R E G U L A T I O N O F G A S M A R K E T I N G A C T I V I T I E S I N M E X I C O *

D a g o b e r t L . B r i t o Rice University

J u a n Rosellón Centro de Investigación y Docencia Económicas

Resumen: Estudiamos las implicaciones de vincular el precio del gas n a t u r a l en México a l del sur de Texas sobre la comercialización eficiente de l gas en el primero. Argumentamos que a Pemex se le debería p e r m i t i r firmar contratos spot o de futuros en la venta de gas. Sin embargo, el precio del gas debería ser siempre igual al precio netback basado en el Houston Ship Channel al momento de entrega. A Pemex no debería permitírsele descontar el precio del gas del precio netback de Houston, incluso si lo hace de una manera no discriminatoria. Esta metodología es transparente, fácil de llevar a la práctica y no el imina ninguna opción legítima de mercado para ninguna de las partes involucradas. Pemex o los consumidores de gas pueden usar el mercado de Houston para cubrirse de transacciones especulativas.

Abstract: We study the implications of l ink ing the Mexican natura l gas price to the Houston price on the efficient market ing of gas i n Mexico. We argue t h a t Pemex should be permit ted t o enter into spot contracts or future contracts to sell gas. However, the price of gas should always be the net back price based on the Houston Ship Channel at the t ime of delivery. Pemex should not be permit ted to discount the price of gas from the Houston netback price even in a nondiscriminatory fashion. This arrangement is transparent, i t is easy to enforce and does not eliminate any legitimate market options for any of the parties involved. Pemex or consumers of gas can use the Houston market for hedging speculative transactions.

J E L Classification: L51

Fecha de recepción: 23 IX 2001 Fecha de aceptación: 25 VI 2002

* The research reported i n this paper was supported by the Comisión Regula­dora de Energía i n a grant to the Centro de Investigación y Docencia Económicas, A . C . We wish to thank the observations of an anonymous referee, brito@rice.edu, juan.rosellon@cide.edu

15

16 ESTUDIOS ECONÓMICOS

1. I n t r o d u c t i o n

The question we are addressing is w h a t restrictions should be placed on Pemex's market ing activit ies i n the n a t u r a l gas market . To address this question, i t is useful to review what are probably well-accepted public- interest goals for regu la t i on . 1 These include the efficient a l ­locat ion of resources, achieving some redistr ibut ive goals, s impl i c i ty , and transparency. W i t h these goals i n m i n d , i t is clear t h a t the de­cision to l i n k the price of n a t u r a l gas i n Mexico to the price at the Houston ship channel by a netback rule solves some very di f f i cult technical and i n s t i t u t i o n a l problems i n a very simple fashion . 2 T h e netback rule l inks the price of gas at any po int i n Mexico to the price of gas i n Houston adjusted by the cost of t ranspor tat ion . T h e n a t u r a l gas market i n Mexico then has all the properties of the gas market at Houston. I n part i cu lar , al l agents are price takers w i t h respect to the market and the Houston market can be used by agents i n Mexico for hedging and other forward contracts. T h e key to the imp lementa t i on of this policy is t h a t there be sufficient pipeline capacity so t h a t the gas markets can clear and rents do not accrue to the pipelines. I f there is not sufficient pipeline capacity so t h a t the n a t u r a l gas markets i n Mexico can clear at the Houston netback prices, i t is impossible to implement the netback rule. A t the net back price, demand w i l l be greater t h a n supply.

A proposal that is being discussed is to change the system so t h a t Pemex sells gas only at the point of in j e c t i on . 3 T h e prices i n the local markets would be set by local supply and demand conditions. These changes would create uncerta inty i n the gas market and also create the possibi l i ty of strategic man ipu la t i on of the price of gas t h a t would be very di f f icult to regulate. Further , the current regulations p e r m i t the net back price to be an upper bound and Pemex can sell gas below t h a t price i f i t does so i n a nondiscr iminatory fashion.

T h e reason t h a t has been given for al lowing Pemex to sell at a price below the Houston netback price, as long as the sales were non-discr iminatory , is t h a t there is no reason to restr ict vo luntary transact ion between parties. However, there is a substantial agency

Polit ical economy goals of regulation are more general since interest group pressure could influence the design of regulatory inst itutions, regulatory frame­works, and industry structures (see Laffont (2000)).

2 See Br i t o and Rosellón (2002). 3 Future challenges of the Mexican natural gas reform are discussed in Comi­

sión Reguladora de Energía (2001).

GAS M A R K E T I N G ACTIVITIES I N MEXICO 17

problem i n these transactions. I t is hard to understand why Pemex (act ing as a agent for the Mex i can people) would want to sell gas i n Mexico for less t h a n i t could net by selling the gas i n Houston . There may be pol icy reasons to subsidize gas i n certain circumstances; however, this does not seem like a decision t h a t should be delegated to Pemex gas market ing .

