Released on July 27, 2005
(Next Release on August 3, 2005)

Growth in U. S. Petroleum Product Demand in 2005
One of the largest uncertainties facing U. S. and global oil markets this year and next is the likely pace of oil demand growth. As has been noted several times in this report, the unanticipated surge in global demand growth last year, on top of a return to pre-'90's growth rates in 2003, stripped out most of the capacity cushion, from the wellhead to the pump, worldwide. Recent oil prices can be largely, if not completely, explained by the lack of such cushions in a world with significant potential for supply disruptions.

While the debate continues as to the permanence of these developments, accurate data on current growth rates from the world's largest and most transparent consumer would seem helpful in assessing immediate and short-term pressures. EIA released its 2004 Petroleum Supply Annual (PSA) on June 24 containing updated final estimates on 2004 demand levels for all products. Today’s Weekly Petroleum Status Report (WPSR) incorporates these revisions. While 2004 demand estimates were revised upward for most products, with total demand revised up by 214,000 barrels per day, some analysts are concluding that this necessarily implies sharply lower demand growth in 2005 than previously projected. True, for the moment, new reported growth rates will decline some, due partly to the uneven nature of the comparisons with incomplete 2005 data, but final growth rates will remain “dampened” only if the 2004 revisions prove to be unusual.

Given recent reports issued by oil consulting companies that have revised 2005 Chinese demand growth down markedly, some oil analysts seemingly have surmised that with demand for the two main engines of recent growth (the United States and China) supposedly slowing dramatically, downward pressure on oil prices may be imminent. Leaving aside the issue of projecting a substantial slowdown in Chinese oil demand, even as China’s economy continues to grow at a remarkable rate, it is instructive to look at the 2004 demand revisions released in EIA’s PSA more closely in order to get a better understanding of the implications for U. S. demand growth this year.

As the chart above indicates, the 2004 data were not, in fact, revised upward by an unusual amount. In two of the previous six years, demand revisions were larger. Even if you analyze the revisions by end-use product, nothing unusual is apparent. For example, gasoline has been revised by a larger amount in 3 of the past 6 years, with revisions in 2000 more than twice that seen last year. Distillate fuel was revised only slightly; residual fuel oil has also been revised by a larger amount in 3 of the last 6 years; and even the source of the largest revision, other products, was more significantly revised in 2001.

Nevertheless, there are three observations that may have made these revisions particularly newsworthy. First, revisions for 2003 were atypically small. This may have led some analysts to expect 2004 data to be similarly unadjusted, but as the chart above shows, 2003 was an outlier, as total demand has been often revised by more than 100,000 barrels per day. Secondly, the 2004 revision comes at a time when demand growth appears (to some) to be unchecked by higher oil prices. Because this seems counter-intuitive to some analysts, there may be some tendency to view revised 2004 data-based comparisons as signaling a demand slowdown, partially attributable to high oil prices. However, the above chart casts doubt on such logic, because revisions are the norm and tend to be systematically biased upward, suggesting higher final 2005 levels are likely. The chart below shows that the main source of revision tends to be underreported imports, which are volatile and difficult to track accurately over a short period.

Another reason the 2004 data revisions have received so much attention may stem from the element of surprise. Why? Historically, EIA has produced a summary of the data resubmitted or revised after publication in the Petroleum Supply Monthly (PSM) in Table C1. However, this table has not been issued since July 2004, after EIA’s petroleum surveys and database system were changed to provide better coverage of petroleum product markets. Without it or some other mechanism for reporting on the evolution of final estimates, analysts arguably lost one source of advance insight into the upward revisions ultimately released in the PSA, especially given the abnormally low revisions in 2003. On the other hand, Table C1 typically has not included the bulk of resubmitted data until well after the fact. That is, most of the recent revisions come so late into the process that a significant gap may yet exist between final PSA data and projections developed using Table C1 information.

