PONay R.swch | IPS1\ ZO WORKING,PAPERS, International Trade International Economics Department The World Bank March 1993 WPS 1120 Policies for Coping with Price Uncertainty for Mexican Maize Donald F. Larson More efficient, market-based policies should be inrduced for coping with uncertainty about international prices for maize in Mexico. Policy RwAisrhWoddngP1p=sdiuntietfindings of woalkin pig rn en0 getheexchngeofides among an saffand alothuintin deve1 opmnaailw.hesepam striuftedbytheResrachAdviswyStff,canythenamesoftheut ,rilnect aolytirviewsv.nddswhdbeaedandckedacrdiey.Thefindingnzo.tios,andon1uionsuatheauthoIsowniTheyshoudd not be asibuted to the Wodd Bank, its Board of Dinctor, its managanan, or any of its manbercounties. Policy Reserch| International Trade| WPS 1120 This paper-a product of the Intemational Trade Division, Intemadonal Economics Department-is part of a larger effort in the department to improve the developing countries'.management of commodity price risk. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Dawn Gustafson, room S7-044, extension 33714 (March 1993, 27 pages). The three goals of recent agricultural pricing offer farmers an inexpensive way to provide in- policies in Mexico for maize have been to raise season price stability. But the farm sector can farm income and c op prifitability by boosting take advantage of these instruments only if the domestic prices through trade restrictions, to domestic distribution system is reformed - by provide some pdice certainty at planting time, liberalizing interstate trade, harmonizing stan- and to reduce year-to-year variations in maize dards and measures (including sanitation stan- prices. dards), and privatizing storage facilities. The govemment pursued all goals jointly, No market mechanisms exist to ease the using import quotas and a state marketing underlying year-to-year price variability for agency to implement a mandated pan-Mexico wheat and maize. But the benefits of govermment price for maize. Farmers benefited primarily intervention to smooth prices are small and, in from the price stippor , and very little from the themselves, do not justify using a price-band other goals. Maize policies were unsustainable mechanism. and enormously expensive, so the govemnment has decided to reform the sector. [The reforms Still, a price-band system might be consid- will be institutionalized in the North American ered as a transitional tool. NAFTA calls for slow Free Trade Agreement (NAFTA.1; liberalization of the maize market, but the Mexican government could liberalize its markets Larson shows that the same price enhance- more aggressively. Levels of transfer under ment and stabilization could have been achieved current policies remain high and the costs of at less cost by using variable border tariffs within adjustment may depend on the path of interna- a price-band mechanism. Moving immediately to tional prices during the transition. The advantage such a policy can lower costs yet produce the of the price-band mechanism is that relief is same effects as current policy. The multiple granted (transparently and automatically) to effects of policy on price can be measured consumers when prices are abnormally high and separately, and a variable tariff/price-band to producers when they are abnormally low. This scheme can be used to target both price levels would help forestall political pressures for ad and price variability. hoc measures. International markets in commodity futures and options (through millers and banks) could The Policy Research Working Paper Series disseminates the frndings of work under way in the Bank. Anobjective of the series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the Policy Research Dissemination Center Policies for Maize Price Variability in Mexico Don Larson Table of contents 1. Introduction 1 2. History of maize and wheat prices in Mexico 4 3. Within-year price uncertainty and free trade 6 4. Measuring the benefits to producers from price variability reductions 9 5. Targeting price variability with a price-band mechanism 13 6. Simulating historic price distributions in maize 20 7. Hedging government revenue risk 22 Conclusions 25 References 27 1. Introdection Agricultural pricing policies often have several targets. First, they may attempt to increase farm income and ctop profitability by engineering a higher domestic price than international markets would dictate. An additional goal may be to provide an indication at planting time of the price farmers can expect to receive at harvest, providing a measure of within-year price certainty. Finally, pricing policies often attempt to smooth farmer incomes across years by reducing price variability, smoothing the peaks and valleys associated with internatioxial commodity prices. In the recent past, Mexican pricing policies for maize attempted to meet all of the targets described above. By setting domestic prizes for maize, defended by import restrictions and domestic interventions in the marketing chain, the government assured farmers a higher-than-international price for their product and a more stable price for their product, and offered a guarantee that, at least in nominal terms, the price that farmers expected at planting time would be the price they received at harvest. However, these policies have been implemented at tremendous cost to consumers and taxpayers. With the government's recognition of the unsustainability of these policies and the overwhelming benefits of reform, analysis was carried out to identify more effective substitutes for these policies within a well- functioning market economy. Table 1 provides a summary of the different characteristics of current' price policies for maize in Mexico. From results reported later in this section and from the results of earlier research, especially 'Since this paper was written, the terms of the yet-to-be-ratified North American Free Trade Agreement were announced. Under the agreement, Mexico will convert its import licensing regime for maize from the United States and Canada to a transitional tariff-rate quota (TRQ). The TRQ will be in effect for 15 years. In the first year of the agreement, the US will enjoy duty-free access on 2.5 million tons of maize. The duty-free quota will grow at 3 % per year for the 15 years that tariffs are in effect. The initial tariff paid -in imported maize above the quota will be 215%. During the first six years of the agreement, an aggregate 24 % of the over-quota tariff will be eliminated and the remainder will be phased out over the rest of the 15-year transitional period. Mexico is free to pursue a more