No account is taken of constraints of quality, transportation (except for Iraq and Nigeria) or pricing which, from time to time, can render marketable spare capacity lower than these levels.  However, prevailing physical limitations on gas flaring and other technical issues are accounted for, with the result that our numbers aim to represent sustainable output potential, rather than short-term surge capacity.

Capacity estimates are updated periodically, taking account of latest reliable estimates of field capability after allowing for assumed decline rates.  Just as for non-OPEC estimates, new field start-ups are factored in based on latest information on start-up date and the likely rate of ramp-up towards capacity.  The vagaries of OPEC output policy play no role in shaping capacity estimates unless they are deemed to be acting to defer investment in new capacity.  The uncertainties over these assumptions, and the secrecy surrounding field operational details among many producers, mean that our capacity estimates may lack precision.  But they hopefully provide a valuable snapshot of prevailing upstream supply flexibility.

As we note above, OPEC effective spare capacity (excluding Iraq, Indonesia, Nigeria and Venezuela which face ongoing security, operational or investment issues) has now slipped below 2 mb/d.  This is one element among many driving high crude prices, as market players digest the implications of such minimal room for manoeuvre in the face of the geopolitical risks hanging over several major producers.

We see OPEC crude capacity increasing by the end of 2008 by some 600 kb/d on a net basis (after netting off capacity lost to decline), to reach 35.6 mb/d from a current 35.0 mb/d.  Saudi Arabia and Nigeria both add 200-300 kb/d on a net basis as the delayed AFK (Khursaniyah) project and Agbami field enter service.  This assumes no further commissioning delays for gas processing equipment for Khursaniyah and the avoidance of a recently mooted strike affecting the Agbami project.  Angola gains a net 170 kb/d to reach 2.05 mb/d crude capacity.  New supplies materialise from the summer onwards from the Saxi and Batuque (Kizomba C), and Gimboa fields, while further increase comes from the already-producing Rosa field.  Enhanced oil recovery work could boost Algerian capacity by a net 50 kb/d.  Prospects also look greater than for some considerable time for Iraq to boost production capacity, potentially by up to 500 kb/d, although in this assessment we have held Iraqi capacity flat through 2008 at 2.5 mb/d.

Higher capacity is always welcome, but may in reality provide limited comfort for a jittery oil market.  Spare capacity of 2.5-3.0 mb/d is better than 2 mb/d, but still represents less than 4% of global demand. Our last review in late 2007 foresaw end-2008 capacity reaching a higher 35.8 mb/d.  OPEC NOCs face the same bottlenecks in raw materials, labour and service capacity, and cost inflation as their IOC counterparts.  Project slippage could derail even these now more modest expansions, and keep markets tight.  Finally, does the impetus for OPEC producers to install new capacity remain as strong as when plans were formulated three to five years ago?  With increased revenue needs and a rising tide of resource conservation sentiment, some producers may be questioning the rationale for tampering with the tight spare capacity-high price mechanism at all.

Contrary to earlier reports, it now appears that the Abu Hadriya-Fadhili-Khursaniyah project (also known as AFK, or Khursaniyah) did not in fact begin producing in April, and will remain idle until the second half of 2008.  Originally due to enter service at end-2007, Khursaniyah has been delayed by problems with gas-processing facilities.  We have removed previously assumed initial volumes from Saudi capacity as a result, taking our current assessment of capacity back down to 10.65 mb/d (excluding Bahrain's share of Abu Safah crude and also Saudi condensates).  As a working assumption, we have pushed Khursaniyah start-up back to 3Q08, building to capacity 500 kb/d output by 2Q09 (although Aramco suggested prior to the latest gas-processing issues that that ramp-up could be quite rapid).

Iraqi crude supply (exports plus use in refineries and power plants) is assessed up by 100 kb/d in May to 2.5 mb/d.  This, together with May exports of 2.02 mb/d, was a post-war record.  April supply was revised up by 60 kb/d from 2.34 mb/d to 2.40 mb/d on higher Kirkuk exports from Ceyhan and domestic crude use (which averaged 440 kb/d and 480 kb/d respectively).  While both Ceyhan exports and local use are estimated at similar levels for May, southern exports from Basrah and Khor al-Amaya came in some 100 kb/d higher than in April, at 1.57 mb/d.

Iraq's Oil Minister said that, subject to a continued disruption-free flow of crude by pipeline to Ceyhan, national exports could attain another post-war record in June, at 2.2 mb/d.  Kirkuk exports themselves are targeted at 600 kb/d in June.  With plans in hand to repair the strategic north-south pipeline and products distribution network, the scene may be set for Iraq to be able to boost supply capacity.  This report has long assessed that hitherto an inability to boost supply beyond 2.0 mb/d was due mainly to power supply, export and refining capacity limitations, ahead of reservoir or wellhead issues.  Regular output close to headline 2.5 mb/d capacity now looks a more realistic prospect, and indeed our base case assumption is that this capacity level is sustained through 2008.  However, if the security situation stabilises and pipeline links are reinstated, the prospect of adding up to 500 kb/d to crude capacity in fairly short order, through recently discussed technical service agreements, begins to look more viable.

Nigerian May supply is assessed up by 55 kb/d from April at 1.9 mb/d.  Renewed attacks affecting Shell's Bonny pipeline network kept an earlier force majeure on exports in force. However, this was believed offset by higher supplies of Forcados crude and from ExxonMobil's Qua Iboe, Erha and Yoho fields after last month's strike action was settled.  Nonetheless, export loading schedules suggest Nigerian supply will likely remain constrained in June, before potentially rising in excess of 2.0 mb/d again in July (security and industrial action permitting).  In this context, a threat by oil union Pengassan to halt Chevron's Nigerian operations, amounting to at least 300 kb/d, in a dispute over welfare and recruitment issues, will be watched carefully.  Nearly 600 kb/d of Nigerian capacity is now shut-in due to Niger Delta unrest.

Non-OPEC Overview

May non-OPEC oil output is assessed at 49.3 mb/d, an increase of 95 kb/d from downward-adjusted April levels.  However, the non-OPEC baseline is revised downwards quite extensively, with 2007 and 2008 output now estimated at 49.6 mb/d and 50.0 mb/d respectively.  Annual totals have been revised down by 100 kb/d and 325 kb/d compared to last month's report.  Downward adjustments to conventional oil production this month are focused in Australia, Mexico, non-OECD Asia, Brazil and the Ivory Coast.

