SUMMARY

1. Short-term supply issues driving up prices. Weather and strikes with no let up in sight.

2. Demand slowdown. Evidence that H1 04 demand growth was indeed exceptional, although collapse seen as unlikely.

3. Global energy infrastructure tight. Until capital expenditure rises significantly the absence of spare capacity both upstream and downstream is expected to support prices. Medium term prices revised up.

4. Normalised oil prices higher and later. Costs still rising and capacity growth slower than expected.

In August's Crude Assessment we noted that in the absence of a material supply disruption, prices were unsustainably high. We highlighted the myriad risks out there (Norway, Venezuela, Iraq, Russia, Saudi Arabia, Nigeria and the US hurricane season) and cited that markets had no room for disruption. Well, two months on and we have seen supplies disrupted in Norway (extension of offshore workers strike), Russia (YUKOS exports to China), Nigeria (civil unrest) and most importantly the historically high level of hurricane activity in the US (Ivan, Jeane etc.). The biggest single disruption has been the impact of Ivan on US Gulf of Mexico production. Some 485kboe/d of oil and 1.8bcf/d of gas remains shut-in with rigs and pipelines damaged. The speed with which facilities are being restarted has been impacted by continued tropical storm activity in the region. Even if production does resume by the end of the month, the way in which production was initially shut down implies a loss in reservoir pressure that could compromise volumes for the rest of the year. The other supply outages have, like many so far this year, been small and temporary in nature. The problem for a market which lacks both spare capacity and sufficiently high inventory levels is that the sequence of events acts to drive prices to new record nominal highs.

The aftermath of hurricane Ivan represents the first large and sustained disruption we have seen this year. It could not have come at a worse time. The closure of production and manufacturing facilities in the US Gulf meant that seasonal inventory builds became substantial draws. This is especially true of heating oil. It seems that US consumers have been delaying the seasonal tank fill in the hope that prices (up nearly 80% this year) would fall back. The impact of hurricane related outages saw inventories fall further and prices rise higher. With weather in the north and north east already getting cooler the decision to delay tank filling is becoming increasingly difficult. Recent DOE inventory data suggest a surge in heating oil buying which has left demand days of inventory cover at only 40.6 days. This is 11 days below last year's level and over 6 days below the five year average. The markets already jittery behavior has been made worse by medium term weather forecasts by various agencies which suggest that this winter will be colder than normal.

So, with the first real sustained supply disruption we are faced with higher oil prices. What else is on the horizon? One should remember that the hurricane season hasn't finished yet. Further storm activity in the Gulf will only act to delay repairs there. The strike in Norway has extended to more fields and has shut in over 150kboe/d. Nigeria's general strike didn't impact export volumes, but the ceasefire is shaky at best and further trouble would spook markets. Brazil is on the verge of a general workers strike which could also disrupt supplies. Heading towards elections Iraq is expected to see heightened levels of violence which could curtail supplies. Plenty of support for prices to remain at these exceptional levels. The only bearish event we can see at this juncture is Kerry's success in the polls in November. Kerry has been campaigning on the message that he would actively use the SPR to cool high oil prices. Despite having to wait until January to put his family photos in the Oval Office, we would expect plenty of commentary about how much of the reserve would be issued even prior to inauguration. We would expect this to take the pressure off prices short-term.

2. DEMAND SLOWDOWN STARTING

We previously noted that H104 demand growth was exceptional compared to prevailing rates of GDP growth. We speculated about possible drivers for this but were unable to substantiate this due to the absence of data related to non-OECD inventories. We are starting to see some signs of demand slowdown. In China, rates of growth during Q2 04 were 25% yoy. In July this was 12% and in August fell to only 6% yoy. For some products (fuel oil) absolute demand has actually started falling. This might be related to new coal-fired power generation facilities which reduce the need to use inefficient independent generators (see chart). Is inventory infrastructure now at full capacity? Has a slowdown in consumption reduced the desire to import? The fact that crude price differentials between WTI and Dubai oil have blown out to historic highs recently suggests that Middle East barrels destined for Asia are having a harder time finding buyers.