How does consideration of the revised 2004 data affect prospects for gasoline demand growth? Demand growth for the first 6 months of 2005 has been reported at a rate of 1.3 percent on average when compared to the unrevised 2004 PSM data. However, when using the revised 2004 data (revised up by a total of 60,000 barrels per day on average for the first 6 months of last year), the growth rate over the first half of this year drops to just 0.7 percent. But is this an accurate depiction of likely 2005 gasoline demand growth, even for just the first half of the year?

As noted in the first chart, including 2004, gasoline demand growth has been revised up in 5 out of the last 7 years, strongly suggesting that 2005 upward revisions are more likely than not. In fact, resubmitted data to date do confirm that higher 2005 demand levels and growth rates for the first six months will eventually materialize. Furthermore, during the second half of 2004, particularly in July and August, gasoline demand growth was reported as sub-par, even after final revisions were included, leaving open the possibility of relatively strong growth rates over the second half of this year given consensus forecasts for continued solid economic activity. While forthcoming revisions suggest that gasoline demand growth in the first half of 2005 will likely exceed the 0.7 percent rate recorded by published data so far, the second half of the year appears likely to grow at an even stronger rate. Thus, when all is said and done, growth in U. S. gasoline demand in 2005 seems more likely to average at or above trend growth (of about 1.5 percent) than below it. A table comparing the demand data in both the PSM and PSA for the major products in each month in 2004, is provided: 2004 Product Supply Data Comparison Table.

In summary, the upward revision in 2004 data will not necessarily pull down 2005 demand growth significantly. While high oil prices have certainly tempered demand, compared with the even higher levels that would otherwise have occurred, it is too early to conclude that demand growth will fall significantly, if at all, below normal rates.

U.S. Average Retail Gasoline Price Falls 3 Cents
The U.S. average retail price for regular gasoline declined for the second week in a row, falling by 2.8 cents from the previous week to 228.9 cents per gallon as of July 25, which is still 38.4 cents higher than this time last year. Prices were down throughout most of the country, but the West Coast saw an increase of 0.1 cent to 249.7 cents per gallon. The Midwest saw the largest decline, dropping 7.0 cents per gallon from last week to reach 219.8 cents per gallon, the lowest regional price in the country. The highest regional price remained on the West Coast, with prices in California even higher at 254.1 cents per gallon.

Retail diesel fuel prices were down 5.0 cents last week to 234.2 cents per gallon, the second week in a row prices have dropped. Prices were down throughout the country, with the Midwest seeing the largest regional decrease of 6.9 cents to 229.2 cents per gallon. Gulf Coast prices were the lowest in the nation, falling 5.3 cents to average 227.8 cents per gallon. Prices in California fell 1.1 cents to 257.8 cents per gallon, while the West Coast remained flat at 253.0 cents per gallon.

Propane Inventories Moderately Higher
U.S. propane inventories continued higher last week, but at a slightly slower rate of 1.4 million barrels, compared with the more robust stockbuilds reported over the past several weeks. Nevertheless, U.S. inventories of propane reached an estimated 59.0 million barrels as of July 22, 2005, just shy of the 60-million-barrel level some industry analysts consider as the minimum level for the start of the winter heating season. Moreover, with more than two months remaining in the summer build season, propane inventories appear well positioned for this time of year to show additional stockbuilding well into the fall months, with the possibility of adding to a fairly comfortable cushion of inventories prior to the start of the 2005-06 heating season. Weekly gains were reported in all regions last week with East Coast inventories up by 0.3 million barrels, while the Midwest and Gulf Coast regions reported higher inventories of 0.2 million barrels and 0.9 million barrels, respectively. The combined Rocky Mountain/West Coast regions rose by 0.1 million barrels during this same time. Propylene non-fuel use inventories remained unchanged at 5.2 million barrels last week, accounting for a slightly lower 8.8 percent of total propane/propylene inventories from the prior week’s 9.0 percent share.

Text from the previous editions of “This Week In Petroleum” is now accessible through a link at the top right-hand corner of this page.