Mexico's Petroleum Exports:
Safe Collateral for a $50 Billion Loan?

Richard C. Duncan, Ph.D.
Institute on Energy and Man

[note: graphics under construction]

It is illusory, and would be harmful, to pretend that petroleum produced and exported in large quantities could become the factotum of Mexico's economy or the panacea for Mexico's economic ills. Mexico does not wish ever to be forced to export such an indispensable energy and chemical resource.
Antonio J. Bermúdez, 1963
Director General
Petroleos Mexicanos, 1947-58


In the mid-1990s, Mexico found it increasingly difficult to attract foreign capital, especially into its energy sector. Then, within a week in December 1994, the Mexican peso fell by 35 to 40%. The Mexican central banks' international reserves, which had stood at $29 billion, plunged to $5 billion. In an effort to stave off a collapse of the Mexican economy, on 31 January 1995 the President of the United States, by executive order, signed a $50 billion "Emergency Stabilization Package" loan to the Mexican government. The collateral for the loan was Mexico's pledge of revenues from its future petroleum exports. (PE, 1995b)

This study uses Mexican petroleum production data and a robust geophysical approach to answer the question, Are Mexico's oil exports safe collateral for the 1995 $50 billion Emergency Stabilization Package (ESP) loan? This question is answered in the context of Mexico's capacity to simultaneously repay the ESP loan, and meet the requirements of the 1993 North American Free Trade Agreement (NAFTA), and complete its plan for industrial development.


The creation of Pemex in 1938 helped establish firm respect for state sovereignty both at home and abroad. Thus, Mexicans believe that what happens to Pemex will directly affect the national dignity, economic independence, and state sovereignty of their country.
David Ronfeldt, 1980

A history of Mexican petroleum production is illustrated in Figure 1. Production began in 1901, and, spurred by a strong demand for oil by the allies during World War I, it peaked sharply in 1921 to mark Mexico's "first oil boom" Although the 1921 peak is small compared to recent rates, from 1918 to 1928 Mexico ranked second only to the United States in world oil production. (Powell, 1956; Bermúdez, 1963)

Figure 1
View image (typically right mouse button)
to see graphs full size

[Note 1: In this study, the notation "DQ/DT" (i.e., plain font, upper case) designates a discrete, historic data point. In contrast, the notation "dQ/dt" (i.e., italic font, mixed case) designates a continuum of values generated by an equation, as discussed later in Sections 4 and 5.]

In the early years, the Mexican oil industry was dominated by American and British companies. However, in the 1930s with national feelings rising and oil production falling, and open conflict between the foreign oil companies on one side and the Mexican government and labor on the other, the days of outside domination were numbered. On March 18, 1938, the Mexican government expropriated all foreign oil assets and, by June 7, 1938, the petroleum industry was nationalized in the shape of Petróleos Mexicanos, or "Pemex" for short. (Bermúdez, 1963)

Mexican oil production stagnated from 1938 until the mid-1940s (Figure 1). Then from 1944 to 1973 production increased at a vigorous average rate of 15% per year (i.e., from 0.11 to 0.55 million barrels per day). After the discovery of large new fields in the 1960s and the sharp jump in world oil prices in 1973, Mexico's "second oil boom" took off. From 1973 through 1982 oil production increased by a whopping average of 50% per year (i.e., from 0.55 to 3.0 million barrels per day). However, from 1982 through 1994 production slowed dramatically (Figure 1) with an average growth rate of only 1% per year (i.e., from 3.0 to 3.3 million barrels per day).

[Note 2: In this report, the words petroleum and oil are used synonymously to mean crude oil and natural gas liquids, "NGLs" - the liquid content of natural gas wherever this is recovered separately.]

The next section focuses on oil production, consumption and "surplus."


Mexico's oil reserves will match those of Saudi Arabia.
Jose López Portillo, 1978
President of Mexico, 1976-82
(in Metz, 1978)
An analysis not dissimilar from that of Venezuela can be applied to Mexico, where newly discovered reserves of oil are, as usual, much exaggerated.
Aubrey Jones, 1981

President Portillo's prediction that Mexico's oil reserves would match those of Saudi Arabia was (and is) flatly wrong. "Oil in place is an abstraction. For Mexico during L-pez Portillo's time, it stood as the foundation of Mexico's line of credit with international bankers." (Baker, 1984) Others too were skeptical.

Is Mexico a new Saudi Arabia? ... By its official estimates, ... Mexico has about 60.10 billion barrels of proven reserves; Saudi Arabia ... has nearly three times that amount or 163.35 billion barrels ... Moreover, Saudi Arabia's production costs are in the neighborhood of thirty-five to fifty cents per barrel, a fraction of the Mexican figures. ... Mexico's large, fast-growing population now consumes six out of ten barrels produced in the country ... In contrast, Saudi Arabia, with fewer than eight million people ... can export almost all of its output. (Grayson, 1980)

But the exaggerated predictions of Mexican petroleum reserves (e.g., described as "an oil bonanza," "prolific wells," "unlimited quantities," and "a veritable sea of oil") were taken seriously by some high American officials on energy. For instance, early in 1979 U. S. Energy Secretary James Schlesinger affirmed, "The Mexican oil reserves prospectively are as great as those of Saudi Arabia." (Gentleman, 1984)

