Is the Petroleum Pricing System Out of Control?
by Matthew R. Simmons
President
Simmons & Company International
On June 15, 1998, crude oil hit its lowest level in over a decade, closing at $11.40 per barrel. Only 18 months earlier, it was over $26.
Regardless of where crude prices go from here, the recent price collapse has already gone into the history books as the second most severe drop in the price of oil in 50 years. The 56 percent drop from the start of 1997 to mid-June 1998 still lags behind the 65 percent drop in 1986.
However, $11.40 oil in June 1998, when adjusted for inflation to 1986 prices, is just over $8 per barrel, 20 percent below the 1986 low.
There is a complete contradiction between these low prices, the consensus view on supply and demand, and the published actual supply and demand statistics. This article discusses the perceptions and the realities of the present oil market situation.
The Perception and its Consequences
The consensus reasoning for the free-fall in oil prices is clear - too much supply and too little demand. The growth in oil demand is believed to have begun to slow in early July, when the Thai currency came under savage selling pressure.
This currency weakness quickly spread like a plague through Asia. General opinion was that the resulting economic slowdown, "the Asian Flu," would depress demand in what had been the fastest growing market for oil.
Compounding the expected demand collapse in Asia was the weather phenomenon "El Nino" and the impact it had on last winter's weather. The Northern Hemisphere recorded one of the mildest winters in recent history, and in fact, it appears the mild weather may have hurt demand even more than Asia's economic troubles.
If oil demand was deemed to be disappointing, conditions on the supply side were regarded as even worse. The view was that OPEC was out of control. Rather than be sympathetic to ever weakening demand as observers had expected, OPEC shocked the market last November by raising its production quotas by almost 2.5 million barrels per day.
If the conventionally accepted forecasts of a growing excess supply turn out to be correct, OPEC's move was a colossal mistake. Through the first quarter of 1998, daily supply of oil was believed by analysts to have exceeded demand by over 1.5 million barrels per day.
As weak as winter oil demand was, it declined even more once spring began, so by May, the supply/demand gap was estimated to be over 3 million barrels per day. Excess supplies appeared to become so plentiful that almost all storage was filled. With no spare storage, the incremental barrel has little value, which further depresses crude prices.
In an attempt to reverse this situation, a coalition of OPEC and non-OPEC producers have agreed to slash their production by 1.5 million barrels per day. Oil prices, however, have remained at low levels, partially because of oil traders' skepticism that the cuts will actually occur.
Where will this end? Could oil prices continue to fall? Could this usher in a prolonged slump similar to the 1930s, when oil sold for only a few cents a barrel? Will the June 24 OPEC and non-OPEC agreement be implemented? If so, will this be enough to "shake the shorts" out of crude oil contracts?
These are the key questions now raised in light of the sharp fall in oil prices. And, the impact of the plummeting oil price is beginning to hurt one industry participant after another. Virtually every economy driven by oil production is reeling from the recent drop in oil prices. Exploration and production companies are also starting to feel the pain. Many long-term supply additions are being postponed and the drilling of new wells is also slowing down.
Reality
As the bad news makes the circuits, however, the continuous revisions to estimates of both supply and demand paint quite a different picture than that suggested by the low oil prices. As real demand data trickles in it shows that, in fact, through the end of 1997, oil demand was still rising, in spite of the mild winter and Asia's troubles.
In fact, the most recently revised estimates show that 1997 worldwide demand was 73.8 million barrels per day, 200,000 barrels per day higher than projected at the beginning of 1997.
Moreover, while actual fourth quarter 1997 demand was indeed less than originally forecast, it was almost 600,000 barrels per day higher than estimated in March 1998, only three months after the quarter ended.
However, these demand revisions, which portray a scenario counter to conventional opinion, received little notice. It seems they were overshadowed in part by the collapse in oil prices and in part by the analysts' tendency to ignore or downplay estimate revisions as much as possible.
The supply side data shows the same picture of divergence between forecasts and actuals. The cumulative downward revisions in non-OPEC oil supply continue to grow and, in terms of a percentage revision, are now very large. As far back as 1996, the IEA estimated that non-OPEC supply would grow by 1.9 million barrels per day.