However, biofuel supply outside of Brazil and the US accounts for 70 kb/d (70%) of the 2007 supply revision and 160 kb/d (50%) of 2008's adjustment.  An annual overview of European and Asian biofuel capacity developments in particular, conducted ahead of the forthcoming Medium-Term Oil Market Report, reveals a more uncertain picture concerning crop economics, government targets and mandates, with consequent new capacity slippage.  Nonetheless, aggregate world biofuels growth in 2008 still attains +330 kb/d, taking total global biofuels supply close to 1.4 mb/d for the year, underpinned by strong US and Brazilian ethanol growth (these elements are included in total US and Brazilian oil output estimates, unlike 'other' biofuels, which are listed as an aggregated line item in our global oil supply/demand balance - Table 1).  World biofuels production growth in 2007 is thought to have amounted to some 210 kb/d.

All told, non-OPEC annual supply growth now looks likely to average close to 0.5 mb/d in 2008, akin to 2006 and 2007 levels. The buoyant growth evident early in 2007, allied to outages, mature field decline and project slippage seen this year, pushes the first half of 2008 into contraction.  However, with the bulk of this year's expected 4.1 mb/d of new non-OPEC capacity start-ups likely to enter service beyond mid-year, the late 2008 and early 2009 supply trend is expected to be strong as production from these new projects builds towards design capacity.  Canada, Russia and Brazil are seen likely to push production higher in 3Q compared to early in the year under the impact of new project start-ups and planned outages which are orientated more to the first half of the year than has been evident in previous years.  Historically, the fourth quarter tends to see a production rebound as autumn outages recede, and with the predominance of 2008 new start-ups later in the year, that trend is exaggerated.  Geographically, the US GOM, Canadian oil sands, Asian and Australasian crude and condensates, Azerbaijan and Brazil are the main drivers of growth in 2008.  Conversely, Mexico, other North America, North Sea and non-OPEC Middle East supplies are expected to fall most sharply.

OPEC gas liquids supply (over and above non-OPEC production) is seen reaching 5.1 mb/d this year, from 4.8 mb/d in 2007.  Qatar, Saudi Arabia, Kuwait and the UAE are largely responsible for this year's increment in NGL and condensate, although here too project slippage has cut some 65 kb/d from our earlier assessment for likely 2008 production.  Accelerating growth is also likely to be seen from this source of supply in 2009.


North America

US - May Alaska actual, others estimated:  March data for US crude production came in some 40 kb/d higher than expected at 5.14 mb/d.  NGL and oxygenate/refinery additive supplies were also a combined 100 kb/d higher than expected at 2.6 mb/d.  While slightly higher GOM and US lower 48 production levels are carried through the forecast, the NGL and oxygenates levels for 2008 are held largely unchanged.  Data for Alaska in May highlighted another disappointing month, with crude supply of some 700 kb/d down by 10 kb/d from April and around 75 kb/d below May 2007.  An early-month power outage affecting Prudhoe Bay lay behind the weaker-than-expected output.

Total US oil production in 2008 is now forecast at 7.55 mb/d, a rise of 80 kb/d versus 2007 but only around half of 2007's observed growth.  Ethanol supply grows by 120 kb/d to 540 kb/d, with NGL expected to gain 40 kb/d to 1.82 mb/d.  Federal offshore GOM supply gains 65 kb/d to attain 1.4 mb/d, with contributions from the Atlantis, Genghis Khan, Neptune, Typhoon, Mirage and Thunder Horse projects. Collectively however, crude production elsewhere in the US could fall by as much as 150 kb/d, offsetting in part the above increases, even though there is anecdotal evidence that higher prices may be helping to stem decline in some mature producing areas.

Canada - Newfoundland April, others March:  Total Canadian oil production in 2008 is expected to grow by a modest 30 kb/d, to average 3.35 mb/d.  That follows two years of growth in excess of 100 kb/d in 2006 and 2007.  Conventional crude supply averages 1.94 mb/d, with the balance shared equally between synthetic crude from mined bitumen and NGLs. In-situ bitumen and heavy oil production (which we categorise as the 'conventional' slice of Albertan oilsands output) is the main source of growth this year, adding 70 kb/d to 2007's level and reaching 615 kb/d.  Growth in 2009 may shift more towards mined synthetic crude once again, notably with an anticipated full year's output from the CNRL Horizon project.  The Horizon project is now expected online in 3Q08, likely attaining 85% of its 110 kb/d capacity by end-year and capacity operating levels early in 2009.

Mexico - April actual:  Pemex officials have acknowledged that their earlier prognoses for attaining 3.0 mb/d of crude production in 2008 are unlikely to be realised.  Revised estimates now stand at 2.9 mb/d for the year, which we still see as somewhat optimistic.  Our own 2008 forecast now stands at 2.86 mb/d, after April production came in a weighty 160 kb/d below previous expectations, at 2.77 mb/d.  Reports suggest that some of the heavy decline in baseload Cantarell production in early 2008 has been due to a damaged pipeline, holding out the prospect for some recovery in months to come.  Nonetheless, we are assuming a weighty 25% decline from Cantarell this year in our forecast in the absence of any unexpected surge in spending to curb such inflated recent decline levels.  Rising deepwater Ku-Maloob-Zaap production only partially counteracts Cantarell's fall, and Mexican oil output slips by over 200 kb/d for the second year running in our forecast.

April crude exports fell by nearly 200 kb/d to 1.44 mb/d, although the impact of this was likely cushioned by North American refinery outages.  Late May also saw renewed port closures due to bad weather, holding out the prospect that exports may have again come in below potential.  NGL April output of 370 kb/d was also marginally below our forecast.