[Note 3: A word of caution before we proceed. There exists a pervasive confusion about the difference between oil resources and oil reserves. Suffice it is to say that (1) reserves are but a small fraction of resources, and (2) the word reserves itself is clouded by an infinite regress of artificial categories such as, "proved," "probable," "possible," "potential," "economic" and "sub-economic." Geologist C. J. Campbell (1991) added yet two more categories, "political" and "suspect." Not to be outdone, a former Mexican Minister of Natural Resources leaped to heavenly heights with "theological" reserves. (Williams, 1979) "We have in the word reserves a chaos of meaninglessness." (North, 1985) In short, most estimates of Mexican oil reserves are not useful for serious study.]

Historic data shows that the Mexican oil industry is now mature. No unexplored frontiers remain. No "major" fields lie coyly awaiting discovery. The last major Mexican oil field was discovered in 1978. (Campbell, 1991; 1995b) ("Major" means an ultimate recovery of 500 million barrels or more.) Fortunately, however, estimates of Mexican reserves are irrelevant to this study, because ample historic data is now available to confidently predict the life-cycle of Mexican petroleum production.

Figure 2 compares Mexican oil production with that of the United States, Texas and Saudi Arabia. Several features are noteworthy. First, the data confirms that Mexico is not "another Saudi Arabia." For instance, Saudi Arabian oil production reached a remarkable 10.0 million barrels per day (in 1980) compared to Mexico's maximum (to date) of 3.3 million barrels per day (in 1994).

Figure xxx

Second, comparing Mexican oil production to that of the United States (Figure 2), even with the addition of Alaskan oil and the remarkable advances in exploratory and production (OE & PO) technologies, from 1970 through 1994 total United States oil production decreased by 25.7% (i.e., from 11.3 to 8.4 million barrels per day) and Lower-48 production decreased by a dizzying 45.0% (i.e., from 9.4 to 5.2 million barrels per day). (API, 1994; BP, 1995) The decline is an example of the Theory of Depletion Dominance: "During the down-side of the life cycle of petroleum production, resource depletion dominates both price and technology." (Duncan, 1995a) In fact, resource depletion is already tightening its jugular grip as Mexican petroleum production nears its all-time peak, discussed later in this report.

Third, the history of Texas oil production sheds light on the future of Mexican oil production (Figure 2). Texas production started in 1889 and peaked in 1972 at 3.6 million barrels a day: i.e., 83 years from start to peak. Moreover, despite heroic efforts to the contrary, Texas oil production has obstinately declined since the peak. (API, 1994; Duncan, 1995a) In comparison, Mexican oil production started in 1901 and reached its greatest rate in 1994 at 3.3 million barrels per day: i.e., 93 years from start to present. Although details differ, in Sections 6 and 7, I show that Mexican oil production is now near its peak, and likely to fall even faster than that of Texas.

Figure 3, a column chart, compares Mexico's cumulative oil production with that of the United States, Saudi Arabia, and Texas. The height of each column represents the cumulative oil production for each nation from the start of production to year-end 1994, abbreviated "Q1994": United States = 191.4; Saudi Arabia = 66.2; Texas = 62.7; Mexico = 23.9 billion barrels. In percentages, the cumulative oil production of Mexico stands at 38% that of Texas, 36% that of Saudi Arabia, and a scant 12% that of the United States.

Figure xxx

Recall that Mexico's oil exports are collateral for the 1995 ESP loan. Mexico's potential for petroleum exports is determined by its "surplus" production. (Surplus oil is defined as total production minus domestic consumption.) Figure 4 plots Mexico's total oil production, domestic consumption, and its surplus for the period 1970 through 1994. The data shows that Mexico's oil surplus grew rapidly from 1974 to 1982, but decreased by 10.6% from 1984 through 1994 (i.e., from 1.795 to 1.605 million barrels per day, a decline in surplus of 190,000 barrels per day).

Figure xxx

The following section describes and develops the Hubbert simulation model.


Hubbert was for many years with the Shell Oil Company. In 1958 he predicted with uncanny accuracy the course of discovery and production of [United States] petroleum.
Philip H. Ableson, 1975

Several decades ago, distinguished geologist and geophysicist M. King Hubbert developed a technique for predicting the production life-cycle of nonrenewable energy resources, such as crude oil and natural gas (Hubbert, 1949; 1956; 1969). By design, King Hubbert wisely avoided the use of energy "reserves." However, his method is critically dependent on exploration and production data. The Hubbert model, as it is called, is based on a simple law of resource geology.

The complete cycle of production of any nonrenewable resource, in any one region or in the entire world, must begin with production rising from zero towards a maximum; there may be several maxima separated by temporary declines; but eventually the cycle is completed by a long-continued decline back to zero via the negative exponential. (North, 1985)

Petroleum expert F. K. North continues,

Hubbert's forecasts for the ultimate oil and gas production of the continental USA survived the passage of time so much better than other forecasts that his methodology was accorded almost magical capabilities by some. In fact, its utility is strictly limited, though if used within its limits the results to which it leads are crucially important. The method obviously makes no pretense of application to unexplored or little-explored areas which provide no historical data to extrapolate. It cannot be used with any reliability until the rate of additions to reserves (dQD/dt) has clearly entered the negative slope. ... In those regions, if in few others, use of the Hubbert technique is likely to yield trustworthy assessments. (ibid.)

