The 1986 Oil Price Collapse: How Saudi Netback Deals Broke OPEC

In August 1985 Saudi Arabia produced about 2.2 million barrels of crude oil a day. Norway, a country of four million people whose first offshore field had come on stream in 1971, was producing roughly three-quarters as much. The Kingdom sat on a quarter of the world's proved reserves, had built the export capacity to lift more than ten million barrels a day, and was pumping less than it had in 1962.

Every barrel of that decline had been voluntary. Under the quota system OPEC adopted in March 1983, Saudi Arabia had taken the role of swing producer: the member that cut whenever the others overproduced, absorbing the arithmetic so that the posted price could stay where the organisation had set it. The arrangement held the marker price at $29 a barrel and cost the Saudi treasury its solvency. Oil revenue, about $119 billion in 1981, came to roughly $26 billion in 1985. The government ran deficits, drew down foreign assets, and watched its customers buy from Mexico, the North Sea and the Soviet Union instead.

On 13 September 1985 Ahmed Zaki Yamani, the Saudi oil minister since 1962, told his fellow ministers that the Kingdom would stop doing it. Within four months the price of a barrel of oil had fallen by two-thirds, hundreds of American banks had begun a slow progress towards insolvency, and the largest foreign-exchange earner of the Soviet Union had stopped earning.

The Arithmetic of the Swing

High prices had done what high prices do. The two shocks of the 1970s — the embargo described in the Yom Kippur war and the oil shock of 1973 and the disruption that followed the Iranian revolution — had pushed crude from under $3 a barrel to a peak above $35, and the response arrived with a lag of about five years. Industrial users switched to gas and coal. Japanese cars replaced American ones. Fields that made no sense at $3 made excellent sense at $30, so Prudhoe Bay, the Forties and Ekofisk fields, Cantarell and a series of Siberian giants all came on stream at once.

Demand and non-OPEC supply moved against the cartel simultaneously. World oil consumption outside the communist bloc fell for four consecutive years after 1979. OPEC's own output went from roughly 31 million barrels a day in 1979 to about 16 million in 1985, and almost the whole of that reduction was borne by one member.

Saudi Arabia was also being cheated, which mattered more than the volumes. Members with quotas of two million barrels a day were selling three, discounting through barter deals, extended credit and adjusted freight terms that did not show up in the posted price. The Kingdom was defending a price that its partners were quietly undercutting in order to sell the barrels it had given up.

Netback

What Yamani chose instead was a pricing formula rather than a volume target, and the choice of formula is the whole of the story.

A netback deal prices a cargo backwards from its destination. Rather than agreeing a dollar figure per barrel at the loading terminal, buyer and seller wait to see what the refined products — gasoline, gas oil, fuel oil — fetch in the market where the cargo lands. From that realised value they subtract freight, insurance and refining costs, then subtract an agreed margin for the refiner. Whatever remains is what the crude cost.

Saudi Arabia offered these terms in September and October 1985 to Exxon, Mobil, Texaco and Chevron, the four American companies that had owned Aramco before nationalisation. The offer was irresistible for a simple reason: a refiner on netback terms cannot lose money on the crude. Its margin is guaranteed in the contract. Refining capacity that had been idled for years came back, and the refiners took every barrel the Kingdom would sell, because there was no price at which the barrel was uneconomic to them.

Saudi output rose through the autumn of 1985 and kept rising. Other producers matched the terms to keep their customers, and by early 1986 a large share of internationally traded crude was moving on netback or netback-linked contracts. Robert Mabro, whose analysis for the Oxford Institute for Energy Studies remains the clearest account of the mechanism, argued that netback pricing did not merely coincide with the collapse but accelerated it: by removing the refiner's incentive to resist a high crude price, it transmitted every weakness in the product market straight into the crude price, and then fed the resulting volumes back into the same product market (Mabro, 1987).

OPEC ratified the strategy at its Geneva conference in December 1985, resolving that the organisation would "secure and defend for OPEC a fair share in the world oil market consistent with the necessary income for member countries' development." The communiqué contained no price and no quota. Traders read it correctly.

Free Fall

Crude was near $31 a barrel at the end of November 1985. It was near $20 by the end of January 1986, near $13 in February, and trading around $10 by the last days of March. Spot cargoes of heavier grades changed hands below that.

Washington could not decide whether this was a victory. Cheap oil transferred something on the order of two per cent of national income from producers to consumers, and the administration had spent five years arguing that markets, not ministers, should set the price. American production was also falling, and the men who ran the Gulf Coast and the Permian Basin were Republicans.

Vice President George Bush, a former Texas oilman, flew to Saudi Arabia in early April 1986 and told reporters before his meetings that he would make the case for stability, adding that it was important "that we not just have a continued free fall like a parachutist jumping out without a parachute." The White House spent the following days explaining that American policy remained a preference for market pricing and that the Vice President had been describing, not proposing. Yamani, asked about the visit, observed that the Kingdom had no quarrel with market forces.

