The fear premium
Do current oil prices reflect market fundamentals or some other set of factors? Over the past year the apparent relationship between industry stock levels and crude oil prices has shifted dramatically, prompting some commentators to claim that oil prices are being inflated by forces other than market fundamentals.
Many analysts have argued over the years that there is a stable relationship between inventories and oil prices, which can be used to determine an equilibrium price that reflects the underlying market fundamentals. This view is justified by a scatter graph comparing historical stock and oil price data that shows a broadly inverse correlation between oil inventories and prices. In other words, high stock levels are associated with low oil prices and vice versa. However, this relationship seems to have broken down since the middle of last year as crude oil prices soared beyond their historical range at a time when industry stocks began to recover from an extended period at very low levels.
Stocks and spreads
There is no doubt that stocks play an important role in commodity markets and that they do affect prices; however, there is a view that the economic rationale concerns the shape of the forward curve rather than the general level of prices. The original investigation into the relationship between inventories and forward price spreads was carried out in the US wheat market by Holbrook Working, an American economist and statistician who specialised in the behaviour of commodity markets, but his findings have subsequently been found to apply to a wide range of other commodity markets, including oil. Working found that spot prices for wheat rose (or fell) relative to wheat futures prices inversely with the level of wheat inventories held by the industry. When wheat inventories were high, spot wheat prices were lower than wheat futures prices, encouraging storage operators to hold the commodity. This shape of the forward curve, with spot prices below those for future delivery, is known as a ‘contango’. On the other hand, when wheat inventories were low, spot wheat prices were higher than wheat futures prices, encouraging storage operators to release the commodity into the market. This shape of the forward curve, with spot prices above those for future delivery, is known as ‘backwardation’.
The same inverse relationship between stocks and forward price spreads can be observed in the oil market (see GOR passim). The best correlation between the two is obtained using forward price spreads for WTI futures contracts and commercial crude stocks in the US Mid-Continent (PADD II) area — chosen because it contains the physical delivery point at Cushing, Oklahoma for the NYMEX light sweet crude oil contract — as recommended by the distinguished oil analyst, Walter Greaves.
Previous work by the CGES demonstrated that there was a good inverse relationship [with a respectable correlation coefficient of around 66%] for monthly PADD II crude oil stocks and forward price spreads between the nearby and 12-month ahead WTI futures contract over a long time-period from January 1995 to September 2002 1(see GOR September-October 2002). A similar, but weaker, correlation was also observed between total US crude oil stocks and WTI forward price spreads, suggesting that the differential movement of stocks in the land-locked PADD II area close to the delivery point for the WTI futures contract has a specific and significant influence on forward price spreads in the WTI market, which reflects a combination of international price influences through its role as a global price reference marker and local price influences as a result of physical constraints in the area.
What matters in the relationship between inventories and prices are the relative price levels along the forward curve rather than absolute prices at the front end of the forward curve. This is because what is being traded is the precise timing of access to the physical commodity rather than the physical commodity itself – which is why the relationship between stocks and forward spreads is stronger than the relationship between stocks and absolute prices. The fact that there appeared to be a stable relationship between stocks and absolute prices over the period from January 1995 to May 2004 therefore merely reflects the fact that oil futures prices moved in a fairly narrow range with most of the variation in the absolute price of oil being driven by changes in the forward spread that — in turn — reflected the changing level of stocks (or stock cover). As Figure 6 shows, while the nearby WTI futures price fluctuated in a wide $25/bbl range from $11–36/bbl over the period 1993 to 2003, longer-term futures prices were much more stable, fluctuating in a narrower $7/bbl range from $15-22/bbl.
What has changed in the past year is not the basic relationship between stocks and spreads — which has remained intact as demonstrated above — but the general level of oil futures prices, which has risen from under $30/bbl at the start of 2004 to over $50/bbl in 2005. Thus, as the oil market has switched from backwardation in 2004 — when crude stocks were still very low — to contango in 2005 as industry crude stock levels have recovered, the forward spreads have continued to behave as before while outright prices have soared into an entirely new trading range.