[Note 4: In the equations and charts that follow, I have dropped the subscript "D" that appears in F. K. North's notation (above), because the subscript introduces a level of detail that is not needed in this discussion.]

The applicability of the Hubbert model to a specific oil-bearing region thus boils down to two factors: (1) the stage of development, and (2) the availability of historic data. No one describes the petroleum life-cycle better than geologist and oil field developer Rick Bass. First, the up-side of the curve.

Geologists, if they work on a new basin long enough, and if the basin holds enough oil and gas, get to see the basin "mature." ... What they mean is that there is now enough information on it, enough control - seismic lines shot, wells drilled - for them to have a very good handle on it. They mean the basin is in the peak of its years, exploration- and development-wise. ... When a basin is "young," there isn't enough information. ... Paradoxically, up to a certain point, the more oil and gas fields found, the easier it becomes to find others. You would think that every time one was discovered, it would make it that much harder to find another - that there'd be one less around. But before you reach a saturation point, it doesn't work that way at all. Because that's the single best way to discover oil: to figure out how it is trapped, under what conditions, nearby, and then look for another area that has similar conditions. ... There are a few golden years in the maturation of any successful basin where, truly, if the homework has been done and all the facts studied and learned, it is . . . easier. (Bass, 1989)

Next, petroleum geologist Rick Bass describes the down-side of the curve.

After a while, of course, the saturation point is reached; it gets harder to find new fields. Because finally a large percentage of the fields have been found. It is so like middle age that it is depressing. Basically, one day you just notice - though it may have been going on for quite some time before you admit it - that you are not finding oil wells as frequently, or with the success rate that you once were. It's not an immediate thing, but it's part of the phase too, part of the cycle. "Overmature" is the term for it. (ibid.)

The model consists of two equally important parts: the Hubbert equation, and the boundary conditions. The foregoing qualitative descriptions by Messrs. Bass and North are translated into the quantitative Hubbert model as follows.

[Note 5: I develop the Hubbert equation in two steps. The resulting Hubbert equation is given by Equation (2) below.]

I. The Hubbert Equation: During the young years of exploration in an oil-bearing region, the characteristic rate of discovery is exponential growth. In this young phase (see Bass above), oil production is approximated by the equation,


where dQ/dt is the rate of oil production, say, in millions of barrels per day, and Q[inf] is the cumulative production of oil, in billions of barrels. The term "a*Q" represents exponential growth. It is like compound interest paid on a bank account: "The rich get richer ... ." Or like a bunch of amorous rabbits multiplying their numbers. Taken alone, Equation (1) means that an oil-bearing region would remain eternally young and oil production would grow faster-and-faster, forever. Unreal. Impossible.

Inevitably, of course, "middle age" sets in (Bass again). The oil-bearing region becomes "mature." Worse yet, after "a few golden years," the region becomes "overmature." Depletion dominates. The process of slowdown and decline is modeled by adding a negative term, -b*Q2, to Equation (1). Doing this, we get,


Equation (2), above, is the Hubbert equation, but (I re-emphasize) it is only half of the Hubbert model.

II. The Boundary Conditions: The boundary conditions ("BCs") are the initial and final states of the system. They are ironclad, inescapable. They lock-in a unique solution. As stated above by F. K. North, energy production, dQ/dt, must be zero before production begins (i.e., at Q = 0) and eventually it must return to zero (i.e., at Q = Q[inf] .) Substituting the boundary conditions into Equation (2), we get,


Equation (3) is the Hubbert model. The "Energy Depletion Arch." Just as the positive term, a*Q, dominates on the up-slope, the negative term (or the "senility gene," so to speak), -(a/Q[inf] )*Q2, dominates on the down-slope, during the "overmature" or "old-age" phase of the oil production life-cycle.

[Note 6: Suggestion. Verify that the boundary conditions are satisfied by the Hubbert model: i.e., substitute separately Q = 0, and Q = Q[inf] into Equation (3) and confirm that dQ/dt equals zero in each case.]

A graph with Q on the x-axis and dQ/dt on the y-axis produces a parabolic arch, one like the majestic Gateway Arch in St. Louis. Mentally translating Equation (3) onto an x-y plot, imagine one "foot" of the arch planted at the origin (i.e., at Q = 0, dQ/dt = 0) and the other planted at Q = Q[inf], dQ/dt = 0. But because countless graphs of the Hubbert arch appear in the literature (e.g., Hubbert, 1982; Duncan, 1995a), I will not repeat it here. For now however, have a look at Figure 7, Section 6, where the "Energy Depletion Arch" for Mexico is graphed.