Source: National Bureau of Statistics, China

Should we be surprised that demand growth appears to be waning? After all, prices are at historically high levels. This is particularly relevant to net importing Asian economies where the pressure to raise product prices is starting to be apparent. Many countries employ a price stabilization fund which attempts to cushion consumers from short term fluctuations in oil prices. That is, domestic product prices are not increased when international oil prices rise, rather the stabilization fund goes into deficit. This is all very well for small movements in oil prices, but this year's sustained rise in prices is starting to bite. Governments have not raised retail prices commensurate with rising crude and product prices, which may have acted to artificially support demand. Smith Barney estimates that Indonesia (now a net importer) has seen its oil deficit rise $6bn this year alone. Thailand has subsidies costing $1bn and Malaysia over $3.5bn. The decision to raise prices is politically unpopular and can result in general strikes (Nigeria being the latest example) but the longer these high prices persist, the greater the financial pressure will come to bear on government finances. With deficits growing daily governments will have to raise consumer prices which could slow demand growth.

There is a wide range of estimates for demand in 2005, with the IEA at 1.5mboe/d (having reduced it by 0.1mboe/d for China slowdown) and the DOE at 2.1mboe/d. Our own estimate of 1.7mboe/d matches that used by OPEC. Winter weather in the Northern hemisphere could well reduce price elasticity of demand during the next few months, but other economies where seasonal demand is less relevant should start to see demand impacted if prices do not fall materially from current levels.

3. INFRASTRUCTURE IS TIED UP

It would seem to be the case that oil infrastructure is insufficient to cope with the levels of demand seen year to date. Whilst some of this demand could well by unsustainable it remains the case that spare capacity is insufficient to prevent material upward price movements in the face of potential supply disruptions. In the upstream, public companies are transitioning towards a portfolio increasingly dependent on non-OECD countries. The move to develop resources which are both technically harder but also further away from major consuming markets has meant a substantial requirement to invest in infrastructure. For OPEC member states national oil companies have failed to grow capacity as much as previously hoped for due to a series of factors. OPEC market share has, until recently been declining. The ability of the cartel to manage prices has meant a surge in oil related revenues which have taken the pressure off governments facing rising social costs. Finally, disagreements over rates of return offered to western oil companies and political restrictions have delayed and prevented investments in countries such as Saudi Arabia, Iraq and Iran. It is also possible that individual grand standing by member states to position themselves for quota increases given the prospect of a post-sanctions Iraq led to aggressive forecasts of capacity growth.

The need to invest can also be felt in the refining sector. An absence of greenfield capacity additions in developed countries has meant operating rates above 90% and margins rising above mid-cycle levels. Global refining capacity actually peaked in 1981. The move by governments globally to tighten product specifications in an industry which traditionally earned cost of capital returns meant additional spend went on better kit, not necessarily bigger kit. Although debottlenecking remains an important source of growth, this is now seen an insufficient to cope with a demand slate which is getting progressively lighter (as transportation become the dominant driver) and a crude supply slate which is getting both heavier and more sour (basically more dirty).

The reinvestment requirement is estimated to be substantial. The IEA estimates over $6trn need to be invested in oil & gas infrastructure over the next 30 years. Goldman Sachs has a similar number of $2.4trn during the next 10 years. To put this in context this is three times the amount spent during the 1990s. If we look at the integrated oil sector we can see that upstream investment has recovered from the levels following the oil price collapse in 1998. That said, the $113bn spent last year is still someway short of the $200bn annual spend Goldman Sachs estimates is required to meet growth in oil demand. This is due, in part to the industry's growing love affair with capital discipline. Many integrated oil companies are planning on oil prices at $20/bbl. With a rising cost profile most companies are now seeing break-even at $20-25/bbl oil, not earning premium rates of return. The opportunity cost of using $20/bbl as a planning assumption when the cost structure has moved up and global infrastructure is tight is a growing concern. As a consequence it appears likely that capital expenditure will, at the very least, remain high. Indeed, with mature basin decline remaining at high rates it is imperative than investment increases simply to cope with conservative estimates of demand growth. The simple fact is that companies are targeting materially higher growth versus previous experience, such that current capex guidance suggests capital intensity actually falls. For the majority of companies in the industry we do not expect this to be the case.