aggressive quota liberalization scheme. I Table 1: Summary of price-policy effects. Components of prking policies Prke level Current pollale rise domestlo puica (by llilng imporu), thus gonerating increased producer incomes, welfare losso to consumen, and net transfers from government revenues. Price risks: - Within-year price uncertainty Since time passes between planting and harvesting, farmers are subject to unexpected price changes after planting decisions are made. Current policies provide an announced guaranteed price, removing this uncertainty to the extent that producers are confident that government resources are available to defend the announced price. - Price variabWty across years International market prices for some commodities are inherently unstable with significant year-to-year variations. Current policit s reduce this variability, but do so at high cost. In addition, the producer benefits generated may be small compared to policies that raise average prices. Levy and van Wijnbergen (1992) for maize, it is clear that large redistributions of incomes and welfare will come from reducing the wedges between domestic and international prices for maize and wheat. As these effects are treated in detail in other papers, they receive only passing treatment here. Rather, this paper deals with the benefits flowing from the two other price-policy effects: reduction of within-year price uncertainty and reduction of price variability across years. Results for wheat are presented along with results for maize to illustra, differences between these two markets. The conclusions of this paper are, first, that within-year price variability can be addressed inexpensively using futures and options markets, once a link between domestic producer and international prices has been established. Unfortunately, recent experiences with sorghum and soybeans indicate that the removal of trade restrictions in and of itself is not sufficient to establish such a link. Additional reforms along the marketing stream, including liberalization of transportation along state highways, harmonization of sanitation regulations between states as well as between nations, adoption of international standards in 2 grading and inspection, and privatization of storage are required before inexpensive market-provided hedging methods can be of use to producers. Secondly, although producers derive benefits from government policies that reduce the variability of annual prices, these benefits are quantitatively small compared to those deriving from policies that attempt to raise price levels. The present mix of policies - Import quotas and domestic management of storage and distribution - generates these benefits in an inseparable fashion. A tariff-based, price-band scheme is described which allows policy makers to separate price-level targets from price-variability targets. Although the average economic returns from reducing price variability in the short run may be small, the price-band system does provide a transparent and systematic method of responding to consumer and producer pressures during periods of abnormally high or low prices. The system is easier to administer than the current system, relying only upon variable border levies, and allows distribution to be determined domestically in response to demand and transportation costs. The remainder of the paper is organized in the following way. The history of Mexican maize and wheat prices is presented in section 2; in section 3, within-year price uncertainty and market instruments for risk are discussed; in section 4, the benefits of past policies in maize and wheat are estimated; section S demonstrates how a variable-levy price-band system can be used to separately target price levels and price variability; section 6 demonstrates that a price-band system could have been used to replicate the historic average level and variability of maize producer prices in Mexico while generating about $300 million in annual government revenues; section 7 discusses how government liability generated by a price-band system can be hedged using existing market instruments; and section 8 concludes. 3 2. History of maize and wheat prices In Mexico For the last three decades maize prices for producers in Mexico have been both higher and less variable than th, price of internationally traded maize (see Figure 1.) In recent years, the high domestic prices paid to producers have been achieved through a state monopoly which limits imports to defend a guaranteed price. The benefits transferred to producers through the higher prices are in effect paid by rural consumers in the form of higher food prices and by taxpayers. The state-subsidized commodity-marketing agency, CONASUPO, buys a significant portion of maize output from producers at national guaranteed prices and imports maize at inwernational prices2. CONASUPO then sells the maize to regional millers at differing prices with a substantially lower price granted to millers in the Federal District. Transportation and handling expenses are absorbed by the agency. These are substantial since the uniform domestic price which CONASUPO is required to pay does not reflect regional differences in transportation, storage, and marketing costs. The current regional wholesale maize and tortilla prices are given in Table 2. Maize prices 1980 pesos per ton 6. 500 8. 000 5. 500 S. Soo I 5. 000 4, 500- 4, 000- 3. 500- 3. 000- 2.5019161 i696 196i 197i 197-7 1981 1985 1989 1963 1967 1971 1975 1979 1983 1987 -Dormestic price -~ US price of border Figure 1: Comparison of maize prices, adjusted for transportation. 2Some poultry pr--Aucers are allowed to import maize at international prices. 4 Table 2:Wholesale corn and retail tortilla prices. Location Maize TortiUa pesos per ton Federal District 510,000 750,000 Veracruz 900,000 1,150,000 Nuevo Leon, Coahuila, Durango 865,000 1,100,000 Rest of Zountry 800,000 *1,050,000 Source: World Bank As illustrated in Figure 2, the differences between international and domestic wheat prices have been fairly small after accounting for transportation and varietal differences. Over the period 1981-90, domestic wheat prices have ranged above and below international prices. However, the average spread between domestic and international prices was positive (about 12%) during this period and domestic prices were less variable. Until recently, CONASUPO handled the ,narketing of wheat in much the same way as maize is currently marketed; however, with liberalization in wheat trade, ASERCA now has responsibility for facilitating wheat marketing. Wheat farmers still enjoy a measure of price protection Wheat prices 1980 pesos per ton 7.500 1 7 000 6,5000 6, 00 5, 000 IA 4. 