[Note 7: Boundary conditions are extremely important in modeling because (1) they eliminate most candidate models, and (2) they lock the model-of-choice into a specific solution. For a familiar example, consider the fine sport of baseball. When the batter hits a ball into the air, the fielder "knows," (1) that the trajectory will be a parabola, and (2) that the ball will come down. Woe be unto the fielder who chooses an ellipse or hyperbola. Getting back to energy, I leave it to you, the reader, to describe the model and boundary conditions that you (subconsciously) apply every time you fill your automobile's tank with petrol.]

[Note 8: The main goal of science is to predict, not explain. Thus, there is no need to have any philosophical worries about what the Hubbert model means, i.e., Equation (3). At the very least, the Hubbert model is a mathematical construction that is too fruitful to embellish or abandon.]

In the next Section, I determine the values of "a" and "Q[inf]" for the Mexican model.


The need for anticipating petroleum-industry development does not derive solely from the immediate function which that industry provides, but also from the depletion character of its operations. Oil and gas are wasting assets; that is, once placed in production, their outputs are subject to inevitable and sometimes rapid decline to exhaustion.
Antonio J. Bermúdez, 1963

Figure 5 below displays Mexican oil production on the y-axis, and cumulative oil depletion, Q, on the x-axis. Although the general profile of the data is similar to that in Figure 1, the change to cumulative oil depletion in Figure 5 emphasizes the causal relationship between oil production and resource depletion. The causality is simple. For every barrel of oil extracted from the earth, the amount of producible oil that remains underground (i.e., the reserve) is depleted by exactly the same amount. As the reserve goes to zero, production goes to zero, inexorably.

Figure xxx

[Note 9: Crude oil is not "produced" in the dictionary meaning that it is made or manufactured. Rather, it is extracted from the earth. Therefore, it would be more accurate to label the y-axis of the graphs, "Extraction, DQ/DT," and the x-axis "Cumulative Depletion, Q," as I have done in Figure 5. However, to avoid confusion, I will defer to the common practice and revert to (mis) labeling the y-axis "Production" and the x-axis "Cumulative Production."]

To apply the Hubbert model to Mexico, specific values for "a" and "Q[inf]" must be determined. The process is known as "parameterization." The method employs a statistical technique called linear regression. First, I divide Equation (3) through by Q to obtain the linear regression equation. The result is,


[Note 10: Equation (4) above is a linear equation. It is related to, but different from, the Hubbert model, as given previously by Equation (3). As such, Equation (4) is called an "auxiliary" equation. Bottom Line: It is the auxiliary equation that is fit to the data (e.g., Figure 6, below), not the Hubbert model.]

Second, I use Microsoft Excel¦ (an application program) to determine the values of "a" and "Q[inf]". Figure 6 displays the ratio of the Mexican annual oil production to the cumulative oil production, i.e., (DQ/DT)/Q data, on the y-axis, and the cumulative oil production, Q, on the x-axis. It also shows the best-fit regression line, i.e., (dQ/dt)/Q, Equation (4), superimposed on the historic data, for comparison.

Figure xxx

In Figure 6, the interception of the regression line and the y-axis estimates the value of "a" to be 0.133 per year. Similarly, the interception of the regression line and the x-axis estimates Q[inf] to be 35.4 billion barrels. Note also in Figure 6, that the historic data displays a definite downward trend and, of course, the regression line reflects that trend by its negative slope.

[Note 11: For comparison; Saudi Arabia Q-ultimate = 275 billion barrels, USA Q-ultimate = 220 billion barrels, Texas Q-ultimate = 64.8 billion barrels, and Mexico Q-ultimate = 35.4 billion barrels. (Duncan, 1995a; PE, 1995a)]

Looking again at Figure 6, the negative slope of the regression line, i.e., Equation (4), means simply that, "The faster Mexico extracts oil out of the ground, the sooner production goes to zero." Or in more picturesque language,

The fat and easy areas have been discovered. We, the world - not just the United States but the petroleum strongholds such as Mexico, the USSR, and the Middle East - are pulling out of the ground twice the amount we are finding. Each year. (Bass, 1989)

[Note 12: No data is omitted in the Hubbert method. Although the regression line in Figure 6 is fit to 1981-1994 data, the rest of the Mexican data (i.e., 1901-1980) is included in the parameter Ns of Equation (6), discussed later in Section 7.]

[Note 13: Figure 6 suggests that, in effect, Mexico's strategy is to stretch out production by reducing the negative slope of the regression line (e.g., by "conservation") - thereby maximizing the ultimate worth of their remaining oil.]

The ensuing section examines the future of the Mexican oil industry.


It is supposed that the foreign currency from petroleum exports would allow Mexico's financial problems to be overcome. ( This position is unsound from Mexico's standpoint. There is no doubt that Mexico is a country rich in petroleum, and that its subsoil contains great volumes of still undiscovered and unexploited oil and gas. But the country is not so rich in petroleum as some suppose - at least not rich enough to throw an abundance of oil on glutted world markets.
Antonio J. Bermúdez, 1963

The values of the parameters for the Hubbert model, as configured for Mexico previously in Section 5, are, "a" = 0.364 per day, and "Q[inf]" = 35.4 billion barrels. Substituting these values in Equation (3), gives,


[Note 14: Compared to the value of "a" given previously (i.e., in Section 5) ¥ In Equation (5) above, I have multiplied "a" by (1,000/365) to convert production from billions of barrels per year to millions of barrels per day.]