North Sea

Norway - March actual, April provisional:  While March data show liquids production (crude oil, NGL and condensate) running some 70 kb/d below our forecast, preliminary April data stand at around 30 kb/d above our earlier estimate.  Crude output under our estimation (whereby some light production categorised as crude in national statistics is reallocated to condensate in our own) came in at around 1.94 mb/d and 1.9 mb/d in March and April respectively.  A combination of an oil leak at the Statfjord field and seasonal maintenance outages potentially pushed crude down to 1.78 mb/d in May, and a maintenance-induced low point for production is expected to be reached in June, when we forecast crude output of 1.54 mb/d.  All told, we see Norwegian production averaging 1.81 mb/d of crude and 565 kb/d of gas liquids in 2008, for a total of 2.37 mb/d, largely unchanged from last month's forecast.  A trend of near-0.2 mb/d decline therefore continues for the fourth year in succession. Indeed the government has just issued a downward-revised oil production forecast with its most recent budget, at 2.4 mb/d compared to an earlier 2.5 mb/d.

UK - March actual:  While still on a declining trend, UK oil production estimates are revised modestly higher for 2008 this month, to the tune of some 20 kb/d.  April loading schedules point to lower spring maintenance than we had been assuming.  Nonetheless, we do expect UK production to dip to 1.4 mb/d in 2Q08 and towards 1.3 mb/d in the third quarter under the impact of maintenance-related outages, before regaining levels closer to 1.5 mb/d late in the year.  UK offshore crude output is now seen averaging 1.24 mb/d in 2008, some 195 kb/d lower than in 2007.  Natural gas liquids (217 kb/d) and onshore crude oil (24 kb/d) take total expected UK output this year to 1.48 mb/d.

UK production has declined by 150-200 kb/d in six of the past eight years, with 2007 proving an exception as ramp-up from Nexen's 200 kb/d Buzzard field offset mature field decline.  This year sees a number of smaller scale start-ups, including Jura (which started condensate output in May), Starling, Brodgar and Callanish, Perth, Chestnut and Ettrick, but collectively these are insufficient to maintain overall production levels.  There are concerns that a growing role for smaller operators in the North Sea, allied to a tighter credit market, may impede marginal and satellite field developments.  The UK government has responded by saying it plans to exempt marginal assets in existing oil blocks from Petroleum Revenue Tax.  Industry representatives are seeking wider-ranging fiscal reform to help sustain investment against a backdrop of rising costs.  UKCS operating costs were reported to have increased from $2/boe in 2006 to $12/boe in 2007, with a prospect that they could remain at inflated levels through the end of the decade.


Australia - March actual:  Extended and unplanned field outages kept March Australian oil production below 500 kb/d for the third month running, with prevailing crude output in 1Q08 of around 430 kb/d and 65 kb/d of NGL. This is around 50 kb/d below both 1Q and 4Q07 levels.  April and May are likely to have seen some recovery in production, to levels closer to 600 kb/d, as among others, the Mutineer-Exeter field in the offshore Carnarvon Basin returned to more normal service.  That said, we have nonetheless curbed our forecast for Australian 2008 production by around 50 kb/d, with lower prevailing output levels of Carnarvon Basin crude and condensate and NGL supply generally underpinning the revision.  However, we do still see Australian oil production enjoying something of a temporary surge (which could last into 2009/2010), rising by 60 kb/d in 2008 versus 2007, to average 610 kb/d.  New field start-ups including Stybarrow (already online), Vincent, Angel and Skua underpin the increase.

Former Soviet Union (FSU)

Russia - April actual, May provisional:  May Russian oil production of 9.95 mb/d (including 9.47 mb/d of crude oil) showed year-on-year decline for the fifth straight month, albeit rising modestly compared to April.  We retain a forecast showing Russian production largely flat in 2008 at 10.1 mb/d, with Rosneft and to a much lesser extent Lukoil showing growth, but offset by a weakening trend amongst most other producers.  Output from Production Sharing Agreements (a key source of 2006/2007 growth) is largely static, as higher and potentially year-round production from Sakhalin 2 counteracts an expected decline from Sakhalin 1. Other 2008 increments come from the Vankor, Salym and South Khylchuskoye projects.

May saw further official pronouncements suggesting extra tax breaks could be forthcoming to reinvigorate production growth.  One company source suggested that a raise in the mineral extraction tax threshold to some $45/bbl would be needed to reverse recent production declines.  Current plans will see the threshold lifted from $9/bbl to $15/bbl in early 2009, and the President has suggested 2010 will see further tax reductions. The Deputy Industry and Energy Minister said in early June that older oilfields might even be made exempt from mineral extraction tax altogether.

Azerbaijan - March actual; Kazakhstan - April actual:   In the face of stagnating Russian production, the two key Caspian producers pick up the FSU growth baton in 2008. A combined increase of 260 kb/d this year takes Azerbaijan's production to 1.06 mb/d, and that of Kazakhstan to 1.45 mb/d.  Azerbaijan sees the lion's share of the increase as output from the Azeri-Chirag-Guneshli complex of fields builds towards 1.0 mb/d by end-year.  Operator BP is reportedly trying to persuade the Azeri government to extend the ACG Production Sharing Agreement which is due to expire in 2024.  The company has said that enhanced recovery could boost proven reserves at ACG to nine billion bbls from a current five, potentially extending plateau production of 1.0 mb/d through until 2019 from a currently envisaged estimate of 2013.

In Kazakhstan, efforts centre on diversifying export routes, with uncertainty still surrounding expansion of the CPC pipeline. CPC expansion was previously believed indispensable to accommodate higher  volumes from the 500 kb/d Tengiz project and new oil from the giant Kashagan field.  Now Tengiz operator Chevron has signed a deal with the Kazakh government to join a project to build a pipeline from Iskene in western Kazakhstan to Kuryk on the Caspian coast.  This would eventually allow up to 1.1 mb/d of Kazakh crude from Tengiz and Kashagan to access the Baku-Tbilisi-Ceyhan pipeline, potentially obviating total reliance on exports via CPC and Russia.  Bottlenecks on CPC and railroad exports are seen potentially limiting increases in Kazakhstan's production in 2008.

Preliminary April data show net FSU oil exports of 8.7 mb/d.  This represents a decrease of 110 kb/d compared to March and is 390 kb/d below April 2007.  A significant monthly drop of 180 kb/d in fuel oil exports led down total FSU product exports by 290 kb/d.  This overshadowed an average increase of 170 kb/d in crude exports.  Although shipments from the Caspian nations were down, a 180 kb/d rebound in volumes via Russia's northerly Primorsk drove a total rise of 290 kb/d through the Transneft pipeline network.  The rise in crude exports has been attributed to refinery maintenance in Russia, which also reduced product exports.  Lower Azeri production in April prompted a 70 kb/d dip in BTC exports.