Changing expectations
The answer lies in the fact that stocks are only part of the story. Inventories held by companies represent less than two months of forward supplies and most of this is effectively inaccessible because it is necessary for the operation of the supply chain. Once minimum operating levels are taken into account the industry only has a few days of discretionary stocks at its disposal – which is why the shape of the forward curve is very sensitive to changes in stock levels. However, stocks and stock cover can only ever be a proxy for the wider forces of supply and demand that ultimately determine the general level of oil prices. They provide a useful indicator to monitor the changing balance of supply and demand and therefore the shifting relationship between prices along the forward curve, but they cannot explain why the oil price is $10/bbl or $60/bbl. This requires a different type of explanation that extends beyond the short-term market fundamentals that are captured by changing stock levels.
Earlier work by the CGES has argued that longer-term futures prices provide a useful insight into market expectations (GOR, Volume 13 Issue 5, Sep-Oct 2002). In particular, it showed that longer-term price expectations were relatively stable for extended periods and that changes in short-term prices could be accounted for by a combination of changes in the shape of the forward curve and changes in price expectations. Thus longer-term price expectations appeared to provide an anchor for the general level of oil prices around which short-term prices move to reflect shifting market fundamentals.
Why are oil prices so high?
A number of factors appear to be driving this dramatic change in price expectations. First of all, the rapid demand growth experienced over the past two years has pushed up capacity utilisation rates both upstream and downstream. OPEC’s margin of spare capacity has shrunk to less than 2 mbpd (just over 2% of the world’s peak winter consumption) as a result of chronic under-investment and political problems in key producing countries. At the same time, global refinery utilisation rates are at their highest levels for over 25 years and are getting close to full capacity at peak seasonal throughputs in key regions such as North America, Europe and Asia Pacific.
Secondly, there are increasing quality constraints that are distorting crude and product price differentials. Growing demand for low sulphur middle distillates is stretching refinery capacity to the limit — opening up huge premiums for both the product itself and the heavy, low sulphur crude oils required to make it. With refinery upgrading and de-sulphurisation plant already fully utilised, the marginal product barrel is supplied by straight-run distillation. At the same time, the marginal crude barrel is getting heavier and more sulphurous as OPEC’s remaining spare production capacity is concentrated in Saudi Arabia, which can only supply extra barrels of Arab Heavy.
Thirdly, there is a growing concern that investment will not be able to keep up with potential oil demand growth — especially in highly-populated and newly industrialising countries such as China and India. Outside OPEC, oil companies remain reluctant to raise the price hurdle used to test new projects. Despite long-term WTI futures prices well above $50/bbl for the next ten years, companies still base their investment decisions on crude prices around $20/bbl, because they are worried that high prices may not persist. This is not only a form of capital rationing, but also reflects the fact that the companies are running out of opportunities for major new upstream investments outside the Middle East and Russia, now that non-OPEC conventional oil supplies seem to getting closer to their peak.
OPEC itself also seems reluctant to boost upstream expenditure on the scale necessary to meet future demand growth. OPEC’s own production capacity has barely changed during the past 25 years and governments will need to commit large sums over the next decade if capacity is to expand to meet the demand potential from China and India. Even Saudi Arabia — owner of the world’s largest reserves of untapped oil — plans to boost its net production capacity by only 1.5 mbpd between now and 2009. Moreover, it is clear that OPEC would be better off in the short-term by restricting upstream investment if it can keep prices high as a result. If non-OPEC producers can no longer compete effectively by expanding capacity when oil prices are high, then OPEC in general — and Saudi Arabia in particular — are in a position to support prices well above the cost of supply.
Taken together, these three factors go a long way towards explaining why there has been such a dramatic upward shift in longer-term oil price expectations. During the 1990s, longer-term expectations lay in a $15-20/bbl price range that reflected the fully-built-up cost of developing new non-OPEC supplies. At the time there were still plenty of opportunities for new upstream developments by companies outside OPEC and higher prices would have encouraged companies to invest more, thus boosting supply and reducing prices once again.