Pemex should be p e r m i t t e d to enter into spot contracts or f u t u r e contracts to sell gas. However, the price of gas should always be the net back price based on the Houston Ship Channel at the time of delivery, Pemex should not be p e r m i t t e d to discount the price of gas f rom the Houston netback price even i n a nondiscr iminatory fashion. T h i s arrangement is transparent, i t is easy to enforce and does not e l iminate any leg i t imate market options for any of the part ies involved. Pemex or consumers of gas can use the Houston market for hedging of speculative transactions.

The Houston market can also serve as a buffer for fluctuations i n Pemex's product ion or i n demand. Pemex can vary its sales i n the Houston market to smooth fluctuations i n Mex i co . 4 This buffer allows Pemex to only sell " p l a i n vani l la" gas w i t h o u t having to engage i n complex market operations in Mexico. Thus , i t is very di f f i cult to see what useful role can be played by Pemex act ing as a gas marketer i n Mexico. I f Pemex wants to engage i n speculative market behavior, i t can do so i n the Houston market . Houston has the advantage of being a well-developed market . Pemex's transactions i n that market would not create any regulatory issues for the Comisión Reguladora de Energía, CRJE, as long as Pemex sells gas i n Mexico-at the Houston spot netback price. As long as there is sufficient pipeline capacity so t h a t there are no bottlenecks in t ranspor t ing gas, this simple rule w i l l result in an efficient and transparent na tura l gas market in Mexico .

A l l o w i n g Pemex discretion in pr i c ing gas becomes an even more complicated problem i f Pemex is allowed to sell gas for future delivery at a price other t h a n the Houston netback price at the t i m e of delivery. For example, Pemex can sell gas for delivery 30 days i n the future at a given price and the next day sell gas for delivery 29 days i n the future a different price. Technically, these transactions would not be discr iminatory . Transactions t h a t involve selling forward gas at a predetermined price would be very dif f icult to moni tor and give Pemex gas marketers a very large amount of power and discretion.

There are i m p o r t a n t and legit imate reasons why private oi l com­panies use forward markets to reduce risk; let us grant t h a t such

This assumes that Pemex is exporting gas.

18 ESTUDIOS ECONÓMICOS

reasons may also apply a nat ional o i l company such as Pemex. Re ­s t r i c t i n g Pemex to sell gas i n Mexico at the Houston spot m a r k e t netback price does not e l iminate any options for Pemex. L i n k i n g t h e price of gas i n Mexico to the Houston market permits Pemex to oper­ate i n these sophisticated markets w i t h out invo lv ing their customers for gas delivered i n Mexico . Further , buyers of n a t u r a l gas i n M e x ­ico can enter into transactions i n Houston w i t h o u t invo lv ing Pemex. Thus , there is no economic reason why Pemex has to operate as a gas marketer i n Mexico.

I t may seem more efficient to permi t Pemex to enter in to such transactions d irect ly i n Mexico w i t h o u t going t h r o u g h the Houston market . However, due to the vert ica l integrat ion of Pemex i n the gas industry , restr i c t ing Pemex to make such transactions only i n the wel l developed Houston market reduces the possibil ity t h a t Pemex could to set entry barriers to other part ic ipants in the gas commercial izat ion business, reduces the regulatory burden i n Mexico, and permits t h e development of proper market ins t i tut ions i n Mexico for futures and forward contracts.

2. P r o b l e m s w i t h F l e x i b i l i t y i n t h e N e t b a c k R u l e

The present regulations p e r m i t Pemex to sell gas at below the Hous­t o n netback price as long as i t does so i n a nondiscr iminatory fashion. Th i s policy is supported by the received wisdom i n regulatory the ­ory t h a t holds t h a t p r o h i b i t i n g a vo luntary transact ion on the p a r t of two competent parties does not improve wel fare . 5 However, th is result does not apply in this case. The linkage to Houston means t h a t all parties i n the Mexican market are price takers. Since Mexican gas can always be sold i n Houston, the value of the marginal cubic foot of gas at the well i n Mexico is the Houston price less cost of t ranspor t . We w i l l demonstrate t h a t a policy to sell at the Houston netback price is Kaldor -Hicks superior to a policy that discounts the price of gas i n a nondiscr iminatory fashion. 6

Suppose the regulator forces the firm to charge prices P . Total welfare would be V(P°) + a7r(P°), where V is consumer surplus, 7T is profits and a i n [0,1]. I f the f i rm is allowed to offer P such that V (P) > V(P°)t this alternative policy would not make consumers worse off and the f i rm would make a greater profit . (See Armstrong, Cowan, and Vickers (1994), p.67).