Instead, actual growth was 1.3 million barrels per day, a reduction of over 30 percent. These downward revisions to non-OPEC supply became even more pronounced in 1997. The IEA's original 1997 estimate was for it to grow by 2 million barrels per day. (This would, in fact, have been the highest yearly increase ever.)
Based on their latest data, the IEA's forecast has been reduced by 1.2 million barrels a day. In fact, non-OPEC supply only grew by 800,000 barrels per day, despite the fact that virtually every rig in the world was at work throughout the year.
The supply revisions were far greater at the "running rate" as 1997 came to a close. As recently as May 1997, fourth quarter 1997 non-OPEC supply was forecast to be 47.2 million barrels per day. Ten months later, actual supply turned out to be a staggering 2.1 million barrels per day lower. This is the equivalent of losing the oil production of a country like Kuwait.
Parenthetically, these non-OPEC supply surprises were, in fact, why OPEC quotas were raised last November. Continuing growth in demand for oil was forcing OPEC to overproduce its old 25 million barrels per day quota by as much as 3 million barrels per day to make up for the massive non-OPEC supply shortfall. They were simply admitting officially to what everyone knew they were doing!
The 1998 non-OPEC supply estimates continue to be adjusted downward. Last summer, fourth quarter 1998 non-OPEC oil supply was forecast to be at 48.8 million barrels per day. The most recent forecast is only 45.7 million barrels per day, a drop of over 3 million barrels per day!
And, these revised forecasts are before the negative supply impact from the recent price collapse, so one might expect further downward supply revisions before the year's end.
In my opinion, the likelihood that non-OPEC fourth quarter 1998 supply could come close to 46 million barrels per day is virtually nil, so the final "mis-estimate" or downward revision could approach 4 million barrels per day.
All this adds up to a strange divergence between the facts, as they are known today, and the official forecasts of oil's supply and demand. As factual demand data accumulates and more factual supply numbers are reported, most of the widely perceived bearish imbalances for 1997 have turned out to be illusory.
In fact, given the magnitude of the non-OPEC supply reductions, one could logically have assumed that oil prices should have strengthened as the year progressed. But, we know they fell. How can such a paradox be explained?
The "Arithmetic Mystery" of Missing Inventory Barrels
As surprising as all the changes in both demand and non-OPEC supply have been, they pale in comparison to what the IEA first described in its April 1998 monthly oil report as "Strange Mathematics" or an "Arithmetic Mystery."
It turns out that the physical measurement of petroleum inventories fails to support not only the perceived imbalance between supply and demand but also the most recently revised supply and demand estimates. As the IEA analysts now state, "the excess supply of oil is hiding" - they think it is there, the traders believe it is there, but the supposed physical excess has not shown up in inventory!
If the latest 1997 supply and demand estimates are correct, petroleum inventories should have grown by 292 million barrels by year-end. However, reported OECD stocks only grew by 100 million barrels. A more significant anomaly was apparent in the first quarter of 1998. Based on the latest estimate for that quarter, oil supply was 1.5 million barrels per day higher than demand.
Thus, petroleum stocks should have risen during the first quarter by a further 135 million barrels. However, actual data appears to show that stocks fell by 69 million barrels. When the imbalance for 1997 is added to the first quarter gap, the amount of oil "missing" from inventories comes close to 400 million barrels!
Oil analysts speculate that these missing barrels could be at sea or stored somewhere outside the OECD stocks; however, both these scenarios are highly unlikely, given the huge volume of missing crude.
For the oil to be "at sea" would have required almost 30 million deadweight tons of idle tankers to have been put into service quickly to handle the surplus for just the first quarter of 1998. But, according to the best estimates of total worldwide laid up tankers, the total available was less than one-tenth this amount.
There is also no evidence of a massive storage building program in non-OECD countries. In fact, non-OECD petroleum demand rose so fast over the past decade that it was hard enough to build storage for internal needs, never mind create a readily available surplus of storage.
These countries would have to have built enough pipeline and tank farm facilities on speculation to replicate the entire crude oil storage available in the U.S. Surely someone would have noticed.
What if the Petroleum Stocks Are Correct?
Unless a massive revision shows up in the reported OECD petroleum stocks, the only logical solution to what the IEA still calls "strange mathematics" is that the supply/demand imbalance so widely reported by analysts is completely wrong. Either demand must have been higher or supply lower than estimated, or both.