FSU net oil exports likely rose in May.  Loading schedules suggest that FSU crude exports via Transneft increased alongside higher Caspian volumes and a rebound in product cargoes.  This should precede lower FSU exports in June coinciding with a rise in Russian crude export duty of 16%.

Other Non-OPEC

Malaysia - March actual:  March crude and condensate production from Malaysia of just over 700 kb/d stood 60 kb/d below 2004 peak levels.  However, production is expected to increase by some 6% (40 kb/d kb/d) in 2008 to average 735 kb/d.  Around 45 kb/d of NGL and 15 kb/d of gas-to-liquids output from Shell's Bintulu facility make up total expected 2008 oil production of nearly 800 kb/d.  The increase this year comes from Murphy's deepwater Kikeh field, which began output in August 2007, and which should attain plateau 120 kb/d production in 2009.  State company Petronas suggests that deepwater areas could contain some 8 billion bbls of reserves, with a second deepwater project, Gumusut-Kakap, due to produce up to 150 kb/d by early next decade.

China - April actual:  Chinese production in April of 3.75 mb/d came in below expectation.  Nonetheless, 2008 is expected to see a continuation of the steady 50-100 kb/d annual growth evident for most of the present decade.  As such China has been a consistent contributor to non-OPEC growth, albeit supply growth is dwarfed by annual demand increases of the order of 300 kb/d in recent years. Production growth this year centres on offshore increments at the Penglai, Lufeng and Bozhong fields, but also at the northern onshore Changqing field.  PetroChina aims for the field to overtake Shengli as the country's second largest oil and gas producer within two years.  Oil output at Changqing has risen by 135% since 2001, averaging 243 kb/d last year.  Our forecast has Changqing rising further to 266 kb/d in 2008.

OECD stocks


  • OECD oil stocks fell counter-seasonally in April, by 8.1 mb, ending at 2,562 mb.  The five-year average April build is around 30 mb.  Decreases of 4.8 mb and 11 mb in OECD North America and Europe were partially offset by a rebound of 7.7 mb in the OECD Pacific from extreme lows.  Upward March revisions, centred in European products, lessen the first-quarter OECD draw to 0.1 mb/d.
  • A 4.4 mb April draw took European crude inventories below their five-year range.  North American crude stocks built by a seasonally meagre 1.3 mb, dented by 6.1 mb drop in Mexico.  However, a 7.6 mb build in Japan helped OECD Pacific crude stock levels to recover from recent historical lows.
  • A steep 11 mb draw reduced US motor gasoline stocks to average levels by end-April, removing the massive overhang which emerged in the first quarter.  Declining distillate stocks in Europe (down 7.1 mb on strong demand and refinery outages) and North America (off by 2.6 mb) reduced cover to the bottom of the five-year range in both regions.  A temporary exemption in a Japanese road tax prompted a demand-led gasoline stock draw of 2.4 mb in April.
  • OECD stock cover remains above average, at 53.4 days of forward demand, despite the April draws.  Preliminary US and Japanese data indicate a 7.3 mb stockbuild in May, far less than usual, with US crude stocks potentially falling by a substantial 17.2 mb.  However, a lower OECD demand outlook means that this unseasonably slight stockbuild would only reduce cover by 0.1 days in May.

OECD Inventory Position at End-April and Revisions to Preliminary Data

The 8.1 mb April draw in OECD stocks marked a counter-seasonal start to the second quarter, a time when stocks normally build.  A sharp decrease of 11 mb in US motor gasoline stocks dominated the stock draw and served to remove the massive 1Q overhang that had been built ahead of heavy seasonal maintenance.  The draw leaves US gasoline stocks near the five-year average.  Perhaps more significant though were some notable counter-seasonal draws in crude and distillate stocks in April, contributing to the sharp rise in prices at that time.

European crude inventories fell below the five-year range in April.  At the same time, a drop in Mexican crude stocks (probably related to production problems) coincided with US crude inventories falling further behind their five-year average.  Preliminary data for the US suggest an unseasonal crude draw of 17.2 mb in May to well below the five-year average.  The latter may be a sign of destocking for economic reasons.  If it wasn't for high crude inventories being reported by the Canadians (possibly a structural trend associated with the Canadian upstream), OECD North American crude stocks would be regarded as extremely tight by historical standards.  OECD Pacific crude stocks recovered in April, with a sharp 7.6 mb build in Japan marking a recovery from the below-average levels seen for over a year.

Distillate stocks, in terms of forward demand cover, are now at the bottom of the five-year range in North America and Europe.  April draws in France and the UK, amid strong demand and refinery outages, prompted a regional decrease of 7.1 mb and reduced forward demand cover to 32.4 days or 1.7 days below average.  US distillate stocks lag the seasonal average and, relative to rising demand, represent an historically thin level of forward cover of 25.5 days.  This is 0.9 days below the 2003-07 end-April average.

Despite these specific areas of tightness, aggregate OECD oil stock cover remains just above the five-year average at 53.4 days (of forward demand) at end-April.  Preliminary US and Japanese data indicate only a 7.3 mb build in May, a month when OECD stocks typically build by 50 mb.  Yet, even incorporating this, forward demand cover is only set to fall by 0.1 days.  This can be partly explained by a higher March base as well as recent downward revisions to OECD demand.  However, it is also clear that the headline OECD total oil stock and stock cover figures do not tell the whole story, with tightness in certain key product categories offset by higher cover in others (see 'Dissecting the Surplus').

Upward revisions to March OECD stocks totalled 9.5 mb.  European product inventories were reassessed at 10 mb higher than reported last month, including a correction of +3.4 mb in middle distillates (with France and Italy corrected around 2 mb higher each).  OECD North American fuel oil stocks were the other significant revision, adjusted down by 3.1 mb.  Despite a 6.1 mb downward revision to February crude stocks, March revisions have left the first-quarter OECD stock draw even smaller than reported last month, at 0.1 mb/d, and further above the five-year average 1Q draw of 0.4 mb/d.