Equation (5) is the theoretical model for Mexican oil production. The resulting curve - a parabola - is plotted in Figure 7 (shaded circles). Selected years are highlighted: e.g., 1989 (the theoretical peak), 1994 (for comparison with the 1994 data point) , and 2000 and 2010 (i.e., the predicted production rates, 4 and 14 years hence). The historic data (X-shaded squares) is superimposed for comparison.

Figure xxx

Note in Figure 7 that the theoretical trajectory (shaded circles) shows a marked decrease in Mexican oil production from 1991 to 1994. But in contrast, the historic data points show a slight increase over the same period. This demonstrates that the model predicts only the general trend of production, not specific values. Sooner or later, however, Mexican oil production will start to fall. And inevitably it will go to zero. At least three factors will determine the steepness of the down-slope trajectory. Phrased as questions, they are:

One: When will Mexican oil production start to fall? In December 1994, while the beleaguered Mexican government faced bankruptcy, I analyzed the Mexican oil industry and predicted that production would peak in 1995 or 1996. On 31 January 1995, I summarized my findings in a letter to President Clinton.

My calculations show that Mexican oil production is now near its all-time peak and poised for a steep and obstinate fall, much like the decline (since 1972) in Texas oil production. In fact, I predict that Mexican oil production will start to fall sharply in less than two years, i.e., before January 31, 1997. (Duncan, 1995b)

My conclusions were (and are) based on (1) the unequivocal negative slope of the regression line (Figure 6, this report), (2) the fact that Mexican oil production has long since passed its theoretical peak (i.e., 1989, Figure 7), and (3) the theoretical production curve is already in sharp decline (as shown in Figures 7 and 9).

The President's response was disquietingly visionary.

My plan to aid Mexico's economy in the face of that nation's financial crisis, which I announced under executive authority, will avert further disruptions in world markets, stabilize the peso, renew confidence in the Mexican economy, and protect American jobs dependent upon exports. The plan will also reduce immigration pressure on the border.

The peso crisis presented us with a situation where decisive action by the United States could prevent a far greater crisis down the line. Recognizing the need for immediate action, we quickly put together a sound financial package that secures collateral from Mexico and places specific restrictions on Mexico's financial practices, which will safeguard U.S. interests. We also worked to gain support from global organizations, such as the International Monetary Fund and the Bank of International Settlements. Allies in the hemisphere, including Canada, Brazil, and Argentina, also made important contributions. Initial reaction of the markets to this U.S.-led effort has been positive.

Our nation's security depends on our continued world leadership -- and only the United States can take this leadership role. In moving decisively, we have acted in the world's interest and in our best interest and that of the millions of Americans whose livelihoods are tied to Mexico's well-being.

America's economic security is inextricably linked with Mexico's. By putting Mexico back on track, we have secured American jobs, preserved American exports, and safeguarded America's borders. (Bill Clinton, 27 February 1995, personal communication)

Two: How fast will Mexican oil production decline? The Hubbert model predicts that, for example, from 1996 to 2010 Mexican oil production will decline at average rate of 4.9% per year (as shown in Figure 7). Petroleum consultant C. J. Campbell describes the Mexican petroleum predicament as typical in the Western World.

Depletion rates for most countries outside the Middle East lie between 3% and 8%, with a few above 10%. The profile is primarily driven by the early large fields which are already in terminal decline and are not being replaced. (PE, 1995a)

Three: What will be the ultimate magnitude of Mexican oil production, Q[inf]? Recall that the value of Q[inf] is an estimate, but (and this is important) the estimate becomes ever more accurate as additional production data becomes available: i.e., as Qi -> Q[inf]. My calculations, and those of other analysts, show that Mexican production is fast approaching Q[inf]. For example, Table 1 gives two forecasts for the Mexican petroleum industry (i.e., Duncan, this study; and Campbell in PE, 1995a).

Table 1. The Down-Side of Mexican Oil: Two Estimates
[Values of Q are in billions of barrels.]
(this study)(PE, 1995a)
Peak Year (prediction)1996 1998
Q-ultimate (estimate) 35.4 52.0
Q end-94 (data source)23.9 21.2
Amount remaining 11.5 30.8
Percent produced 67.5%40.8%
Percent remaining 32.5%59.2%
Depletion rate (prediction)4.9%3.1%

Note especially the two bold rows in Table 1. Both studies predict that the peak of Mexican oil production is imminent, i.e., 1996 and 1998, respectively. Both likewise predict steep rates of decline, i.e., 4.9% and 3.1% per year, respectively. Because the down-slopes will be socially, politically, and economically devastating in either case, the difference of 1.8% does not alter the conclusions of this study.

[Note 15: Table 1 gives estimates of Mexico's Q-ultimate equal to 35.4 and 52.0 billion barrels. Of course there are other estimates, ranging from 48.5 all the way up to 700 billion barrels. (Grayson, 1980; Metz, 1978) Take your pick.]