If the first quarter 1998 petroleum stock data turns out to be essentially correct, this also implies that the world oil markets were basically close to being in balance for most of the past year, despite the combination of mild winter weather throughout the Northern Hemisphere, Asia's troubles and the deep reductions which have already occurred in non-OPEC supply.
Moreover, if the missing barrels do not show up, then the collapse in oil prices was not a result of a vastly oversupplied market, but a trading reaction to the off-target supply and demand estimates.
What such a scenario also implies is that demand could have easily outpaced supply had the world not experienced the "El Nino" winter and the crisis in Asia. In fact, if we had had a really cold winter, world oil demand could have increased by as much as 2 million barrels per day during the winter months.
Whether the world's oil logistics could have coped with such a large stock drawdown is an open question.
But, Oil's Low Price is a Fact so We Must Have a Surplus
That the oil markets might have been close to being in balance is deemed almost impossible by most analysts who fervently believe that oil prices are a perfect indicator of fundamentals. Their view is that the price of oil always signals whether supply is too tight or too loose. Since the 1997/98 collapse in oil price was a fact, then they say it follows that the market must have been over-supplied.
This sounds logical, but it presumes oil's current pricing mechanism creates the accurate barometer that people believe. It is possible, however, that falling oil prices could have created a misleading impression of industry fundamentals.
This phenomenon certainly occurred in 1993 when MG (the German trading firm) drove oil from $18 to $13 a barrel as it unwound its massive speculative crude trading positions.
The fact is that daily oil prices are now set in the pits of the New York Mercantile Exchange (NYMEX), where the contracts for WTI crude oil are traded. This is self evident when one looks at the relationship between the daily closing price for the NYMEX crude contract and the price of West Texas Intermediate in the cash market on the same day. The two rarely vary by more than a penny or two.
Some analysts might argue that this demonstrates what a perfect market we have between the paper and the wet barrel. In fact, it highlights how imperfect this price mechanism is, since the closing price of the NYMEX paper barrel has nothing to do with oil on that specific day.
It is a contract, if held to the expiration date, to purchase 1,000 barrels of WTI equivalent grade oil in Cushing, Oklahoma, at a future period of time.
For example, June 22, 1998 was the last day for the July oil contract. The holders of contracts have 30 days in August to settle this transaction. So, the NYMEX closing price on June 22, 1998 had nothing physically to do with the actual oil markets on that day.
A day later, the closing price of the NYMEX paper barrel represented an ability to either sell or buy oil sometime in September, leaving the June 22 NYMEX price even further removed from any physical relationship to the fundamentals of supply and demand for oil in the third week of June.
But, the mechanism the oil business now uses to price oil ignores the difference in these dates. Ever since the pricing of the paper barrel supplanted the posted price for crude, the final price for the NYMEX oil contract each day almost always translates into a real cash spot price for West Texas Intermediate oil on the same day.
Then, the price of this relatively small stream of land-locked crude quickly sets the price for almost all other grades of crude around the world.
Since the correlation between the daily NYMEX price and the real cash price of WTI oil is essentially perfect, it is hard to refute the argument that the NYMEX oil traders now set the price for worldwide crude.
This is not necessarily good or bad, it is simply how crude prices are currently set. It serves, however, as a reminder that the world has come a long way from the days when posted prices were set by the Seven Sisters or when the Texas Railroad Commission set a floor on the price of domestic oil.
The need to understand how crude is now priced is not understanding the mechanism, per se, but to recognize that the price fluctuations created by NYMEX traders are not necessarily a perfect indicator of the real fundamentals of supply and demand. Those statistics are at complete variance with the consensus view which is reflected by the traders' pricing activity.
Could Poor Forecasting Have Created A "House of Mirrors?"
The real question which the "missing barrels mystery" raises is whether poor forecasting by industry analysts has inadvertently created a gigantic "House of Mirrors," in which bad supply/demand estimates are reflected from one observer to another, creating a misled consensus on fundamentals compared to what was happening in the physical market.
This creates a self-fulfilling prediction where, as prices continue to fall, the analysts grow more bearish, and tilt their subsequent demand guesses lower and their supply forecasts higher. This vicious circle creates more bearish news which encourages the traders to short oil even more, with the result that prices fall further.