OECD Industry Stock Changes in April 2008

OECD North America

OECD North American industry stocks fell by 4.8 mb in April, a month when they usually build by around 20 mb.  Regional crude stocks only rose by 1.3 mb which was well short of the 10 mb-plus rise normally observed in April.  Production problems in Mexico, centred on the Cantarell field, were apparently behind a 6.1 mb decline in crude inventories.  This almost entirely offset a 7.4 mb increase in April US crude inventories which was also short of seasonal norms.  The US build might have been larger, given the substantial lag in refinery crude throughputs to the five-year range, were it not for below-average crude imports in early April (notably from Mexico).  Preliminary weekly data suggest that US stocks are set to decrease by over 17 mb in May, a significant counter-seasonal trend which could indicate deliberate destocking as a means to release capital in the current environment of high oil prices.  Canada continues to report very high crude stock levels, most recently for March.  Accelerating a very long-term rising trend, Canadian crude inventories have risen by almost 15 mb, to nearly 120 mb, since last summer.  This is not a seasonal trend and may relate to increase operational stocks needed in association with growing synthetic crude production.

Regional product inventories also bucked seasonal trends in April, posting a 4.1 mb decrease on aggregate.  A steep draw of 11 mb in US motor gasoline stocks, a result of heavy spring refinery maintenance and unplanned outages, cancelled out the seasonal rebuild of propane stocks (included in the 10.4 mb increase in US 'other products').  Moderate middle distillate draws of just over 1 mb were also observed in US and Mexico respectively. US stocks of gas liquids decreased by 2.1 mb.

OECD Europe

OECD European crude and product stocks also fell counter-seasonally in April.  The headline stock draw was seen in middle distillate inventories, which fell by 7.1 mb.  Respective distillate declines of 2.4 mb and 2.3 mb in France and the UK were the result of disrupted refinery production (due to strike action at Fos-sur-mer and Grangemouth plants) amid apparently firm demand.  New York Harbour diesel prices were level with those in Rotterdam in early April before rising to a $10/bbl premium in the second half of the month, discouraging transatlantic diesel trade flows to Europe.  Strong gasoil demand was also probably instrumental in April drops in the Netherlands', German and Italian distillate inventories of 1.5 mb, 0.8 mb and 0.8 mb respectively.  Independently held gasoil stocks in the ARA region registered a decline in mid-May to their lowest point since 2Q2004 before rebounding in early June.  In other products, gasoline stocks rose by 1.6 mb, including a 1.6 mb increase in France.

The European April crude draw totalled 4.4 mb, dragged down by a 7.4 mb decrease in the Netherlands.  French crude stocks also fell by 2.4 mb in April in line with a rebound in refinery throughputs of 0.2 mb/d.  By contrast, German, Italian and UK crude stocks each built by between 1.5 mb and 3 mb, with the German increase attributable to a drop of 0.2 mb/d in refinery crude runs.  The only countries currently with crude stocks above the five-year average are Ireland, the Netherlands, Poland, Slovakia and Switzerland.  Of these, the largest surplus is in Poland, where stocks structurally rose in 2007 in accordance with the scheduled building of stocks to meet IEA/EU emergency requirements.

OECD Pacific

A 7.6 mb hike in Japanese crude stocks led a regional total oil stockbuild of 7.7 mb in April.  The increase in Japanese crude inventories was prompted by a reduction in crude throughputs in the run-up to peak maintenance but also helped by higher imports (up 0.15 mb/d to 4.55 mb/d), notably from Saudi Arabia and United Arab Emirates.  If confirmed, the hike brought Japanese crude inventories back in line with five-year seasonal averages, after 13 months below average.  Korean crude imports, by contrast, decreased by almost 0.1 mb/d (to 2.25 mb/d) according to reports.  This contributed to a Korean crude stock draw of 1.1 mb reflected by preliminary April data, despite refinery throughputs sinking by 130 kb/d to 2.3 mb/d.  The changes in Japan and Korea led a regional April rise of 6.6 mb, bringing crude oil inventories back within the five-year range after the March low.  This trough was lowered further by a 1.6 mb monthly drop reported in Australian crude stocks (to well below seasonal norms).

OECD Pacific product stocks rose by 0.5 mb in April.  A 2.4 mb draw in Japanese motor gasoline stocks, was related to a demand spike caused by a temporary suspension of a special road fuel tax (with widespread reports of long queues at service stations).  However, rising distillate and residual fuel oil stocks, especially in Korea where product exports were reduced in April, pushed Pacific product stock changes into positive territory.  With Korean heating degree days significantly lower than the 10-year average in March and April, lower heating demand may have contributed to increases in these fuel categories.

Recent Developments in Singapore Stocks

Independent middle and light distillates stocks in Singapore rose above five-year ranges in May.  Middle distillate inventories reached 9.8 mb in early June, having increased by 1.6 mb in May.  Smaller rises in light distillates and residues pushed total products stocks up by 2.3 mb in total.

Dissecting the Surplus

Comparing OECD total oil stocks and stock cover to the most recent five-year average allows an at-a-glance measure of the OECD stock buffer against a consistent yardstick.  However, this comparison is not without problems, most notably a glossing over of regional- and product-specific stock trends.

With the exception of November 2007, OECD total oil stocks have trended above the five-year average since October 2004.  OECD total stock cover has exceeded the five-year average since September 2005.  If we dissect the absolute surplus in total oil stocks in April to the five-year average, we can see that the only product categories in surplus are North American crude stocks and European product stocks (see chart).  Within the former category only Canadian crude stocks are now in surplus.  For European products, it appears that the excess gasoline, residual fuel oil and 'other products' produced by refiners in their attempt to meet middle distillate demand, are now making up the surplus in European product stocks.  This reflects the problem currently facing European refiners, namely that product yields are out of line with product demand growth patterns, causing a structural long in gasoline and fuel oil and a shortage of distillates in the region.  There were also increases in Polish crude and Turkish product stocks in 2007 related to meeting IEA/EU emergency requirements.  These structural stockbuilds inflated European totals for these products, implying that stocks would be even tighter without them.  Dissecting the OECD stock surplus in this way suggests that oil stocks are indeed tight in key areas, namely distillates and crude, but this fact is lost in the country/product aggregation and its comparison with the five-year average.

A look at forward demand cover for individual products displays a similar picture.  Very high levels of forward demand cover in North American and European fuel oil stocks, alongside European gasoline, are potentially inflating regional total cover in April.  Again, the OECD total oil cover disguises the tightness in crude and distillate cover which has pushed prices to such high levels in recent months.