6 Under the Kaldor-Hicks test, state A is preferred to state B i f those who gain f rom the move to A can hypothetically compensate those who lose and yet

GAS M A R K E T I N G ACTIVITIES IN MEXICO 19

F i g u r e 1

I t is Kaldor -Hicks superior to have the price of gas i n Mexico equal to the Houston price adjusted for t ranspor tat i on costs. Assume t h a t gas is produced at Burgos and shipped to Houston and M o n ­terrey . 7 Let PH be the spot price at Houston and PM be the spot price at Monterrey. Let Ch be the cost of moving gas f rom Burgos to Houston, c m be the cost of mov ing gas from Burgos to Monterrey. T h e netback rule would lead to the condit ion t h a t the price of gas less t ransport cost is the same at Houston and Monterrey,

Pm - Cm = Ph - Ch (1)

Suppose a customer i n Monterrey had a demand curve Qi = Di{p) for the gas. Pemex can sell the gas to the customer i n Monterrey

be better off. The Kaldor-Hicks criterion suggests that A is preferable even i f compensation does not actually occurs.

n The gas fields of Burgos are located in the northeast of Mexico, close to

the Texas border. I n recent years over 44 t r i l l i o n cubic feet of non-associated gas have been discovered in Burgos (over 35 years of reserves). Burgos' reserves represent 57.1 percent of t o ta l natura l gas reserves in Mexico but contribute only 17.3 percent of t o ta l natura l gas production.

20 ESTUDIOS ECONÓMICOS

or sell the gas i n Houston. Suppose Pemex sold the consumer Qi amount of gas at p < Pm. I t is feasible for Pemex to sell the gas i n Houston and transfer an amount ApQi + AP^Q to the M o n t e r r e y consumer. (See figure l ) . 8 Th i s would lead to greater revenue, , to Pemex and make the Monterrey consumer no worse off. T h u s i t is Kaldor -Hicks superior to have the price of gas i n Mexico equal to t h e Houston netback price and sell the balance of the gas on the Houston market rather t h a n to sell gas i n Mexico at a price below the Houston netback price.

3. R e g u l a t i o n of P i p e l i n e R a t e s

Pipelines have a h igh fixed cost, and for a substantial p o r t i o n of the i r operat ing region low marginal costs. T h e capacity of the pipel ine is u l t i m a t e l y l i m i t e d by the pressure l i m i t s of pipe. Figure 2 i l lustrates the cost curves for a 48-inch pipeline 100 miles l o n g . 9 T h e dashed lines represent w h a t the cost curves wou ld be i f the pressure l i m i t s were not b ind ing . A t a pressure l i m i t of 1,500 pounds per square inch, the pipel ine reaches its l i m i t at approximately 3,800 m i l l i o n cubic feet per day. A t this po int i t becomes impossible to increase t h r o u g h p u t by increasing power and i t becomes necessary to add compressor, s ta ­tions whi ch increases throughput w i t h o u t exceeding the l ine l i m i t by increasing the pressure gradient.

We have shown t h a t the netback-pr ic ing rule is. the so lut ion of a static welfare o p t imiza t i on problem i f the fee for t r a n s p o r t i n g gas is the marg ina l cost of t ranspor t ing gas . 1 0 However, marginal-cost pr i c ing results in a loss of rents. (See figure 2). One so lut ion to th is problem is to set a fee t h a t yields a regulated rate of r e t u r n over the life of the project sufficient to cover al l costs. A n alternative , more sophisticated alternative is a two-part ta r i f f w i t h a price cap. C R E current ly regulates Pemex t ranspor ta t i on (and d i s t r ibut i on ) tarif fs t h r o u g h a (average-revenue) price cap over two-part t a r i f f s . 1 1 T h e

8 Recall t h a t under the net.back rule, the revenue after transportat ion costs of selling gas in Houston or Monterrey would be the same.

9 The parameters used in constructing this example are based on numbers reported in the O i l & Gas Journal, November 27, 1995.

1 0 . See Br i t o and Rosellon (2002), and Br i to , L i t t l e j o h n and Rosellon (2000). 1 1 Pemex estimates its fixed, variable and financial t ransportat ion costs ( i n ­

cluding an 11.5 percent rate of return) and sets its two-part tari f f according to i ts revenue requirements.

GAS M A R K E T I N G ACTIVITIES I N MEXICO 21

cap prevails d u r i n g each five-year period. The i n i t i a l value of the cap is set i n each per iod t h r o u g h cost of service and adjusted by inf la ­t i o n , efficiency, pass t h r o u g h and correction factors. Average revenue is calculated as the rat i o of t o t a l revenue to o u t p u t i n the current period. Ramirez and Rosellon (2002) show t h a t this regime creates a stochastic effect t h a t implies higher levels of consumer surplus for higher levels of r isk aversion and u n c e r t a i n t y . 1 2

F i g u r e 2

More generally, the economics l i terature on gas (and electric ity) t r a n s p o r t a t i o n shows how usage congestion charges can be used to

We must point out that when only one product is supplied, as in the trans­por tat ion service, average-revenue regulation coincides w i t h tariff-basket regula­t ion . For the case of gas d istr ibut ion , Ramirez and Rosellon (2002) also show that the CRE's average-re venue regime creates incentives for setting two-part tariffs strategically. The usage charge is typically dropped to its lowest feasible level while the fixed charge can be raised to compensate for the loss of operating profit .