500 4. 000 ,kTA4t 3. 500 3. 000 1961 1965 1969 i973 1977 1981 1985 1989 1953 1967 1971 1975 1979 1983 1987 -- US price at border Domestic price Figure 2: Comparison of wheat prices, adjusted for transportation S and ASERCA must resort to subsidies in order to market the wheat in the presence of wheat flour imported under low tariffs. It is an indication of the price distortions generated under current pol'ties that ASEkCA recently found it cost effective to subsidize the use of wheat as animal feed In northwest Mexico rather than transport it to urban centers and subsidize Its conversion to flour. 3. Within-year price uncertainty and free trade Under current policies, Mexican producers know that the price announced prior to planting is the price they will receive at harvest. While this provides a benefit to producers, it is a benefit which could be readily and inexpensively provided by private markets for hedging - primarily international markets for futures contracts and options on futures. It is not necessary for the individual farmer to understand and participate directly in such markets since it is often in the interest of intermediates such as millers or banks to offer the hedging to farmers. For example, a wheat miller close to an urban center may have a transportation-cost advantage over competitors if local wheat can be secured. Figure 3 provides an illustration of how a miller anxious to secure local production might bundle hedging services in his price offer to local risk-averse farmers. Another example of how market forces might provide for hedging services is that of a bank already lending to local farmers. Such a bank might offer the same type of program to reduce the chance of default on farm loans. For example, the bank might purchase a put option (which gives the owner the right to sell a certain amount of a commodity at a specific price) in Chicago to protect a customer/farmer at the beginning of the crop year from an unexpected price decline at harvest. The bank would loan the producer the cost of the put along with the balance of the loan and recover the costs at the end of the crop year. Should prices unexpectedly fall, the value of the put would increase in a compensating manner, protecting the producer from an income loss and, indirectly, the bank from a default on the loan it extended to the farmer. Protecting the farmer against downward price risk is in the self interest of the bank since it also reduces the likelihood of producer default and increases the bank's recovery rate. 6 Chicago closing priCes: ODeember 4, 1991 tutrt opt IOwE ,Juiy '92 mail. .luly 92 put with 82 . 60 atrike prIc 82 . 6075 i bu. S 0 `13 / bu. buyS put / Miler 12 . 50 - transport Coats - 8 . 13 if price goes up, ferivar caaoults but .waa mre than miller 'a offer If Orice goes down, miller obiigated to wY U . 60, b;jt 1 coe,enasted by Increased val ua of opt ion Figure 3: Private market hedge for guaranteed maize price. International markets for hedging instruments for many commodities are readily available and inexpensive. For example, the cost of a six-month forward put in Chicago for maize at 'expected' prices, as indicated by the futures market, was slighty more than $5 per ton on December 4, 1991 - about 5% of the spot price. Physical distance is usually of little consequence in hedging commodity price risk. The Cocoa Marketing board in Ghana actively hedges cocoa prices and has an office in London to do so, while copper producers in Peru n-Aake use of markets in New York. However, it is vitally important that the links between one location and another be efficiently made. The removal of trade quotas is a necessary condition for such a link to be established but is not, by itself, sufficient. Recent experience in the Mexican soybean market tjest additional reforms along the marketing chain are required before Mexican producers have access to inexpensive markets to hedge within-year price risk. The producers themselves will not undertake the hedge; this is more effectively done by intermediaries such as central storage facilities, millers, exporters, or banks. 7 Recently, ASERCA asked Merrill Lynch to investigate the use of futures markets for soybeans which are currently freely traded. The chain of markets between a soybearn producer in Mexico and Chicago can be segmented in the following way. The future price of soybeans in Chicago in SUS plus forward Chicago-to-border transportation costs in $US times the future peso/dollar exchange rate plus local transportation costs should equal the local wholesale price; otherwise, arbitrage opportnities exist. Forward markets exist for the peso as well as for sea transportation from the United States. However, Merrill Lynch found that the variability over time in local transportation costs was so large that the benefits of using the Chicago markets to hedge domestic price risk in Mexico were lost. Put another way, Merrill Lynch concluded that the primary uncertainties faced by the Mexican farmer were not the variations in international prices, international transportation, or exchange rates, but were associated with domestic marketing costs. Two physically separate markets will be linked as long as there are ways in which unwarranted price - 'fferentials can be arbitraged away. Ultimately, this requires the opportunity for physical goods to be anoved from one market to another. Several bottlenecks occur in the local marketing system which must be addressed before the ability to arbitrage distant markets can be established and full advantage taken of the opportunities available to hedge the prices. These include: 1) the liberalization of transportation along state highways; 2) the harmonization of sanitation regulations both among the Mexican states and between the United States and Mexico; 3) the adoption of international standards in grading; and 4) the privatization of storage. The following examples illustrate the difficulties that remain with lingering regulation even as most sectors in Mexico are being rapidly deregulated. At one time, freight rates were regulated along both state and federal highways. While freight regulations on federal highways have beer. dropped, not all state regulations have been eliminated. As a result, transportation costs may be substantially different 8 from one locale to another. In addition, some state governments have been accused of using sanitation regulations as a barrier to interstate trade. In order to arbitrage markets, some certification is required to demonstrate that the good is of sufficient quality