  • Crude futures spiked at just under $140/bbl on 6 June on Israeli statements regarding Iran.  Prices had previously slid from a 22 May peak of $135 to around $122/bbl, but overall remain high and volatile, driven by strong fundamentals, especially for distillates, and tight effective spare capacity.
  • Refined product markets remain out of kilter, with gasoline crack spreads to crude untypically weak and fuel oil still suffering from a global glut.  Distillates in contrast remain strong on healthy global demand and structural shortness in Europe and parts of Asia.
  • Refining margins mostly fell in May given the above-mentioned product imbalance.  Gasoline-driven US cracking margins were mostly negative in May, unusual at the start of the summer driving season.  In contrast, European Brent and Urals cracking margins remain healthy due to diesel's strength.
  • Dirty tanker rates in the Middle East Gulf and West Africa were pushed to extreme highs in May, with Iranian floating storage a major constraint on VLCC supply, while firm long-haul demand was boosted by increased Saudi exports.  Clean tanker rates in East of Suez markets rose in May and remained firm in the Atlantic.


On 6 June, crude prices surged to record heights near $140/bbl on Israeli sabre rattling over alleged Iranian uranium enrichment.  On 22 May, crude futures had already topped $135 as the market balance remained tight, especially for distillates, before sliding down to around $122 in early June on the perception that rises in government-set (subsidised) retail prices in Asia would curb demand growth.

Global market fundamentals showed continued tightness, with constrained supplies and robust non-OECD demand growth.  Saudi Arabia made headlines when it announced a 300 kb/d rise in crude exports in early May, but this implies tightened effective spare capacity in the light of a further-delayed start-up to its major Khursaniyah project.  At the same time, the season's first tropical storm hit the Caribbean just in time for the US summer driving season, forcing Mexican ports to close and reminding markets that forecasts are for higher-than-usual hurricane activity this summer.  In India, an explosion at an offshore oil field shut-in some 40 kb/d, while US legislators forced a halt to SPR filling, freeing around 70 kb/d of crude for the market.

On the demand side, while government-mandated retail price rises in several Asian countries are likely to tame demand growth to an extent, pricing policy currently looks unlikely to change in the world's growth centres China and the Middle East, ensuring that consumption will remain robust there.  India, Indonesia, Malaysia and others announced hikes in government-set retail prices due to ballooning subsidy bills.  Elsewhere, while US gasoline demand growth is slowing due to high prices and growing ethanol production, diesel demand in Europe and elsewhere is surging.

Distillate crack spreads remain unseasonably high versus gasoline in all regions, reflecting a tight market.  Restarts at the Grangemouth refinery in Scotland and the hydrocracker at Porvoo's refinery in Finland should in theory take some heat out of the constrained European market.  News reports also indicate continued supply shortages in China.  Gasoline cracks showed signs of recovery in May as the US summer driving season started, but gasoline and fuel oil's relative weakness is still keeping refinery cracking margins low, at least in the US.

First Brent and later WTI forward curves shifted into near-term contangos from around mid-May as evidence of slightly looser crude supply pressured near-month contracts, even as expectations of longer-term crude tightness lifted up longer-term contracts.  Several well-known investment banks raised price forecasts for the rest of the year, underpinning the widespread impression markets will remain tight for the foreseeable future.  CFTC Commitment of Traders data showed retreating open interest since mid-May.

Spot Crude Oil Prices

Crude prices strengthened as refineries ramped up throughputs for the summer.  Given diesel's strength, distillate-rich crudes were sought after.  In the US market, LLS rose to a premium of around $4.50/bbl over WTI, though the latter was also pressured by rising crude stocks at the US delivery point of Cushing, Oklahoma.  High transatlantic freight rates also curbed US interest in Nigerian and other West African crudes.  Domestic US sour grades such as Mars narrowed their discounts to WTI, but again, this was rather due to the latter's easing, as Mars cracking margins remained heavily negative.

In Europe, Dated Brent remained strong on healthy cracking margins in NWE and the Med.  Brent strengthened against WTI over the course of May, eventually rising to a premium at the end of the month, though this has subsequently reversed again.  But in early June other light sweet crudes such as Azeri Light, Bonny and Tapis saw premia to Brent rising as gasoline cracks picked up.  Medium sour Urals maintained a wide discount to Brent on weak fuel oil cracks and growing Iraqi Kirkuk exports. 

In Asia, where refiners are set to come out of peak seasonal maintenance and diesel demand remains very high, again distillate-rich grades were on the rise.  Abu Dhabi's Murban shot up to a healthy premium to Dated Brent while Dubai and Oman wavered.  Heavier sour crudes were under even more pressure.  Press reports, backed up by shipping data, indicated that, in May, Iran was storing an estimated 15-30 mb of crude in tankers offshore Kharg Island, though some of this may now be flowing out of the Middle East.  Strong buying of West African and other Atlantic Basin grades continued to be seen, with Indian IOC for example boosting its July-delivery imports to the highest volume this year, around 375 kb/d.

Refining Margins

Refining margins were mostly down in May, month-on-month, as gasoline cracks fell on average, despite rising from their early month lows.  As to be expected, this is hitting US refiners harder than others, given its higher relative gasoline use, and US cracking margins were mostly negative.  US West Coast cracking margins fell particularly sharply as gasoline cracks hovered around $20/bbl - only half the spread for distillates.  Fuel oil discounts to crude also widened.  Singapore margins were mixed, as fuel oil's weakness kept Dubai's margins well below those of Tapis.  In Europe, margins were little changed, with cracking margins remaining far higher than elsewhere on distillates' strength.

Spot Product Prices

Gasoline crack spreads began to pick up in May and early June as the summer driving season (in the northern hemisphere) started.  Nonetheless, they remain well below their usual levels for this time of year, largely due to demand weakness in the US, exacerbated by growing volumes of ethanol blended into the pool.  On average however, May gasoline crack spreads were actually down, weakening refining margins.  Naphtha cracks also picked up, but remain negative, with the exception of Singapore, where they inched into positive territory in early June.