22 ESTUDIOS ECONÓMICOS

give proper incentives for capacity inves tments . 1 3 However, there are two caveats to the use of a flexible pr i c ing mechanism to regulate Pe-mex t r a n s p o r t a t i o n tariffs t h a t seem to make preferable a f ixed-price regulat ion t h a t does not allow Pemex to carry out price d i s c o u n t s . 1 4

T h e first problem is Pemex's vert ical integrat ion i n gas p r o d u c t i o n , t ranspor ta t i on , and market ing , which allows Pemex to carry out cross subsidies among these three economic activit ies. A second prob lem is the l i m i t e d i n s t i t u t i o n a l capacity of a smal l regulator to ob ta in t r u e cost i n f o r m a t i o n on al l segments of the more t h a n 9,000-kilometer-long Pemex t ranspor ta t i on network. Under these conditions, Pemex could strategically manipulate pipeline rates to maximize its revenues but reduce consumer surplus.

A s an example of the la t ter assertion, suppose Pemex is p r o ­duc ing gas i n Burgos and Ciudad P^mex, and selling gas i n H o u s t o n and Mexico C i t y (see figure 3. The arrows indicate which way gas is moving) . Los Ramones is the arbitrage po int . Assume t h a t ti is the " rea l " (cost-reflective) price per mile for t ranspor t ing gas t h r o u g h the pipeline segment Burgos-Los Ramones and t h r o u g h the C i u d a d Pemex-Mexico C i t y segment. Let ¿2 be the corresponding price for the Houston-Burgos segment and the Los Ramones-Ciudad Pemex segment (ti < ¿ 2 ) - 1 5 The dashed line i n figure 4 i l lustrates the regu­lated price p a t t e r n that would result under perfect in f o rmat i on .

Suppose however t h a t the regulator does not have i n f o r m a t i o n on the real cost i n each pipeline segment, and t h a t Pemex can set a price for t ranspor t ing gas through the pipeline network i n the range between t\ and ¿2 per mile. Pemex can then exploit its f l ex ib i l i ty to set the pipeline tariffs to increase revenues. Pemex can charge the

Vogelsang (2001) proves how a two-part tari f f can be used to solve short -run congestion problems as well as the long-run capacity expansion problems of an electricity transmission network. Under capacity congestion, the variable charge becomes a pure congestion charge and, whenever congestion charges are greater than the marginal costs of increasing capacity, the transmission company wi l l have incentives to expand capacity.

1 ^ This is not equivalent to the use of cost-of-service regulation. Rather, we propose to keep calculating the in i t ia l value of the (average revenue) cap i n each regulatory period through cost of service and adjust i t along the period by inf lat ion and efficiency factors, but wi thout allowing price discounts.

1 5 Then, according to the netback pricing rule, the price of gas at Mexico C i t y wi l l be equal to the benchmark price in Houston less the transport costs f rom Houston to Burgos, plus the transport costs from Burgos to Los Ramones, less the transport costs from Los Ramones to Ciudad Pemex, plus the transport costs from Ciudad Pemex to Mexico City.

GAS M A R K E T I N G ACTIVITIES IN MEXICO 23

F i g u r e 3

Mexico City

Houston

Burgos

Los Ramones

Ciuäd Pemex

low t ranspor t charge t\ between Houston - Burgos, the h igh t ranspor t charge ¿2 between Burgos - Los Ramones, the low transport charge t\ between Los Ramones - C iudad Pemex and the high transport charge ¿2 between Ciudad Pemex - Mexico City . This is i l lus trated by the solid l ine in figure 4. This pr i c ing policy maximizes the revenues for Pemex by cross-subsidizing its pipeline segments. The result is a higher price of gas i n Mexico as compared to the one that wou ld prevail i f Pemex charged the real t ransport charges per segment.

T h e CRE then needs Pemex to provide accurate i n f o r m a t i o n on its t ransport costs by segment. Under the vert ical ly integrated struc­ture of Pemex, wh i ch allows cross-subsidization among gas produc­t i on , t r a n s p o r t a t i o n and market ing , the regulator should implement a fix-price regulat ion by t ranspor ta t i on region so t h a t Pemex cannot make discounts i n t ranspor ta t i on charges. Of course, the (f irst) best solut ion would be to vert ical ly separate Pemex - a l l o w i n g u n b u n d l i n g and compet i t i on i n m a r k e t i n g - and to regulate t ranspor ta t i on charges w i t h the incentive scheme already i n place.