to trade on both markets. Generally, issues of inspection and certification are handled routinely either by government or private Institutions. However, when a central agency has monopolized international trade as has CONASUPO in Mexico, the existence of such institutions cannot be taken for granted. Finally, much of domestic storage is government-owned in many parts of Mexico and, according to anecdotal evidence, not properly priced. During harvest periods, storage may be rationed, forcing a larger quantities onto the local market than would be justified if storage costs and supplies were market-dictated, thereby depressing local prices. Without adequate access to markets for commodity price risk, trade liberalization will unnecessarily impose further costs on those farmers and intermediaries who would desire to be hedged at market prices. While efficient domestic markets for storage, transportation and marketing are necessary to take advantage of international markets for commodity price risk, they are also essential for the efficient allocation of resources within the domestic economy. Though raised here in connection with the use of hedging instruments, reforms in domestic distribution networks have important implications for the entire economy. Addressing such issues early during trade liberalization will greatly ease the costs of transition. 4. Measuring the benefits to producers from price variability reductions While the benefits to producers of receiving a higher price are obvious, the benefits of a more stable price are dependent upon how farmers view risk. To the extent that producers prefer stable incomes to unstable incomes, additional benefits accrue over the life of a price stabilization program 9 based on the efficacy of the program in reducing income variability by reducing the variability of the price component of producer income. Newbery and Stiglitz (1981) derived a quantifiable measure of the value of the income stabilization achieved based on assumptions concerning the relative risk aversion for producers. Assuming that producers can be treated as a single aggregated agent whose utility can be represented by a Von-Newman Morgenstern Utility function of income U(Y), average benefits relative to income are defined as: _ i Y _ I(f/0_2 B Y,-o 2 1 y 0 2R('o)[o271(Y/Y)2-o2rO] 1 where B is the money value of the stabilization benefits; YO, Y, represent income without and with a stabilization program, respectively, and a bar over a variable represents the variable's mean; a2y is the square of the coefficient of variation for the income Y; and R is the coefficient of relative risk aversion given by: R u -y.rL(nŽ (2) U() Tme first term in (1) is a transfer benefit resulting from any change in the mean level of income, while the second term measures the benefit directly attributable to a reduction in the variance of income resulting from a reduction in the variability of price. By making some assumptions about supply elasticities and risk-aversion, the value of the two separate benefits can be approximated. Given the assumptions listed in Table 3, government intervention generated transfer gains of SUS 1.964 billion a year for producers since 1981, and generated price stabilization gains equal to about $US 54 million. Had international prices prevailed in Mexico, maize prices would have been on average 38% lower, and consumption of maize about 18% higher. Several 10 Table 3: Partial equilibrium effects of pricing maize at intemnational levels. Assumptions price: equal to international levels adjusted for transport land-use elasticity: -irrigated 1.50 -rain-fed 1.03 yields: remain at historic levels currency units: constant 1980 pesos demand elasticity: -0.80, with 60% of demand at rural prices re aive risk aversion: unitary Simulation results Elistoric Alternatve mean price (1980P/ton): 5,507 3,399 coefficient of variation (CV) of price: .096 .183 irrigated area ('000 ha.): 920 395 rain-fed area: 5,891 3,567 producer revenut (billion 1980 pesos): 67.33 24.09 coefficient of variation (CV) of revenue: 0.184 0.440 demand ('000 tons): 14,928 17,598 points should be noted right away. First, the transfer loss to maize farmers is a partial equilibrium result. For the producers, part of the 'transfer' will be offset by growing other crops. In addition, losses to producers, along with general efficiency gains accrue to others - primarily consumers - in the larger economy.: Secondly, even though the transfer benefit represents a maximum value, the benefits of the existing programs for the maize farmer come overwhelmingly from raising the average price, not from reducing price instability. This second point becomes especially clear once inflation Is considered. Thirdly, risk-aversion probably rises with a reduction in wealth. A large once-and-for-all drop in maize pAces will result in the devaluation of farm-land used to produce maize and the income of many maize farmers will fall. To the extent that an arbitrary measure of risk aversion is correct before maize prices TFor more general results for Mexico, see Levy and van Wijabergen (1992) and Robinson et. al. (1991). Braveina et. al. (1992) show that similar methodology provides a good approximation of more generalized results in the cm of Brazilian agriculture. 11 are lowered, the measure will be higher after maize price reductions. At the same time, stabilization benefits still remain small relative to transfer benefits, even when the measure of risk aversion used In the calculations is doubled or tripled. Farmers deal in nominal prices and must distinguish between inflation effects and real effects in prices. In the analysis presented above, farmers are assumed to distinguish completely and correctly between movements in real prices and general inflation - i.e., there is no money illusion. However, with rapid inflation and variable inflation, a large component of price instability comes from the inflationary component and the assumption of no mon.ey ilusion may be too strong. To the extent that money illusion does occur, it tends to further reduce perceived stabilization benefits. To illustrate this, note that from 1980 to 1990 the coefficient of variation (CV) for domestic maize prices equalled 1.21; the CV for international prices converted to pesos and adjusted for transportation equalled 1.22. From a statistical point of view, it is impossible to reject the ncti'n that the two variances arise from the same distribution.' The effects of moving to international prices for wheat are less dramatic (see Table 4.) Past policies generated income transfer effects amounting to $47 million per year, while price stabilization generated $21 million a year in risk benefits. Risk-benefits for wheat, though smaller than the risk- benefits generated for maize, were a much greater proportion of total benefits in wheat than in maize. However, the same model can be used to show that very small levels of average protection (4.5%) would have fully compensated wheat farmers for the risks implied by international price variability. 