In Figure 8, I have sketched a hypothetical trajectory for the down-side of Mexican oil production. Note that the trajectory is consistent with the historic data (Figure 1) because it starts at the 1994 data point (i.e., Q1994 = 23.9 billion barrels, and DQ/DT = 3.3 million barrels/day), and it ends at my estimate of Q[inf] = 35.4 billion barrels and dQ/dt = zero.

Figure xxx

Our main concern here is, Will the down-side of Mexican oil production be gentle or steep, a slope or a cliff? Although the fine details are not predictable, there are telling clues. If the range of estimates of Q-ultimate in Table 1 is correct (i.e., if 35.4 < Q < 52.0 billion barrels), then sometime between 1 January 1996 and 31 December 1998, Mexican oil production will, so to speak, tumble headlong over a sheer cliff.

It will be very interesting to see how we handle it. (Bass, 1989)

The following section shows the life-cycle of Mexican oil production, in years.


The rebirth of the Mexican petroleum industry concerns not only the destiny of Pemex, but of all Mexico, because if Pemex fails, the nation will fail.
Díaz Serrano, c. 1976
Director General, Pemex
(in Williams, 1979)
I believe the next 10 to 15 years are our last chance. Failure to provide young people with a stake in the system may have serious consequences for a regime whose commitment to "revolutionary" goals often seems more rhetorical than real. [If Mexico does not succeed] social tensions will grow difficult. There will be invasions of property, insecurity in the cities, new political leaders.
Victor Urquidi, c. 1979
President, Colegio de México
(in Grayson, 1980)

More than fifteen years have passed since the warnings of Messrs. Serrano and Urquidi. In preparation for such contingencies, the U.S. Immigration and Naturalization Service (INS) developed a scenario for field exercises held along the U.S.-Mexican border in 1995. It reads,

Suddenly, a vast flood of illegal immigrants - Mexicans driven to desperation by some unspeakable and unspecified social catastrophe - surges across the Southwest border, inundating entire communities as it washes north into the American heartland. (Dillon, 1995)

Fortunately, we are no longer dependent on scenarios and vague warnings. We can now use historic petroleum production data (Figure 1) and the time-dependent Hubbert model configured for Mexico [i.e., Equation (6), discussed below], to reliably predict the future of Pemex petroleum production, and, if Mr. Serrano is correct, the future of Mexico itself.

The Hubbert model, configured for Mexico in the Q-domain by Equation (5), can be transformed into the time-domain by separating the variables and integrating. The result is Q(t), the cumulative production of oil as a function of time, t. Then, by differentiating Q(t) once with respect to time, we get the equation for oil production, dQ/dt, in the time domain. The result is,


Equation (6) is the time-dependent counterpart of Equation (3). As before, dQ/dt is the production rate of oil, e.g., in millions of barrels per day. The values of "a" and "Q[inf]" are given in Section 6. Time "t" means "year minus regression start year" [i.e., t = (year - 1981)]. To simplify the notation, Ns is defined as (Q[inf] - Qs)/Q[inf], where Qs is the cumulative production at the regression start year (i.e., Q1981, Figure 6).

For those who wish, details of the Hubbert model are discussed in Hubbert (1971; 1981; 1982), Smith, et al. (1992), and Duncan (1994a; 1994b; 1994c; 1995a). What matters here is that Equation (6) represents the time-dependent life-cycle of Mexican oil production.

Figure 9 shows the theoretical life-cycle of Mexican oil production in the time-domain, i.e., a plot of Equation (6) for the years 1950 to 2030. The historic data is superimposed on the theoretical curve for comparison.

Figure xxx

Considering the consequences of collapse (e.g., "If Pemex fails, Mexico will fail," and "America's economic security is inextricably linked with Mexico's," and "No one can estimate how far the contagion might spread."), the down-slope of Mexican oil production is stark, to say the least. Equation (6), graphed in Figure 9, indicates that Mexican oil production is now at or near its all-time peak -- tottering at the brink of a precipitous cliff. Twelve years ago, this possibility (including what are now called "maquiladora" industries) was foreseen by political scientist Judith Gentleman.

Unfortunately, the past always constitutes a preface to the future, and in Mexico's case, it does not make for pleasant reading. ... Rather than constituting a harbinger of a new level of material well being for the mass of the population, the development of Mexico's petroleum resource signaled the opening of a new era of dependence for the nation. ... The opening of the floodgates for a new period of direct foreign investment fashioned from the demand profiles of foreign economies will mean the further elaboration in Mexico of a productive structure that neither provides a satisfactory level of employment nor produces goods that are compatible with the needs of the mass of population. (Gentleman, 1984)

Equally prescient, but decades earlier, natural resource expert Harrison Brown warned that the United States was slouching toward an ill-conceived, unworkable NAFTA-like involvement with Mexico. His afta' NAFTA scenario follows.