4. P e m e x S e l l i n g G a s O n l y at t h e P o i n t of I n j e c t i o n

One advantage of using the netback rule w i t h a fixed fee for trans­p o r t i n g gas is t h a t all parties act as price takers at all points along

24 ESTUDIOS ECONÓMICOS

F i g u r e 4

Price

/ •

y

— 4 > 4 y

Houston Bums Los Ramones i n i M [ ^ í a

6 Pemex City

the pipeline. Restr i c t ing Pemex to sell gas only at the point of i n ­j e c t i on and al lowing local market conditions to set the price creates the possibi l i ty t h a t marketers could acquire some degree of market power. Parties could buy the gas at the po int of inject ion and ship either to the Houston market (where they face an essentially flat de­mand curve) or to the Mexican markets where they face an inelastic demand curve. Collusive behavior on the part of marketers w o u l d allow them to equate marginal revenue i n b o t h markets and explo i t the fact t h a t demand curves i n the local markets are very inelastic and earn monopoly rents. I t then becomes necessary to regulate t h e activit ies of the marketers. The regulatory problem is much simpler i f Pemex sells at all points on the pipeline system Using the netback rule to determine the price. This would not el iminate other marketers ' act ivit ies .

5, F o r w a r d M a r k e t s a n d P i p e l i n e C a p a c i t y

I f Pemex is required to sell gas on the spot market at the Houston Ship Channel price adjusted by the netback rule, can Pemex use

GAS M A R K E T I N G ACTIVITIES I N MEXICO 25

its monopoly power over the pipeline to get monopoly rents i n this forward market? To address this question let us consider a simple model. Assume a two per iod model. Gas is produced at Burgos and shipped to Houston and Monterrey. Let poh be the spot price at Houston at t ime 0, pom the spot price at Monterrey at t i m e 0, p\h the spot price at Houston at t i m e 1, p i m the spot price at Monterrey at t ime 1, fth the forward price at Houston at t ime 0, and pm the forward price at Monterrey at t i m e 0. Let be the cost of mov ing gas from Burgos t o Houston, crn be the cost of moving gas f rom Burgos to Monterrey, and A c = c m — c .̂ Let Qm be the capacity constraint on the pipeline f rom Burgos to Monterrey. I f the capacity constraint does not b i n d , the netback price at Monterrey is prrn — prh + A c , (see figure 6 l e f t ) . I f the capacity constraint binds, the price at Monterrey is Vrm — Prh + A c -h where R are the rents associated w i t h the capacity constraint , (see figure 6 r i g h t ) .

F i g u r e 5

• < • : > ^ M B H

Figure 6

& ft I f the pipeline capacity does not b ind , anyone who desires to engage i n f orward transactions can do so i n the Houston market and Pemex

26 ESTUDIOS ECONÓMICOS

does not have an effective monopoly of the forward market and w i l l capture no rents. However, i f the pipeline capacity does b i n d , Pemex can capture the rents associated w i t h the pipeline constraint by sel l ing o u t p u t forward. Pemex can become a monopoly i n the f o rward f i r m -service market . Note t h a t i f the pipeline capacity does b i n d , rents w i l l exist and the only question is who w i l l appropriate t h e m . G iven t h a t the capacity constraint on the pipeline is b ind ing , there are no real effects.

T h e key regulatory issue i n this context appears to be insur ing t h a t Pemex invests sufficiently in pipeline capacity so that capacity constraints are not a serious issue.

6. O p t i m a l P i p e l i n e C a p a c i t y

A necessary element i n implement ing a policy where the Houston gas market is the reference po int for pr i c ing gas i n Mexico is t h a t there be sufficient capacity so that the market for gas can always clear at t h e netback price. T h e obvious question is whether the cost of m a i n t a i n ­i n g such capacity is warranted . Th is is a very dif f icult question i n t h a t there are economic, po l i t i ca l and i n s t i t u t i o n a l constraints involved i n the basic question of pr i c ing gas along the Mexican pipelines.

A benchmark for discussion is the p a t t e r n of investment t h a t wou ld be followed by a planner who is a t t e m p t i n g to maximize a measure of welfare. Such a policy may involve periods where the capacity constraint binds. The length of this period is a measure of the cost of the deviat ion f rom " o p t i m a l " t h a t results f rom the pol icy of using the Houston gas market as a benchmark for pr i c ing gas. W e w i l l show t h a t a policy that insures sufficient pipeline capacity so t h a t the gas market can clear at the Houston netback price deviates f rom an " o p t i m a l " pol icy by only a matter of weeks.