'Testing the hypothesis that the two variances are different resulted in an F-statistic of 0.35. A value of 1.83 is required to reject the hypothesis at a 95% confidence level. 12 Table 4: Partial equilibrium effects of pricing wheat at international levels. Assumptions price: equal to international levels adjusted for transport land-use elasticity: -irrigated 1.10 -rain-fed 1.02 yields: remain at historic levels currency units: constant 1980 pesos relative risk aversion: unitary Simulation results for 1981-1990 Hitoric Akernative mean price (1980P/ton): 4,193 4,020 CV of price: .06 .12 irrigated area ('000 ha.): 860 828 rain-fed area: 151 142 producer revenue (billion 1980 pesos): 13.867 13.443 CV of revenue: 0.28 0.47 Transfer benefit (SUS): 47 million/year Risk benefit (SUS): 21 million/year 5. Targeting price variability with a price-band mechanism Under current policies, the average level and the variability of maize prices are controlled by restricting imports and by government controls over managing domestic distribution. Besides being economically inefficient in distorting the allocation of resources, the system is cumbersome and costly since it requires the government to intervene at all stages of trade and marketing, assuming responsibility for anticipated demand and supplies, providing storage and distribution systems, and managing imports to clear physical markets at predetermined prices. These activities provide no direct benefit to either producer or consumer. Any average level of support to producers provided by current interventions in the physical markets can be duplicated by intervening only at the border with variable tariffs. In a price-band scheme (where domestic prices are only allowed to vary within pre-determined bounds) variable levies can be 13 used to independently determine price levels and price variability while allowing domestic markets to efficiently handle allocations. Far fewer resources would be required to manage the scheme compared to the current system. Moreover, a variable tariff scheme is a useful instrument in the context of trade negotiations since support levels can be phased out in a smooth and continuous fashion. The scheme has the added advantage that it defines in a transparent manner the way in which the government will intervene to provide additional relief to consumers during periods of "high' prices and to producers during periods of "low" prices, preempting calls for ad-hoc assistance. A price-band system uses a moving-average of border prices to establish a reference point for domestic prices. A general tariff level is applied to all prices when prices fall within some predetermined range of the moving-average (say 10%).' When international prices move above the predetermined upper bound, the general tariff Is reduced; when international prices move below the lower bound, the general tariff is increased. A price-band mechanism allows policy makers to address separately the two components of price policy in the following manner: the average level of price support can be targeted by fixing the general tariff level, while the price-variability component can be targeted by adjusting the width of the band. For a price-taking country like Mexico there are nine exhaustive possible states defined by the relationships between border prices, the price band, and trade flows; the price band may fall In a range in which the country would always be a net exporter or a net importer, or it may straddle the point at which domestic supplies equal domestic demand. In addition, the border price may fall above the band, 'Since the mid-1980s, Chile bas oporated a price-band schome for wheat, oflseeds, and sugar. In assessing the program for wheat, Muchnik and Allue (1991) found the mechanism weasy to implement and has been well administered. But the procurement price determination mechanism used ... has been the object of bargaining and political pressures exercised by farmers, while consumers have had no voice in the process (p. 73). 14 below the band, or within the band. In operating a price-band program, price risk is transferred from the producers and consumers to the government. The government gains revenues at some times and pays out at other times. In four of the nine possible states the stabilization mechanism would produce revenue; in two of the nine states the mechanism would generate a loss; and in three of the states there would be neither a pay-out nor revenues generated. Figures 4-6 illustrate the three possible states if the country always exports the "stabilized' commodity. In Figure 4, the border price (Pw) falls above the upper limit of the price band. In order to peg domestic prices at the top of the band (Pu), the government would impose an export tax equal to Pw minus Pu. If producers correcdy anticipate the prevailing domestic price, Qs will be produced. At the prevailing price of Pu domestic demand will be Qd. Qs minus Qd will be exported generating (Pw- Pu)*(Qs-Qd) in revenues for the government. Figure 5 illustrates the state in which the border price falls below the lower range of the band (PI). In this case, the government must first impose an import tax equal to (Pl-Pw) to prevent less expensive foreign supplies from filling domestic demand, then subsidize exports (Qs-Qd) by (Pl-Pw). The import tax will generate no revenues, and the net loss of revenues from the government or stabilization fund would equal (Qs-Qd)*(PI-Pw). In Figure 6, the border price falls within the price band; import and export taxes are set to zero and the fund neither gains nor loses revenue. Figures 7-10 illustrate the three importing states. In Figure 7, the border price falls above the price band. The government must impose an export tax equal to (Pw-Pu) to prevent domestic supplies from flowing to the more profitable export market. In addition, imports (Qd-Qs) must be subsidized by .. '-Pu). In this state the fund loses (Pw-Pu)*(Qd-Qs). Figure 8 illustrates the importing case when the border price falls below the price band. The government imposes an import tax equal to (Pl-Pw) which 15 generates (Qd-Qs)*(PI-Pw) in revenue for the stabilization fund. Figure 9 shows the state where the border price falls within the band, generating neither revenue nor losses. Finally, Figure 10 illustrates the case when the price band straddles the point at which domestic supplies equal domestic demand. When the border price falls above the upper range of the price band the country will be a net exporter. To bring the domestic price in line with the price band (Pu), the government must impose an export tax equal to (Pw-Pu) which generates revenues equal to (Pw-Pu)*(Qs- Qd). When the world price (Pw') falls below the price band, the country is a net importer. The government imposes an import tax equal to (PI-Pw') and collects revenue equal to (Pl-Pw')*(Qd'4Qs'). If the price falls within the band, the country may either import or export, but tariffs will be set to zero and no revenues will be generated or lost. 16 Pv ~~~~~~Supply Supply band.