As for OPEC, actual capacity has fallen by 1.7mboe/d during the past two years and expectations for future capacity additions have been reduced at a time when utilisation rates are at historically high levels. OPEC ex-Iraq is currently estimated to be operating at 99% utilisation. Capacity growth projections in countries such as Venezuela, Kuwait and Nigeria have all fallen. Deutsche Bank estimates a shortfall of capacity of 1.35mmbl/d within OPEC during the 07-04 period versus forecasts made nine months ago. We assume lower rates of growth in Iraqi capacity due to ongoing troubles there. The issue which requires resolution is the fact that western oil companies are planning on $20/bbl oil and the resource holders assume higher prices and therefore offer lower returns. The success or otherwise of current discussions in Libya will provide us with the first real evidence of how such an issue will be addressed given a $50/bbl backdrop. We expect the listed oil companies to next year usher in higher assumed oil prices for planning purposes which, at the same time, will be accompanied by higher capital expenditure budgets.

This requirement to invest, together with stronger than expected FY 04 demand growth and disappointing growth in OPEC capacity means that our oil price assumptions for 05-08 must be revised up. Part of the role played by higher prices is to ensure that the necessary investment is seen as sufficiently attractive to go ahead. This has resulted in material changes to our price assumptions which are detailed in the table below. We still see current prices as unsustainable and expect a correction of $15/bbl during Q2 05 (versus current prices). That said subsequent near-term price falls to below $30/bbl only look likely if we see a demand shock take place (keep watching Asia?).

4. HOW LOW CAN IT GO, AND WHEN?

The question of normalised oil prices is a vexed one. At the start of this year consensus suggested 2006 (most common year for normalisation) would see oil prices at $20/bbl. Now that number is $27.50/bbl, its rise coupled with strong sector relative outperformance. We have been using a normalised number of $20.25/bbl for 2006 but have reappraised this in light of 2003 cost performance and experience year to date from leading industry players.

Industry costs continue to rise, not only due to increased capital intensity but also due to higher operating costs and government take. Previously we had assumed that average unit costs would fall from levels inflated by disappointing volume growth and shift to non-OECD areas. We still believe this to be the case, but now expect costs to have to rise further prior to falling back. Costs of raw materials such as steel and cement have increased substantially this year and oil field equipment and service rates are starting to rise in many areas where previous capital excesses had worked in favour of the oil companies.

Governments have also acted to take some of the windfall from higher prices and redirect it into their own coffers. This is especially true of less developed countries but this is exactly where a large proportion of future investment is being made. So far this year we have seen taxes rise in Russia, Venezuela, Argentina and Kazakhstan with increased government take also likely in Bolivia and Angola. There is a degree of circularity in the argument that high oil prices result in higher tax take, but many of the changes announced recently suggest structural and not cyclical shifts.

As a result we have reworked our model of typical oil field economics to allow for higher capital, operating and tax costs. Assuming the required rate of return has remained unchanged we now estimate the normalised price $3/bbl above old estimates (i.e. $23.25/bbl). Previously we had assumed normalisation as a trigger for OPEC to cut production such that prices would average $1.75/bbl above normalised levels. On the new basis this gets us to $25/bbl Brent. We think this 'premium' is reasonable as the analysis is based on average costs for major integrated oil companies. This is on the generous side and certainly takes no account of the social costs which have to be met by OPEC member states. Also, the typical barrel of hydrocarbon resource is changing. Unconventional resources like gas-to-liquids, oil sands and tight gas are becoming increasingly important. These resources incur higher unit costs and therefore the analysis used to estimate normalised costs could be overly conservative.

Source: Schroder Global Energy

Having calculated our new normalised oil price we now have to ask ourselves when it actually occurs. Our previous assumption had been 2006 and indeed this remains the consensus view. However, due to the need for additional infrastructure and further growth in OPEC capacity we now expect normalisation in 2008. Again, the only likely driver to seeing prices at this level before 2008 is a demand shock.

The changes to our macro price deck are substantial. However, we would caution that in spite of positive revisions to equity price targets, we are becoming more negative on the sector at the margin. With the prospect of continued pressure on underlying returns and rising operating costs per barrel, the sector is increasingly reliant on a high and sustained commodity price deck. We would caution against aggressive overweight positions in this environment. For this reason changes to fair value will be more muted than the revisions suggest.