The Oil Patch

The American drilling industry recorded the collapse more precisely than any price series. The rotary rig count compiled weekly by Baker Hughes had peaked at 4,530 in December 1981. By the third week of July 1986 it stood at 663 — a reduction of about six rigs in seven. Crude production in the lower forty-eight states began a decline that ran for a quarter of a century, and the share of American consumption met by imports, 27 per cent in 1985, passed 40 per cent before the end of the decade.

Banking damage followed with a lag, because loans secured on oil reserves are revalued slowly. Energy lending in Texas, Oklahoma and Louisiana had been written against reserve estimates discounted at $30 a barrel and above, with no covenant contemplating $10. When the reserve engineers repriced the collateral in 1986 and 1987, the loans were not merely impaired; they were unsecured. Commercial property in Houston and Dallas, built on the assumption that the oil economy would keep hiring, emptied at the same moment, so the two largest asset classes on every regional bank's book failed together.

First RepublicBank Corporation, the largest banking company in Texas with about $33 billion of assets, was resolved by the Federal Deposit Insurance Corporation in July 1988 — the biggest bank failure in American history to that date. MCorp followed in March 1989, Texas American Bancshares that summer. Nine of the ten largest Texas banking organisations failed or were sold under federal assistance between 1987 and 1990 (FDIC, 1997). The thrifts in the same states, already carrying the interest-rate losses described in the savings and loan crisis of 1980–1995, were finished by the same collateral.

Debtors and an Empire

Outside the United States the accounting was starker, because oil exporters had borrowed against the barrel.

Mexico earned about $14.8 billion from oil exports in 1985 and about $6.3 billion in 1986. An economy that had defaulted four years earlier, in the episode covered in Mexico's 1982 moratorium and the lost decade, contracted by nearly four per cent in 1986 while inflation passed a hundred per cent. The standby arrangement Mexico negotiated with the International Monetary Fund that year carried an oil contingency clause, a device the Fund had not used before: additional financing would be released automatically if the price fell below a stated floor. Nigeria, Venezuela and Indonesia went through smaller versions of the same year.

The Soviet Union had no such facility. Oil and gas supplied the hard currency that paid for imported grain, machinery and consumer goods, and the price of both fell together. Yegor Gaidar, the economist who became Russia's acting prime minister in 1992, dated the terminal fiscal crisis of the Soviet state to the autumn of 1985, when Saudi Arabia stopped restraining output, and put the resulting annual loss to Moscow at roughly $20 billion (Gaidar, 2007). Grain purchases were sustained with borrowed money for another five years.

For importers the transfer ran the other way. American consumer prices rose 1.9 per cent in 1986, the slowest annual rate in twenty-one years. The Federal Reserve cut the discount rate four times that year, from 7.5 per cent to 5.5 per cent. Thirty-year Treasury yields, above ten per cent in mid-1985, fell towards seven per cent by the spring of 1986, and the Dow Jones Industrial Average finished 1986 at 1,895.95 against 1,546.67 a year earlier — the middle of the advance that ended in the crash of October 1987. James Hamilton's work on the macroeconomics of oil shocks notes the asymmetry that makes such episodes hard to read: large price increases have reliably preceded American recessions, while large decreases deliver a weaker and slower stimulus than the symmetry of the arithmetic would suggest (Hamilton, 2011).

Eighteen Dollars

By the summer of 1986 the market-share war had achieved its stated aim and made the Kingdom poorer doing it. Saudi output was back above five million barrels a day, roughly half again what it had been in 1985, but revenue per barrel had fallen by more than half. OPEC met in Geneva on 5 August and agreed a temporary return to quotas for September and October; the price recovered above $14 within weeks, which told the ministers what a credible ceiling was worth.

The political reckoning came on 29 October, when King Fahd dismissed Yamani. He had been oil minister for twenty-four years, had been kidnapped by Carlos the Jackal at the 1975 OPEC meeting in Vienna, and had been the public face of the price the collapse destroyed. His successor, Hisham Nazer, inherited a negotiation already under way.

On 20 December 1986 OPEC agreed to abandon netback pricing and return to a fixed reference price of $18 a barrel for a basket of seven crudes, with a production ceiling of 15.8 million barrels a day from 1 February 1987. Prices did recover to something near that level for a time. The system did not hold. Members discounted, the basket drifted, and within about two years Saudi Arabia and the rest had moved to formula pricing — crude sold at a differential to a published spot benchmark such as Brent or West Texas Intermediate, with the benchmark set by traders rather than by ministers.

That is the change 1986 actually made. Before it, the price of oil was a number a committee announced. After it, the price of oil was whatever the futures screens said, which is why the hedging strategies that later destroyed an oil hedge worth $1.3 billion at Metallgesellschaft were possible at all, and why a state's budget could be undone by a quarterly move in a contract traded in New York, as Russia found in the rouble crisis of 2014. Daniel Yergin's account of the period reads the collapse as the moment the producers discovered that the instrument they had seized in 1973 could not be held without cutting production until it hurt (Yergin, 1991).

The $18 reference price agreed in Geneva two days before Christmas 1986 survived less than two years of quiet discounting, and no one has tried to post a fixed price for crude oil since.