~ ~~~~~~~Din Demand ,2u _P\/ Pv I i\ band.~~~~~~~~~~~ ~P Dband. 1<~~~~~~~~ I Figure 6: Exporter fadng world price above Flgure 7: Importer faing world price below band. band. pi &and Pi I I nd Os Od 0d Qd Figure 4: Emporter facing world price above Figure 5: Exporter facing world price betlo band, band. P1~~~~~~~~~~~1 Supply Pv Pu PI~~~~~~~~~~~~~~ Qd 0si ad' as Figure 10: Marginal trader. By intervening in the domestic market through a variable tariff, price-band scheme the goverunment enforces welfare transfers between consumers and producers as well as giving rise to efficiency losses. These transfers occur for each period the govermnment or fund manager intervenes and may have off-setting effects. In fact, one of the advantages of a price-band mechanism is that the average effect on domestic prices is neutral - see Coleman and Larson (1990). In addition, by stabilizing the price component of producer income the program also generates a stabilization benefit which occurs over the life of the program. Efficiency losses and the income transfers between producers, consumers, and the governranent are readily calculated through changes in the traditional nmeasures of consumer and producer surplus. These measures give a general indication of the welfare gains and losses for each is period in which the government intervenes.6 Figure 11 illustrates the consumer, producer, and government surpluses generated by imposing an \ export tax on an exported good in order to lower P 8 Ii d P.* -' domestic prices. Domestic prices fall from the j I border price of P to P' as the government imposes a X tax equal to (P-P'). Demand increases from Qd to Qd Qd' Qs's Qd' and supplies decline from Qs to Qs'. The Figure 11: Income transfers and effirlency government receives revenues equal to area d; losses under an export tax. producer surplus drons by an amount equal to the sum of areas a,b,c,d, and e; consumer surplus increases by an amount equal to areas a and b, leaving an efficiency loss equal to areas c and e. The way in which income is transferred, or lost to inefficiencies, will vary depending upon the type of government intervention and will differ from period to period as the type of intervention needed to defend a price band changes with international price movements. In general, the transfers between producers, consumers, and the government stabilization fund are likely to be offsetting. Inefficiency losses, however, are not offset and are a social cost incurred by operating a price stabilization program. Table 5 lists the income transfers associated with each of the possible nine states that can occur under a price-band scheme. 60'ly under very strong assumptions do ordinal monetary measures of consumer surplus correspond to unique measures of consumer welfare. However, for applied work there are few practical alternatives. See Just, Hueth, and Schmitz (1982), chapter 5 for a discussion of consumer surplus in applied economic analysis. 19 Table S: Producer, consumer, and government surplus gains and losses under a price-band stabilization scheme. Do_der prieo Trade State Above band Witbi band Below band Exportor Consumer surplus gain neutral los Producer surplus lou neutral gain Fund revenue gain neutral loss Marginal trader Consumer surplus gain neutrl lose Producer surplus lou neutrml gain Government revenue gain neutral gain Impoiter Consumor surplus gain neutral lo" Producer surplur lou neutral gain Government revenue loss neutra gain 6. Simulating histk ilc prize distributions in maize A simple, but revealing way of showing how a price-band system works is to identify the type of system which would provide the same domestic price characteristics for maize that the current system has provided over the last ten ) ears. As it turns out, the Government of Mexico could have achieved the same level of price support for maize producers as achieved by current policies by applying an average tariff of 60.5% on maize imports and operating a price-band plus/minus 8% of a moving five-year average of border prices.7 The results are represented in Figure 12. The size of the fixed tariff component puts into perspective the magnitude of the transfers and distortions involved and gives an indication of the magnitude of the adjustments which will occur. Since the simulated variable tariff MThe upper bound for the price band therefore becomes (1.605)*(l.0&)*p, where p is a 5-year moving average of the maize border price. The lower bound becomes (1.605)0(0.92)*p. The use of a 5-year moving average in the simulations is arbitrary, but is similar the 60-month moving-average used in Chilean price-band schemes. Coleman and Larson (1991) review various price-band mechanisms in the context of Venezuela. 20 Simulated effects of a price-bond on malze prices In Mexico 7. 000- 6,500 > =2 S. 000 5. 500 *4. 500 CD 4. 