This could easily be what is now driving crude prices. By the fall of 1998, the real supply/demand oil numbers for the first and second quarters of 1998 should be clearer. We will then know whether any major revisions to oil inventories were made. If there have been none, it would appear that the ultra-bearish "spin" of the energy analysts drove the price of oil down to levels which began to threaten the political stability of key oil producing countries.
If so, then perhaps it is time for senior industry executives to question whether some structural changes are needed in the oil industry's information system to ensure this does not happen again.
How Can the Information Be Improved?
Can the oil industry's "management information" system be improved? Unfortunately, forecasting petroleum supply and demand will always remain an art rather than a science. Start with the problems of forecasting demand accurately. The only way demand for most petroleum products can be estimated is by observing what disappears from primary inventory levels.
It can take years to get real data measuring consumer demand for some petroleum products. This is why reliable petroleum inventory data is so important (and why the missing barrel issue looms so large).
Forecasting supply is also difficult because oil production can be impacted by so many unpredictable factors. The ever-present "depletion" or decline curve makes it hard for a given producer to forecast precisely its net supply in any given period. Also, it is hard to forecast monthly production to 100 percent accuracy.
But, if the world supply estimate is off by as little as 3 percent, this creates a revision, one way or another, of over 2 million barrels per day. Given this lack of precision, one can question why the monthly publications are filled with scores of demand statistics that go down to tens of barrels. Moreover, supply information often contains multiple references to specific oil fields producing as little as 5,000 to 20,000 barrels a day.
All of this "trivia" creates an illusion that the industry crystal ball is almost perfect. As subsequent revisions show, it never is, and given the way petroleum data is created, it never can be.
Analysts compound the problem by rarely commenting on the magnitude of subsequent revisions other than simply noting that it is a revision from the prior monthly report. In fact, the cumulative revisions to both supply and demand are often far more relevant to understanding what drives the oil markets than the reams of supply/demand data for individual countries that are published each month.
Some readers of the data have questioned whether organizations like the IEA should even try to forecast oil supply and demand, given industry's dismal forecasting track record in recent years. If the reports dealt only with current data, and had far greater detail and analysis of the revisions which have been made to their historical data, might this not form a better statistical data base for the world's body of petroleum analysts to sort through?
Through such a change, might the IEA not get closer to its original mission of creating transparency in the oil markets? At the same time, readers would have a much clearer view of what is actually happening.
How the Pricing
Structure Might Be Improved
Correcting the industry's information system is a simple task compared to attempting to change the current NYMEX paper barrel system that sets world oil prices.
But, there are ways in which a more rational basis for setting oil prices could occur. One way would be for most buyers and sellers of crude oil and petroleum products to return to using longer-term supply contracts at prices which reflect the real cost to insure a safe and reliable supply.
The other would be for the industry participants to play a more aggressive role in the trading of the NYMEX oil contract, but using trading not as a potential profit center but as a way to smooth out a price for the end product. This is common for agricultural companies today, and through practicing this type of trading, industry executives would be forced to analyze what the product price should be.
Is The System Out of Control?
Only time will tell whether the system is out of control. The answer almost certainly lies in resolving the mystery of the missing barrels of petroleum stocks. But, even if the "missing" stocks are finally discovered, we are still left with the issue of the integrity of the industry's supply and demand estimates.
Any time supply estimates fall by close to 3 million barrels a day in less than a year, an alarm bell should go off that something is amiss with the system of forecasting.
What has become crystal clear is that a $11.40 price for crude will quickly eliminate a large amount of supply, because few key producers can tolerate such low levels. And most of the announced OPEC cuts will likely happen, since the need to get prices higher is important to the well-being of too many key producing countries.
However, if the price collapse turns out to merely be an artificial reaction to bad data, then the current system must be corrected, as it is out of control.
Mr. Simmons is President of Simmons & Company International, a specialized investment banking firm serving several segments of the worldwide oil service industry. The firm has helped its client base in over 300 mergers, acquisitions, public and private financings, restructurings and other financial advisory tasks at a combined value exceeding $20 billion.
He graduated cum laude from the University of Utah and received an MBA with distinction from Harvard Business School. Mr. Simmons founded Simmons & Company Intern'l in 1974 with a goal to concentrate the firm's services on companies providing service and equipment to the oil and gas industry. Over the past 23 years, the firm has played a leading role in helping oil service companies execute mergers, acquisitions and restructurings along with strategic financial advice.
|