Distillate crack spreads on average gained ground in May, remaining a significant driving force behind strong oil prices.  Globally, supply remains tight, though the situation may be easing somewhat after the restart of several key refineries in Europe and following US refiners' steady hike in distillate yields.  Europe continues to draw in cargoes from around the world, with around 380,000 tonnes reportedly booked to load in June from the US and another 530,000 tonnes from Asia.  In Asia itself, China reportedly imported around 145 kb/d of diesel in May and plans to bring in nearly 160 kb/d in June, as regional product shortages remain, stockbuilding ahead of the Olympics reportedly continues and as continued coal shortages require more diesel generator use.  Some cooling in demand for the region may yet come from big importers such as Indonesia, which have recently raised government-set fuel prices.

Fuel oil discounts to crude remained wide in May, though were narrowing slightly in early June.  Some support may however begin to come from electricity use for summer cooling, especially in the Middle East, where the season has begun with hot temperatures and natural gas is, in some countries, in short supply.  Chinese demand - a key factor - remains lacklustre and is reflected in high Singapore stocks in independent storage.  Two million tonnes are due to arrive in Asia from the west, slightly lower than average volumes so far this year.

End-User Product Prices in May

End-user product prices on average increased by 8.1% in May in surveyed IEA countries - in US dollars, ex-tax - again hitting record levels.  Gasoline prices on average rose by 7.5%, but were even higher in the US and Canada, where prices jumped by 11.8% and 12.3%, respectively.  Diesel prices on average rose by 8.6%, with the highest increases observed in European countries at around 10%, in US dollars, ex-tax.  US drivers are now paying just under $1/litre for gasoline and Europeans around $2.20/litre for diesel (both with tax).  Heating oil retail prices on average rose by 10% and low-sulphur fuel oil prices increased by 5.5% in May 2008.  Compared to May 2007, retail prices are now approximately 60% higher than a year ago - ex-tax.


Tight vessel fundamentals pushed dirty tanker rates in the Middle East Gulf and West Africa to extreme highs in May.  With Iranian floating storage a major constraint on VLCC supply, sustained firm long-haul demand including increased Saudi exports lifted May rates to levels not seen since late 2004, excepting the end-2007 spike.  Shipping delays reduced vessel supply in the Caribbean, boosting regional Aframax rates.  Clean tanker rates in East of Suez markets rose in May.

According to the Saudi Arabian oil ministry, Saudi exports increased from 10 May and reports emerged that spot chartering by Vela, Saudi Aramco's shipping arm, was higher in May.  Middle East Gulf VLCC rates on benchmark routes to Japan rose to a peak of $37/tonne by mid-May.  Slow-steaming to save on growing bunker costs (now at around $580/tonne in Singapore) continues to act as a drag on tanker supply.  Tanker movement reports suggest that eastbound sailings (to China especially, according to reports) from the Middle East Gulf rose in May at the expense of westbound volumes.  If confirmed, this might explain an easing in VLCC rates later in the month, as a greater proportion of shorter eastbound routes reduces aggregate tanker demand.

VLCC and Suezmax freight rates leaving West Africa rose to even higher historical levels in May, reaching over $39/tonne on the 130 kmt WAF-US Atlantic coast trade.  West African light, sweet grades which have a high yield of light distillates have been in particular demand given the recent tightness in the diesel sector, with Bonny Light reaching a premium of $6/bbl over Brent prices in May.  Low VLCC availability may have also prompted cargo-splitting into Suezmaxes as well, spreading market strength across vessel sectors.  Still, while the higher freight costs may have added an extra $2/bbl-plus to US Gulf Coast refiners' delivered crude costs, their impact on refining margins has been overshadowed by the absolute volatility in physical crude and product prices.  Elsewhere, port delays due to bad weather have reportedly disrupted tanker traffic in the Caribbean, with export terminals closing in some cases.  The resultant drag on vessel availability boosted Aframax rates from $10/tonne in early May to around $24/tonne by the end of the month.

Clean tanker rates for large vessels exporting naphtha from the Middle East Gulf to Japan increased by $17/tonne in May to end the month at $41/tonne.  This is well above seasonal norms and coincides with the advance of Asian refinery maintenance and high product demand, particularly from China.  Transatlantic clean freight rates remained firm in May at around $36/tonne.



  • Global refinery throughput increased by 0.2 mb/d in May to 73.3 mb/d.  A strong recovery in US crude runs was overshadowed by weakness in the OECD Europe and Pacific regions.  Non-OECD crude throughput similarly increased during the month as higher Russian and Middle Eastern runs are estimated to have more than offset a decline in Chinese throughputs.
  • 3Q08 global crude throughput is forecast to average 75.7 mb/d, 2.1 mb/d higher than 2Q08 and 1.0 mb/d higher than 3Q07.  Year-on-year growth is entirely based in the non-OECD regions.  Crude runs are forecast to peak in August at a record-high 76.1 mb/d, before the start of autumn refinery maintenance in September weighs on throughput levels.

  • OECD 2Q08 crude throughput is projected to average 38.1 mb/d, 0.2 mb/d higher than assessed last month, as higher-than-expected April European and May US data boost the quarterly average.  Furthermore, upward revisions for US throughputs have been carried through to June and July. Overall though, as refining activity remains constrained by weak hydroskimming margins, 3Q08 OECD crude runs are forecast broadly in line with last year's average at 39.2 mb/d, but are still 1.1 mb/d above 2Q08.
  • 2Q08 non-OECD crude runs are forecast to average 35.5 mb/d, 0.1 mb/d lower than estimated in last month's report, but remain 0.5 mb/d above levels of a year ago.  Year-on-year growth accelerates during the third quarter, to an average of 1.0 mb/d, driven by China, Other Asia and Africa.  Chinese throughput growth will recover from the heavy refinery maintenance in May and should receive a further boost with the start of Sinopec's 200 kb/d Qingdao refinery in late May.  Further upside in 3Q08 is possible if the 580 kb/d Reliance Jamnagar refinery starts commercial operations ahead of our assumed 4Q08 start date.

Global Refinery Throughput

Global crude throughput levels should increase in June to 74.5 mb/d, from May's 73.3 mb/d, driven by stronger US crude runs.  However, weakness in refining margins remains critical to the activity level seen in many regions, as do planned (and unplanned) maintenance shutdowns.  Nevertheless, Atlantic Basin runs have recovered over the course of the second quarter and should continue to strengthen into the third quarter as refiners gear up for the peak gasoline demand season in the Atlantic Basin.