000 300 2950 1984 1988 1982 1986 1990 -i- Bordwr price - Simulated price Figure 12: Effects of a 60.5% tariff and a plus/minus 8% price-band on a five-year moving average of international maize prices. program was intentionally designed to mimic the historic price distribution (mean and variance) of maize in Mexico, demand, supply, and import levels are only slightly affected by the switch in the protection mechanism. The single quantitatively differenit effect is the tax revenue generated by the tariff. For four of the simulated years, 1982, 1985, 1986, and 1988, the tariff-level would have stood at 60.5%. For five years the tariff would have been raised in order to defend the lower range of the price band (to 85% in 1980, 92% in 1981, 89% in 1987, 63% in 1989, and 74% in 1990); and for two years (1983-84) the tariff would have been lowered (to 32% and 53%) to defend the upper-band. Ihe scheme would have generated about $300 million in government revenues annually. The net savings to the government would be more since CONASUPO operates at a net subsidy. Against this gain to the government, consumers would take on the costs of distribution. In addition, certain livestock producers would no longer enjoy the privilege of importing international maize tariff-free. 21 7. Hedging government revenue risk The underlying variability of commodity prices is determined by a combination of deterministic market forces and stochastic forces of nature. As such, the risks associated with unanticipated prices changes cannot be eliminated, but rather reassigned, pooled, or otherwise managed. By operating a price- band scheme to stabilize domestic prices, the risks faced by producers are transferred to the govermnent budiget. (Of the nine possible states relating prices and trade that can occur under a price-band mechanism, only two create a liability for government revenues, while four of the states generate revenues, see Table 5.) In addition, while the liability faced by the government is limited, the limit may be quite large. Conversely, the potential for tax revenue is not bounded. Figure 11 illustrates the case for an export tr. The area above the price band and between the supply and demand curves is unbounded above and represents the potential area that could be used to finance price stabilization via an export tax. Te area between the lower range of the price band and the demand and supply curves at P1 down to the axis represents the maximum payout from the government for an exporter. However, the value of this area goes to zero as the lower range of the price band falls. A similar situation exists for the other fliableN state, the case of subsidized imports.' For the exporter case, the maximum government liability has the following characteristics as prices fall: lim ,I(-p)[SWP)w-D(P1)] 0, O3) where S(p) = D(p). For the case of the importer: limp. (p8-p) [StP)-D(p)]= O, (4) where S(p.) = D(p.). 'In the importer case, total import tax revenues are bounded as well, but only by the non-negativity condition on prices. 22 For commodity markets in which futures are available, the government can further restrict the payout by b'iying options to hedge its liability. Consider the following example. In period one, producers, consumers and the government-risk manager know what the price band will be, but do not know the stochastic component of the international price. Furthermore, the risk manager knows the demand and supply curves of the relevant commodity. In period two, the stochastic component and therefore the international price will be revealed and the risk manager will have to defend the price band. In order to limit the potential payout by the government, the risk manager looks at the range of the prices covered by the band. If the country would export over the entire range of the band (see Figure 4), the government faces a liability only if the intetnational price in period two were to fall below the lower range, that is, below PI. Therefore to hedge that liability, the risk manager hedges the quantity S(PI)- D(PI) by purchasing put-options at a strike price of PI, for a delivery date corresponding to period four." Should the border price actually fall below P1, the added value of the put option would compensate the government for additional outlays. The government's liability is thereby limited to the purchase cost of the options. This amount is not inconsequential, but it is also not overwhelmingly large when compared to the costs of a 60% tariff. By way of a realistic illustration, the Mexican government could find itself hedging about 2.5 million tons, or about 98.5 million bushels. A typical cost of a 6-month-ahead put- option for maize when the strike price matches the futures price might be about 7 cents/bushel. The government's losses would therefore be limited to about $6.9 million. In the case where the range of the price-band implies that the country will be an importer, the government is liable when the border price moves above the upper bound of the price-band range (see Figure 8.) In this case, the risk manager can limit tne government's liability by hedging the quantity 9An optimal 'hedge-ratio' based on the price of the option and the underlying variability of the commodity price determines the cost-effective number of options to purchase. It is possible that 'no hedge' is the optimal hedge. 23 [D(Pu)-S(Pu)] through the purchase of calls for a strike price of Pu for delivery in period two. Should the border price move above Pu, the increased value of the call options would compensate the government for additional payouts. When expected revenues from the tariff are significantly large, and when a short-fall in revenues generates budgeting problems, the government may want to 'lock.-in' expected revenues. To illustrate this, suppose the government is operating a price-band stabilization scheme and expects prices to be below the lower price band, generating revenues of $300 million. An unexpected price rise would greatly reduce the levy applied to protect the lower band. The unexpected price change would not generate a liability, but would cause a short-fall in revenue. The government could lock-in an expected price, and therefore an expected revenue level, by buying a call on the expected Import level at the expected price. If maize prices were to rise, the value of the call option would increase offsetting the loss of revenue due to the reduced levy. Should the price of maize fall on the international market, the value of the call would decline, diminishing any windfall coming from a higher-than-expected levy. Before deciding to operate a hedging mechanism to help budget a price-band scheme, the government should analyze its sources of revenues and look for correlations among the sources. Any hedging should be considered within the context of the goverment's portfollo of revenue sources. For example, if the government is operating more than one price-band system, negative correlations among prices may reduce the need for hedging - the particular commodities may generate a natural portfolio effect. If the prices are positively correlated - as is more likely - then the aggregate effect of a sudden set of price aberrations may be exacerbated and hedging becomes more important. 