Let us consider a case where pipeline capacity is given by Q: Demand is growing at a rate A. Let PM be the price i n Monterrey based on Houston netback. Assume t h a t demand reaches pipel ine capacity at t = 0 so t h a t PM — PM for Q < Q and PM = &(Q)-I f the pipeline capacity binds, PM = 9(Qext and the excess burden associated w i t h this bottleneck is given by:

v _ A p A Q _ Q(ext - \)[e{Qext-pM] f 0 .

This is the tr iangle a-b-c i n figure 7. The bottleneck results i n rents being generated and these rents result in the loss given by

GAS M A R K E T I N G ACTIVITIES I N MEXICO 27

X2(t) = 7 l Q A p = nQ[9(QeXt) - pM] (3)

where ji is a parameter t h a t can range f rom 0 to 1, and is the weight of the rents generated by the bottleneck in the welfare funct ion . The loss i n equation (3) represents the fract ion of rents t h a t are consumed i n transfer and reflects such factors as rent seeking and X-inefficiency. This is the rectangle PM — PM — a — 6 i n figure 7. Define the t o t a l loss in welfare as:

X(t) = X1(t)+X2(t). (4)

F i g u r e 7

P

i ! t u m

Q Qe"

Opening a second pipel ine reduces the marg inal costs of t ranspor t ing gas mov ing the operat ing range of b o t h pipelines to ^ . W i t h capacity constraints b ind ing , the marginal cost of t ranspor ta t i on depends on the construct ion of the new pipeline. The marg inal cost of mov ing gas w i l l then be reduced by AMC. This w i l l reduce the cost of moving

28 ESTUDIOS ECONÓMICOS

gas by (p — AMCQ, Let 72 be the weight of the cost savings of open­ing a second pipel ine i n the welfare funct ion . A welfare m a x i m i z i n g planner would want to pick the t ime of opening the second p ipe l ine to min imize the welfare loss less the savings in operat ing costs w h i c h is given by:

Q(ext-l)[0(Qext)-pM} 2 (5)

(6)

+liQ[e(QeXt) - PM] - 72¥>jdt + e~rTCo

T h e first order condit ion for this max imiza t i on is

• Q ( e * T - l ) [ f l ( 6 e * r ) - p M ] 2

+liQ [6(QeXT) - PM] ~ 72<P } - rC0 = 0

which can be w r i t t e n as

^ - " ' f ^ 1 + 7 i 3 [ g ( g e * r ) - Pm] ~ 7 2 ^ Co — = r ( ? )

We construct a numerical example to calculate the value of T for those values of the parameters and get a rough approx imat ion of t h e length of the period where i t is efficient for the capacity constraint to b i n d . Assume, as before, that a 48-inch pipeline reaches its l i m i t capacity at 3,800 m i l l i o n cubic feet per day when i t is 100 miles long.

Figure 8 i l lustrates the solution of the min imiza t i on problem for a 48-inch pipeline, 300 miles long. The curve labeled 71 = 1, 72 = 1 depicts the loss to the consumers. I f we examine the curve we see t h a t even for a very h igh rate of r e t u r n on the order of 30 per cent, the " o p t i m a l " investment t ime is about two weeks after the capacity constraint begins to b ind . For a rate of r e t u r n of 15 percent, the consumers w i l l never want the capacity constraint to b ind . Consumers of n a t u r a l gas are w i l l i n g to pay for the facilities to t ransport t h e gas they demand at the Houston netback price. Thus i t is feasible to construct a rate structure t h a t w i l l compensate the operator of the pipeline for ma inta in ing sufficient capacity t o t ransport the gas demanded at the Houston netback price. Note that such a policy is Pareto superior.

GAS M A R K E T I N G ACTIVITIES I N MEXICO 29

F i g u r e 8

o io 20 36 To So Week

The curve labeled 71 = 0.1, 72 = 1 depicts the welfare loss i f we assume that 10 percent of the rents transferred to Pemex are lost to X-inefnciencies. For a rate of r e t u r n of 30 percent, the " o p t i m a l " period for the capacity constraint to b ind is 15 weeks. For a rate of r e t u r n of 15 percent, i t is not o p t i m a l for the capacity constraint to b ind . T h e savings i n operat ing costs are sufficient to warrant the investment.

The curve labeled 71 = 0, 72 = 1 ignores the transfers f rom consumers and includes the savings i n operat ing costs and the dead­weight loss. The curve labeled 71 = 0, 72 = 0 ignores everything but the deadweight loss. Even using this measure of welfare loss the o p t i m a l per iod for the constraint to b ind is less t h a n one year for a rate of r e t u r n of 15 percent.

The weather i n Mexico does not f luctuate as much as i n the U n i t e d States, however, peaks that could result in seasonal bo t t l e ­necks do occur. Assume the bottleneck starts at t = T\ and ends at t = T 2 , (see figure 9). The welfare loss associated w i t h such a

T 2

bottleneck is then J X(t)dt. I t pays to invest i n addi t ional pipel ine

capacity to e l iminate the bottleneck if:

30 ESTUDIOS ECONÓMICOS

T 2

fX(t)dt-j2f - , * = * ^ - r Q - ^ > r , (8)

Co Co

where X is the average of welfare loss, (see figure 10).