Elsewhere, the completion of maintenance work in China and Russia should increase the demand for crude in coming months.  Refiners remain caught between the markedly different product cracks which have continued to diverge in recent weeks.  While distillate production remains tremendously profitable, gasoline cracks remain subdued and fuel oil cracks have reached record lows against benchmark crudes.  Consequently, refiners are forced to seek ways to maximise distillate production at the expense of fuel oil and, more recently, gasoline.

Global crude throughputs for 2Q08 are estimated at 73.6 mb/d, 0.1 mb/d higher than in last month's report, following better-than-forecast US crude throughput in May (which we have carried forward into June and July) and higher than estimated OECD Europe April crude runs.  3Q08 global crude throughputs are forecast to average 75.7 mb/d, some 1.0 mb/d higher than the comparable period in 2007 and 2.1 mb/d higher than 2Q08.  Year-on-year growth is driven by China, Africa and Other Asia.  Crude runs are forecast to peak in August at 76.1 mb/d, before the start of seasonal autumn refinery maintenance in September reduce crude throughput to an estimated 75.3 mb/d.  The forecast does not include any crude processing at Reliance Petroleum's 580 kb/d Jamnagar refinery, which we assume will be commissioned in 4Q08, so further upside to our forecasts is possible.

OECD 3Q08 crude throughput is forecast to remain weak compared to the five-year range, reflecting the forecast 0.4 mb/d contraction in 2008 OECD demand and a weaker margin environment.  Utilisation rates could also be affected by more protracted maintenance.  Discussions with US refiners indicate that the weaker refinery economics are having an impact on the speed with which operational problems are resolved.  Where previously refiners would endeavour to resolve process issues as soon as possible, through the use of additional outside contractors, additional overtime and third party resources, management are taking a more sanguine view of the available solutions.  Consequently, the downtime from any particular problem has in some instances increased since last year.  While a strengthening of margins could shorten downtimes, in the US at least, weakening gasoline demand, rising ethanol blending and strong gasoline imports make such a scenario appear remote.

The OECD North American 2Q08 crude throughput forecast is revised up by 0.2 mb/d, largely on the back of the higher-than-forecast US crude runs in May.  3Q08 crude runs are forecast to average 18.5 mb/d, an increase of 0.1 mb/d from 3Q07, reflecting the offsetting impacts of the return to service of BP's Texas City and Whiting refineries, and the weaker margin environment.

US refinery crude throughput averaged 15.2 mb/d in May, according to provisional weekly data, some 0.1 mb/d ahead of our expectations.  Crude runs increased by more than 0.2 mb/d in the Midwest, and Gulf and West Coasts, as refineries completed planned maintenance and unplanned outages were resolved.  Further crude throughput increases are anticipated during June and July, particularly given the apparent dearth of planned refinery maintenance in the coming weeks.  US crude runs are forecast to peak at 15.8 mb/d in August, slightly ahead of the 3Q07 level.

The 2Q08 OECD Pacific throughput forecast is unchanged at 6.7 mb/d, despite provisional data for April indicating that runs were 0.2 mb/d higher than estimated.  Weekly data for Japan suggest that May runs were some 0.1 mb/d lower than forecast in last month's report, and upwardly revised maintenance for Japan in June similarly lowers our forecast for the month.  Crude runs at the end of May were reported at 3.4 mb/d, suggesting that the peak in maintenance activity is imminent.  Korean refiners are not expected to see peak maintenance activity until July, delaying the recovery of regional crude runs until August.  Crude runs at Australian refineries have been revised down for March and were weaker than forecast for April, in line with reports of problems with hydrotreating capacity.

OECD Europe crude throughput in April averaged 13.4 mb/d, 0.3 mb/d above forecast, almost entirely due to higher-than-anticipated Italian crude throughput (+0.4 mb/d vs. forecast).  Our assumption of heavy Italian planned maintenance in April appears misplaced.  Work may have been delayed to take advantage of the strong distillate cracks available, and may therefore occur later in the quarter, although strength in diesel cracks during May could push refiners to delay work until the autumn.  May crude runs are estimated to have dipped to 13.1 mb/d, following work at several German refineries and the delayed restart of INEOS's Grangemouth refinery following the late-April strike and subsequent hydrocracker fire.

Furthermore, reports indicate that in addition to the normal routine maintenance that occurs at European refineries during the spring some refiners have already started to upgrade hydrotreating capacity in preparation for the 10 ppm sulphur gasoline and diesel limits from 1 January 2009.  We had assumed that most refiners would undertake the necessary work in autumn 2008, but the fact that some refiners have already started may reflect tightness in engineering contractor markets.  OECD Europe 3Q08 crude throughput is currently forecast to average 13.8 mb/d, 0.4 mb/d above second-quarter levels and 0.1 mb/d above 3Q07 levels.

Chinese April crude runs were 6.7 mb/d, 0.1 mb/d ahead of forecast. Reports of ongoing product shortages would appear to keep the pressure on state oil companies to maximise crude runs in the coming months.  However, the poor refinery economics, due to domestic price regulation of transport fuels, will require heavy financial subsidies from the Chinese central government, in order to balance state oil company cash flows.  Furthermore, to bolster product supply and offset the shortfall from poorly utilised teapot refineries additional imports of gasoil to meet domestic needs are likely.  Despite all of the above, several refiners are reported to have carried out maintenance during May, leading us to project a 0.2 mb/d dip in estimated May crude runs to 6.4 mb/d. The start of commercial runs at Sinopec's Qingdao refinery was announced in late May, at 120 kb/d, equivalent to 60% capacity utilisation.  It is expected to remain at this level over the course of June and the third quarter.  These additional crude runs and the return to service of refineries from maintenance should restore Chinese crude runs to an average of 6.8 mb/d during June and 3Q08./p>

African crude throughput forecasts for 2Q08 are revised down by 0.1 mb/d on the back of much weaker-than-estimated crude runs in Libya in April (-0.2 mb/d), presumably following maintenance at the Ras Lanuf refinery, and planned maintenance at Sudan's Khartoum refinery during much of the quarter. Similarly, 2Q08 FSU crude throughput is reduced by 0.1 mb/d, following much weaker Russian crude throughput data for April. It appears that the heavy maintenance we had assumed would be taking place in March was largely seen in April.