True federation between countries of high population growth potentials and those of low is extremely difficult to imagine if abolishment of economic barriers is to be an integral feature of federation. For example, a federation between the United States and Mexico at the present time could have disastrous consequences. Although the standards of living in Mexico would probably be improved by such a move, at least temporarily, the standard of living in the United States would almost certainly be lowered, and, far more important, the sudden incorporation in our society of a major group possessed of high [population] growth potential would lead to an accelerated rate of population increase, together with the numerous difficulties associated with an accelerated rate of increase. (Brown, 1954)

The ensuing section discusses the 1995 Emergency Stabilization Package (ESP) loan in the context of Mexico's plan for industrialization and its obligations to the NAFTA.


The United states can find an inexpensive solution to the problem now, or an expensive one later. No one can estimate how far the contagion might spread, or how many jobs might be lost. Mexican default is extremely unlikely.
Robert E. Rubin, 1995
U.S. Secretary of the Treasury
(in Sanger, 1995a; 1955b)
Unless decisive political leadership is shown now, this ancient Aztec nation could find itself in the mid-twenty first century with the same deformed economy, but it may have exhausted the abundant oil reserves with which to pursue the equalitarian ideals of the 1910 upheaval.
A North American Analyst
(in Williams, 1979)

The Republic of Mexico - historic, independent, and proud - faces several major problems simultaneously. The triad includes, (1) its ambitious plan for industrial development, (2) the requirements of the NAFTA, and (3) its external debt - with the focus here on the 1995 $50 billion Emergency Stabilization Package (ESP) loan.

1. Industrial Development: The most fundamental requirement for "sustainable" development (of Mexico, or any country) is that its economic infrastructure - i.e., agriculture, transportation, commerce, and industry - must be based on renewable and permanent primary sources of energy, e.g., photovoltaic and nuclear fusion power, should either become practical. (Duncan, 1993a; 1993b)

If Mexico is to avoid serious and substantial problems in the future, its government will soon, in this decade, have to take steps to restrain domestic energy demand, and to begin developing alternative energy sources. ... Mexico will need to develop substitutes for oil, especially in the generation of electricity. Currently, 90 percent of its electrical production relies on oil and gas. (Ronfeldt, 1980)

However, in the 16 years since the Ronfeldt document appeared, Mexico's dependence on nonrenewable energy has gone from bad to disaster. In fact, the Mexican economy is among the most unsustainable in the world. Mexico is now dependent on fossil fuels for more than ninety-six percent of its primary energy supply. Specifically, the tally for year-end 1994 is; oil = 71.3%, natural gas = 21.7%, and coal = 3.7% for a total fossil-fuel dependence of 96.8%. (BP, 1995) Therefore, if Mr. Ronfeldt is correct, Mexico's plan to use fossil fuels as the basis for industrial development is unrealistic. Unattainable.

2. NAFTA Requirements: As I understand them, the NAFTA requirements (both written and "understood") include,

[Note 16: Mexico's major environmental problems include: Untreated sewage and industrial effluents polluting rivers in urban areas; deforestation; widespread erosion; desertification; serious air pollution in the national capital and urban centers along the U.S.-Mexican border. (EIA, 1995)]

There is no doubt that each of the foregoing NAFTA requirements, taken alone, could be achieved to North American standards. But taken together, and given the present discordant state of the Mexican government and the distressed state of the economy, the reality is that the NAFTA requirements (even in part) are unachievable in the foreseeable future.

3. The ESP Loan: Theoretically, i.e., taken alone, out of context, Mexico could use the revenues from its oil exports to repay the full amount of the 1995 $50 billion Emergency Stabilization Package (ESP) loan. But that is not the problem.

The real problem facing Mexico, taken simultaneously, in context, is to repay the 1995 ESP loan, and meet the requirements of the 1993 NAFTA, and complete its plan for industrial development. To accomplish this, Mexico would have to quintuple (i.e., increase by five times) its oil exports over the next decade (i.e., by 2005, it would have to increase its oil exports from the present 1.6 million barrels per day to some 7.5 million barrels per day, roughly the 1995 export level of Saudi Arabia). Quintupling its oil exports is not feasible for several reasons: financial, political, engineering, geological, and moral.

3.1 Financial: To quintuple oil exports by 2005, Pemex would have to spend billions of dollars each year on oil exploration, development and capital infrastructure. But Pemex is now financially strapped. "Between 1995 and 2000, Pemex requires more than $20 billion in investment [just] to maintain current production." (Grayson, 1995) Moreover, it makes no sense for Pemex to invest a lot of up-front capital that, in a glutted market, would only reduce the price of its oil exports.

3.2 Political: Conceivably, Mexico could privatize its entire oil industry, and thereby obtain a large injection of foreign capital to repay the ESP loan. But at best, this would only be a temporary fix because the government would lose control of its Cash Cow, Pemex. Further, Article 27 of the Mexican Constitution forbids foreign ownership of crude oil reserves, and transportation and production facilities.

3.3 Engineering: Even if 100% financing were assured and engineering design started today, it would take more than 10 years to construct the physical infrastructure necessary to quintuple Mexican oil exports. But by that time, Mexican oil production will already be in abrupt decline. (Campbell, 1991; PE, 1995b; Duncan, this study)

3.4 Geological: Oil production at most Mexican wells is severely "rate limited." Thus, even if Mexico solved problems 1-3 above, its rate of oil production would not significantly increase, and the ultimate recovery, Q , would, in fact, decrease.