F i g u r e 9

0 10 ft 20 30ft 40 50 Weeks

Let A T = T 2 — T i , figure 11 depicts the relationship between A T and X for r=0 .15 .

Consumers of gas are w i l l i n g to pay to el iminate a five-day peak whose average is 10 percent over capacity. A planner t h a t assigns a 10 percent cost to transfers w i l l invest to el iminate a 35-day peak whose average is 10 percent over capacity.

T h e need for concern about the possibil ity of capacity constraints i n gas pipelines follows f rom projections about demand. Demand for n a t u r a l gas i n the Pemex t ranspor ta t i on system w i l l grow at an annual rate of 11.0%. These estimates are based on increases i n the demand for n a t u r a l gas of electricity generation, industr ia l consumers, and

GAS M A R K E T I N G ACTIVITIES I N MEXICO 31

LDC's (see table 1). The northeastern and northwestern regions w i l l register a g rowth of 12% and 18%, respectively, d u r i n g the 1999 - 2003 period due to the CFE's projects. ( C F E is the nat ional electric ity monopoly.) These two regions w i l l represent 36% of t o t a l market demand.

F i g u r e 10

I n 1999, demand and supply for natura l gas in Mexico w i l l be 4,824 and 4,838 m i l l i o n cubic feet per day (mcfd) , respectively, i n 2000-2001 5,096 mcfd and 5,111 mcfd, respectively, and i n 2002-2003 5,259 mcfd and 5,275 mcfd, respectively. Accord ing to the permi t granted by C R E to Pemex to t ransport n a t u r a l gas , 1 6 Pemex w i l l face this increase i n demand by expanding its t ranspor tat i on capacity, (see table 2 ) . 1 7

Comisión Reguladora de Energía (1999). This permit states all the technical details and investment plans that Pemex wi l l have to ful f i l l during the next five years i n its transportat ion activities.

17 These calculations are based on estimates of injection and extraction require­

ments at each node (Comisión Reguladora de Energía (1999), appendix 3.1), flow

32 ESTUDIOS ECONÓMICOS

F i g u r e 11

As shown, the increase i n pipeline capacity w i l l barely cope w i t h the increase i n demand, and there could be bottlenecks d u r i n g peak p e r i ­ods. Especially i m p o r t a n t is the 1597 ki lometer- long pipeline system i n the Reynosa and Monterrey operat ing sectors where a huge i n ­crease i n demand is expected and where two of the three compression stations are o l d . 1 8

A very s trong case can be made f rom these calculations t h a t a policy t h a t makes sure t h a t there is always sufficient pipeline capacity so t h a t the gas market can always clear should be followed. Such a pol icy would generate sufficient savings to the consumers of gas so t h a t they w i l l be w i l l i n g to pay for such an investment.

and capacity technical information for each transportation sector (annex 3, ap­pendix 3.1 and 3.2), repowering needs at each compression station (appendix 3.1), and investment needs for expansion of the pipeline network (annex 6.2.1).

18 There are three compression stations located in these sectors. I n the Mon­

terrey sector here are two old "reciprocate" compression stations Ojo Caliente, and Santa Catarina, w i t h more than 30 years of operation, and w i t h huge drops i n pressure and low volumes. I n the Reynosa sector there is a " turbo compression" stat ion" that was constructed i n 1997.

GAS M A R K E T I N G ACTIVITIES I N MEXICO 33

T h e only argument t h a t can be made against investing i n this pipeline capacity is t h a t the government loses the revenue created by rents to the pipeline. However, the Mexican government can at present capture the rents t h a t would be generated by pipel ine con­gestion by tax ing gas. I f we take as given t h a t addit ional t a x a t i o n of n a t u r a l gas is not desired, then a pipeline investment policy t h a t pre­vents pipeline congestion can be Pareto superior. Consumers w o u l d be w i l l i n g to pay for this capacity and the only cost to the government is not col lecting rents i t can now collect and has chosen not to do so.

T a b l e 1 Natural Gas Demand: Annual Growth Rates

by Consumer Type

1994 - 1997 1997 - 2003

CFE 7 17 Industrial 5 5 Cogeneration — 76 Pemex 1 5 Vehicles — 51 Distribution 1 13

Source: Escenarios de oferta y demanda en el sistema nacional de gasoductos de Pemex-Gas, Comisión Regula­dora de Energía (1999).

T a b l e 2 Maximum Average Transport Capacity of

Pemex's National Pipeline System

Units Year 1 Year 2 Year 3 Year 4 Year 5

MMGcal/Year 421.5 445.3 445.3 459.5 459.5

MMPCD 4,824 5,096 5,096 5,259 5,259

Source: Comisión Reguladora de Energía (1999).