24 Conclusions In the recent past, the Government of Mexico has addresspd issues of seasonal price uncertainty, year-to-year price variability, and producer support in the maize and wheat sectors by limiting trade and managing domesdc distribution. There has been a recogniton of the costs and inefficiencies of past policies. Trade liberalization has begun for wheat and is anticipated in maize and this will reduce the income transfers to these sectors. It also raises questions about the policies to be adopted with respect to the volatility of international prices. While the effects of past policies on the average price of maize have been large, the beneflts associated with reductions in year-to-year price variability have not. In the case of wheat, past policies have generated smaller effects both on average prices and price variability. International markets in commodity futures and options offer an inexpensive way to provide within-season price stability for farmers and it will be in the interest of millers and banks to provide such services to farmers. However, before the farm sector can take fuil advantage of these instruments, reforms in the domestic distribution system are required, including the liberalization of state highways, harmonization of sanitation standards, harmonization of standards and measures, and the privatization of storage facilities. Market mechanisms do not exist to ease the underlying year-to-year price variability for wheat and maize. At the same time, the benefits from government intervention to provide a smoothing of annual prices are small and, of themselves, would notjustify a price-band mechanism. However, a price- band system might be considered as a transitonal tool. While NAFIA calls for a slow liberalizadon of the maize market, the Mexican government is not precluded from liberalizing its markets more aggressively. The level of transfers under current policies remain quite high, and the costs of adjustment may depend on the path of international prices during the transitional period. The price-band mechansm 2S has the advantage that relief is granted to consumers during periods of abnormally high prices and to producers during periods of abnormally low prices in a transparent and automatic way, helping to forestall the political pressures for ad-hoc measures. 26 REFERENCES Braverman, Avishay, Ravi Kanbur, Antonio Brandao, Jeffrey Hammer, Mauro Lopes, and Alexandra Tan, 1992. 'Conunodity Price Stabilization and Policy Reform: An Approach to the Evaluation of the Brazilian Price Band Proposals" Regional and Sectoral Studies, Washington D.C.: World Bank. Coleman, Jonathan, and Donald F. Larson, 1990. 'Tariff-Based Commodity Price Stabilization Schemes: The Case of Venezuela" Policy. Research. and External Affairs Working Paper 611, Washington D.C.: World Bank. Just, Richard, Darrell Hueth, and Andrew Schmitz, 1982. Aonlied Welfare Economics and Public Elcy. Englewood Cliffs: Prentice-Hall. Larson, Donald F., and Coleman, Jonathan Coleman, 1990. "Tbe Effects of Option-Hedging on the Costs of Domestic Price Stabilization Schemes." Policy. Research. and External Affairs Workinz Paper 653, Washington D.C.: World Bank. Levy, Santiago and Sweder van Wijnbergen, 1992, "Maize and the Free Trade Agreement Between Mexico and the United States", World Bank Economic Review 6(3):481-502. Muchnik, Eugenia and Marcela Allue, 1991, "The Chilean Experience with Agricultural Price Bands: the Case of Wheat", Food Polic 16(1):67-73. Newbery, David and Joseph Stiglitz, 1981, The Theory of Commodity Price Stabilization: A Stuy in the Economics of Risk, Oxford: Oxford Press. Robinson, Sherman, Mary Burfisher, Raul Hinojosa-Ojedo and Karen Thierfelder, 1991, "Agricultural Policies and Migration in a US-Mexico Free Trade Area: A Computable General Equilibrium Analysis", Departnent of Agricultural and Resource Economics. Working Paper 617, University of California at Berkeley. United States Departmnent of Agriculture, 1992, NAFTA Agriculture Fact Sheet: Commodities and Other Topics, mimeo. 27 Pollcy Research Working Paper Series Contact Title Author Date for paper WPS1 102 Did the Debt Crisis or Declining Oil Andrew M. Wamer February 1993 M. DMno Prices Cause Mexico's Invstment 33739 Collapse? WPS1103 Capital Mobility In Developing Peter J. Montiel February 1993 R. Vo Countries; Some Measurement Issues 31047 and Empirical Estimates WPSI 104 Trade Policy Reform In Latin America Asad Alam February 1993 J. Troncoso and the Caribbean In the 1980s Sarath Rajapatirana 37826 WPS1 105 Estimating Quasi-Fiscal Deficits in Philippe Le Houerou February 1993 N. Velasco a Consistency Framework: The Case Hector Sierra 34346 of Madagascar WPS1 106 Improving Women's Access to Halil Dundar February 1993 S. David Higher Education: A Review of World Jennifer Haworth 33752 Bank Project Experience WPS1 107 Flnancial Reform Lessons and Gerard Caprbo, Jr. February 1993 W. Pitayatonakarn Strategies kzak Atiyas 37664 James Hanson WPS1 108 Public Output and Private Decisions: Thanos Catsambas February 1993 A. Correa Conceptual issues in the Evaluation 38549 of Govemment Activities and Their Implications for Fiscal Policy WPS1109 Risk Management and Stable Andrew Sheng March 1993 M. Raggambi Financial Structures Yoon Ja Cho 37664 WPS1110 What Would Happen If All Developing Will Martin March 1993 D. Gustafson Countries Expanded Their Manufactured 33714 Exports? WPS1 111 Foreign Investment Law In Central Cheryl W. Gray March 1993 M. Berg and Eastern Europe William Jarosz 31450 WPS1 112 Privatization, Concentration, and Ying Cian March 1993 S. Llpscomb Pressure for Protection: A Steel Ronald C. Duncan 33718 WPS1113 The Lucky Few Amidst Economic Christiaan Grootaert March 1993 E. Vitanov Decline: Distributional Change In Ravi Kanbur 38400 Cote d'lvoire As Seen Through Panel Data Sets, 1985-88 WPS1 114 Does Price Uncertainty Really Anita George March 1993 D. Blevenour Reduce Private Investment? Jacques Morlsset 37899 A Smal Model Applied to Chile Pollcy Research Working Paper Series Contact Titb Author Date for paper WPS1 115 Looking at the Facts: What We Know Ross Levine March 1993 D. Evans about Policy and Growth from Cross- Sara Zervos 38526 Country Analysis WPS1116 Implications of Agricuitural Trade Antonio Salazar Brandao March 1993 D. Gustafson Liberalization for the Developing Will Martin 33714 Countries WPS1117 Porifolio Investment Flows to Sudarshan Gooptu March 1993 R. Vo Emerging Markets 31047 WPS1 118 Trends in Retirement Systems and Olivia S. Mitchell March 1993 ESP Lessons for Reform 33680 WPS1119 The North American Free Trade Raed Safadi March 1993 J. Jacobson Agreement: Its Effect on South Asia Alexander Yeats 33710 WPS1 120 Policies for Coping wih Price Donald F. Larson March 1993 D. Gustafson Uncertainty for Mexican Maize 33714 WPS1121 Measuring Capital Flight: A Case Harald Eggerstedt March 1993 H. Abbey Study of Mexico Rebecca Brideau Hail 80512 Sweder van Wijnbergen