Oil and gas cannot be extracted efficiently from natural reservoirs excepting at relatively low rates. ...From 1911 to 1925 ... there were well fields in which salt water broke through the oil after only a few months of uncontrolled production. Had these fields been exploited with the proper technique and the application of proper conservation measures, they would have remained productive for anywhere from ten to fifty years. (Bermúdez, 1963)

3.5 Moral: David E. Sanger covers the moral issue.

If all goes well, American taxpayers could even profit from the [ESP] strategy, because Mexico is paying hefty fees for the loan guarantees. The United States is also insisting on first rights to Mexico's oil export earnings in case of default, an annual income flow of roughly $7 billion.

Many people say, however, that it would be impossible for the United States to demand those oil profits if Mexico falls into deeper crisis and its unemployment rate soars. (Sanger, 1995b)

The main goal of this study, you will recall, is to answer the question, Is the revenue from Mexico's petroleum exports safe collateral for the 1995 $50 billion Emergency Stabilization Package (ESP) loan? The answer is no. Specifically, if the Mexican government meets the requirements of the North American Free Trade Agreement (NAFTA), then it must default the Emergency Stabilization Package (ESP) loan, and vice versa.

However, if President Clinton (1995) and Secretary Rubin (in Sanger, 1995a; 1995b) are correct, then the ESP-NAFTA problem is much worse. Historic data shows that the remaining amount of Mexico's producible oil is small and rapidly depleting (Figures 6-9, Sections 5-7). This means that Mexico's present plan for industrial development, because it is uniquely dependent on petroleum, is bound to fail. [See Note 17.]

[Note 17: This does not mean that Mexico can never develop. It only means that Mexican development, should it occur, must rely on renewable and permanent sources of energy, not petroleum.]


Summary: On January 31, 1995 the President of United States signed a $50 billion Emergency Stabilization Package (ESP) loan to bailout the near bankrupt Mexican government. In case of default, the United States has first rights to revenues from Mexico's oil exports. This study uses a robust method to answer the question, Are Mexico's petroleum exports safe collateral for the 1995 $50 billion Emergency Stabilization Package (ESP) loan? (Section 1)

Mexican oil production grew rapidly from 1973 to 1982, but stagnated from 1982 through 1994 (Section 2, Figure 1). For comparison, Mexico's cumulative oil production through 1994 was only 38% that of the adjacent State of Texas, and Texas production is now in sharp decline (Section 3, Figures 2 and 3). Moreover, Mexico's oil surplus increased rapidly from 1974 to 1982, but decreased by 10.6% from 1984 through 1994 (Figure 4).

The Hubbert model (i.e., the method used in this study) is described and configured for Mexico (Sections 4 and 5, Figures 5 and 6). The ultimate production of Mexican oil is estimated to be 35.4 billion barrels, an amount that is only 55% that of Texas, and a minuscule 13% that of Saudi Arabia. Although for the past four years (through 1994), Mexican oil production has increased slightly, this study predicts that it will reach its all-time peak in 1995 or 1996, and thereafter decline at about 4.9% per year (Section 6, Figures 7 and 8, and Table 1).

The theoretical life-cycle of Mexican oil production is graphed for the years 1950-2030 (Section 7, Figure 9). The question of repayment of the 1995 ESP loan is discussed within the context of three major problems that simultaneously confront Mexico: industrial development, NAFTA requirements, and external debt (Section 8).

Conclusions: Mexico's petroleum production is now near its all-time peak. Decline is imminent. Mexican oil production will follow a down-slope trajectory similar to that of its U.S. neighbor, Texas. More to the point of this study, Mexico's oil exports are falling and, overall, they will continue to fall. As a result, the Mexican government will be forced to default or renege on, or abandon, one or more of its major commitments, respectively, (1) the 1995 $50 billion Emergency Stabilization Package (ESP) loan, or (2) the 1993 North American Free Trade Agreement (NAFTA), or (3) its plan for industrial development. The contagion will spread far, wide, and deep.


Copies of the paper, "Mexico's Petroleum Exports: Safe Collateral for a $50 Billion Loan," are available from the Institute on Energy and Man.

U.S. Mail Distribution: Hard copies are available. Send a regular letter to the Institute on Energy and Man; 3821 NE 45th, #37; Seattle, WA 98105. Be sure to include the title of the paper and your U.S. (or international) mailing address.

Internet Distribution: Digital copies are likewise available by either of two procedures. (1) File Transfer Protocol - Host:, User ID: anonymous, Password: , Directory (or e.g., Path): /pub/local/duncanrc. Or (2) Hypertext Transfer Protocol - at URL The information is contained in four files: [1] "MexOil_README," [2] "MexOil_COVERPAGE," [3] "MexOil_TEXT," and [4] "MexOil_GRAPHS." The text files are in Word[tm] 6.0 (for Macintosh) and the graphics file is in Excel[tm] 5.0.

© 1996, 1997

Richard C. Duncan Ph.D., Director
posted 1996 June 30
updated 1997 December 18