34 ESTUDIOS ECONÓMICOS

C o n c l u s i o n s

Pemex should be p e r m i t t e d to enter into spot contracts or f u t u r e contracts to sell gas. However, the price of gas should always be the net back price based on the Houston Ship Channel at the t i m e of delivery. Pemex should not be p e r m i t t e d to discount the price of gas from the Houston netback price even i n a nond iscr iminatory fashion. T h i s arrangement is transparent, i t is easy to enforce and does not e l iminate any legi t imate market options for any of the parties involved. Pemex or consumers of gas can use the Houston market for hedging speculative transactions.

The Houston market thus serves as a buffer for f luctuat ions i n Pemex's product i on or i n demand. Pemex can vary its sales i n the Houston market to smooth f luc tuat ion i n Mexico. This buffer allows Pemex to only sell " p l a i n vani l la" gas w i t h o u t having to engage i n complex market operations in Mexico. Thus , i t is very di f f icult to see w h a t useful role can be played by Pemex act ing as a gas marketer i n Mexico. I f Pemex wants to engage i n speculative market behavior, i t can do so in the Houston market . Houston has the advantage of being a well-developed market . Pemex's transactions i n t h a t market w o u l d not create any regulatory issues for the C R E as long as Pemex sells gas in Mexico at the Houston spot netback price.

The key to this policy is t h a t there be sufficient investment i n pipeline capacity so that bottlenecks do not develop. A very strong case can be made from our calculations t h a t a policy t h a t makes sure t h a t there is always sufficient pipel ine capacity should be followed. Such a policy would generate sufficient savings to the consumers of gas so t h a t they w i l l be w i l l i n g to pay for such capacity investment i n the rate s tructure . The only argument t h a t can be made against invest ing i n this pipeline capacity is the loss of revenue created by rents to the pipeline. However, the Mexican government can at present capture these rents and does not do so. I f this is the correct policy, then a pipeline investment policy t h a t prevents pipeline congestion can be Pareto superior.

R e f e r e n c e s

Adelman, M . A . (1963). The Supply and Price of Natural Gas, Oxford, Blackwell Publishers.

Armstrong , M . , S. Cowan and V . Vickers (1994). Regulatory Reform, Economic Analysis and British Experience, Cambridge, M I T Press.

GAS M A R K E T I N G ACTIVITIES I N MEXICO 35

Br i to , D. L . and J. Rosellón (2002). "Pricing Natura l Gas in Mexico: A n A p p l i ­cation of the L i t t l e Mirrlees Rule", The Energy Journal, 23, 3.

Br i to , D. L . , W. L . L i t t l e j o h n and J. Rosellón (2000). "Pric ing L i q u i d Petroleum Gas i n Mexico", Southern Economic Journal, 66, 3.

Comisión Reguladora de Energía (2001). Los nuevos retos de la industria del gas natural, 2001 - 2005. Hacia una estructura más eficiente y competitiva en la industria del gas natural en México, www.cre.gob.mx/boletines/2001 / consulta.pdf (1999). Permiso de transporte de gas natural otorgado a Pemex gas y petro­

química básica para el sistema nacional de gasoductos, www.cre.gob.mx / permisionarios/gas/ g061tra99.html. (1999). Escenarios de oferta y demanda en el sistema nacional de gaso­

ductos de Pemex-Gas. (1998). Resolución sobre el programa temporal de cobertura de precios del

gas natural para el invierno 1998/99, www.cre.gob.mx/registro/resoluciones /1998/Resl41-98.pdf.

Dahl , C. (1992). Energy and Energy Product Demand, Elasticities for the De­veloping World: A Survey of the Econometric Evidence, Colorado School of Mines, Working Paper 92-14.

Laffont, J . J . (2000). Incentives and Political Economy, Clarendon Lectures in Economics, Oxford, Oxford University Press.

Mas-Colell, A , M . D. Whinston and J. Green (1996). Micro-economic Theory, Oxford, Oxford University Press.

OGS Special (1995). Oil and Gas Journal. Pemex (1998). Indicadores petroleros y anuario estadístico. Ramírez, J . C. and J. Rosellón (2002). "Pricing Natura l Gas D i s t r i b u t i o n in

Mexico" , Energy Economics, 24, 3. Secretaría de Energía (2001). Prospectiva del mercado de gas natural, 2001 -

2010. www.energia.gob.mx/secundario/publicaciones.html. Stauffer, T . R. (1996). The Diseconomies of Long-Haul L N G Trading, Interna­

t ional Research Center for Energy and Economic Development, Occasional Paper No. 26.

Vogelsang I . (2001). "Price Regulation for Independent Transmission Compa­nies" , Journal of Regulatory Economics, 20, 2.