SUGAR AND SWEETENERS February 08, 2001 January 2001, ERS-SSS-230 Approved by the World Agricultural Outlook Board --------------------------------------------------------------------------- SUGAR AND SWEETENERS is published three times a year (includes yearbook) by the Economic Research Service, U.S. Department of Agriculture, Washington, DC 20036-5831. This release contains only the text of the report -- tables and graphics are not included. Subscriptions to the printed version of the report are available from the USDA order desk. Call, toll-free, 1-800-999-6779 and ask for stock # SUB-SSS-4033, $25/2 issues. ERS-NASS accepts MasterCard and Visa. --------------------------------------------------------------------------- Contents Summary U.S. Sugar Year Ended, FY 2000 Prices Current Year, FY 2001 High Fructose Corn Syrup Special Article U.S. Sugar Price Indices and Stocks-to-Use Ratios Report Coordinator Stephen Haley (202) 694-5247 FAX (202) 694-5884 E-mail: shaley@ers.usda.gov Principal Contributors Stephen Haley Nydia Suarez Karen Ackerman Database Coordinator/Graphics & Table Design Fannye Lockley-Jolly Editor Martha R. Evans Layout & Text Design Wynnice Napper Approved by the World Agricultural Outlook Board. Summary released January 25, 2001. The next Sugar and Sweetener Situation and Outlook is scheduled for release on May 24, 2001. Summaries and full text of Situation and Outlook reports may be accessed electronically via the ERS web site at www.ers.usda.gov. The Sugar and Sweetener Situation and Outlook is published two times a year and supplemented by a yearbook. To order, call 1-800-999-6779 in the United States or Canada. Other areas please call (703) 605-6220. Or write ERS-NASS, 5285 Port Royal Road, Springfield, VA 22161. SUMMARY The U.S. Department of Agriculture (USDA) reported the impact of the Payment-in-Kind (PIK) Diversion Program on fiscal year (FY) 2001 sugarbeet and beet sugar production in the October 2000 Crop Production report and the October World Agricultural Supply and Demand Estimates (WASDE) report. Sugarbeet area harvested fell by 101,600 acres, or about 7 percent from the September 1 forecast; and sugarbeet production dropped by 2.030 million tons, or about 6 percent lower than its September 1 forecast. National yield was forecast upward by 0.2 ton per acre. The production of beet sugar was reduced by 320,000 short tons, raw value (STRV), or by 6.9 percent from the September WASDE. Total sugar production in FY 2001 is presently projected at 8.538 million STRV, down 130,000 tons from the December projection, and 6 percent lower than the FY 2001 estimate. Beet sugar production for FY 2001 is currently projected at 4.370 million STRV. Sugarbeet production is forecast at 32.521 million tons, about 3 percent lower than last year. Area harvested is estimated at 1.378 million acres. The yield estimate is 23.6 tons per acre, a record. Based on historical relationships, it is projected that beets sliced during the current crop year will exceed 30.4 million tons, or about 3.7 percent less than last year. Based on to-date sugar recovery rates and their average relationship with respect to the final sugar recovery rate, it is projected that the final recovery rate, net of sugar from molasses, will equal 270 pounds per sliced ton, down 24 pounds from last year. This rate, along with a projection for sugar recovered from molasses, implies sugar production for FY 2001 of 4.37 million STRV, or about 12 percent less than last year. Cane sugar production for FY 2001 is projected at 4.168 million STRV, about 2.5 percent above the estimated total for FY 2000. Production increases in Florida (154,000 STRV), Texas (75,000 STRV), and Puerto Rico (19,000 STRV) are expected to more than offset declines in Louisiana (92,000 STRV) and Hawaii (53,000 STRV). Florida cane sugar production is projected at 2.130 million STRV. The National Agricultural Statistics Service (NASS) estimates sugarcane production at 664,000 tons more than last year. Sugar recovery to date has been at a record high, implying that the season-average recovery rate will likely approach 13.2 percent. Texas cane sugar production is projected at 180,000 STRV. NASS estimates Texas area of sugarcane for sugar and seed at 46,600 acres, up 15,600 acres from last year. Yield is estimated at 37.7 tons per acre and the crop is estimated at 1.734 million STRV. Puerto Rico sugar production for FY 2001 is projected at 22,500 STRV. It is expected that the two sugar mills on the island will be back in production this year. Louisiana cane sugar production is projected at 1.570 million STRV. Although sugarcane area harvested has increased 35,000 acres over last year, sugarcane production is estimated down 275,000 STRV due to a continuing lack of adequate moisture. Sugar recovery is expected to be 5 percent lower than last year. Hawaii cane sugar production for FY 2001 is projected at 265,000 STRV. One of the three remaining Hawaiian sugar companies ceased operations in November 2000, and one of the other companies closed a processing facility. Sugar imports under the raw and refined sugar tariff-rate quotas (TRQs) are currently projected at 1.275 million STRV. As of January 8, 2001, sugar imports under the TRQs have amounted to 317,703 STRV, or about 25 percent of the amount projected to enter for FY 2001. Sugar imports outside the sugar TRQ for FY 2001 are projected to total 515,000 STRV, including 365,000 STRV under the combined Refined Sugar Re-export Program, the Sugar- Containing Products Program, and the Polyhydric Alcohol Program. The USDA projects an increase in U.S. sugar supply to 125,000 STRV as a result of sugar syrup imports under HTS 1702.90.40. High-tier tariff sugar imports for FY 2001 are projected to increase to 25,000 STRV, up from 6,000 STRV in FY 2000. Sugar exports under the Refined Sugar Re-export Program for FY 2001 are projected at 175,000 STRV. Deliveries to domestic food and other products manufacturers under the Sugar-Containing Products Re-export Program are projected at 125,000 STRV. Deliveries for the Polyhydric Alcohol Program are projected at 15,000 STRV. Total deliveries for FY 2001 are projected at 10.385 million STRV. After netting out deliveries made for the Sugar-Containing Products and Polyhydric Alcohol Programs, along with deliveries for livestock feeding (20,000 STRV), domestic food and beverage deliveries are projected at 10.225 million, about 2.3 percent higher than FY 2000. Ending stocks are currently projected at 1.987 million STRV, for an ending stocks-to-use ratio of 18.8 percent. Of the total, the Commodity Credit Corporation (CCC) owns 793,555 STRV or 39.9 percent of total projected ending stocks. The ratio of privately- held ending stocks-to-use is projected at 12.5 percent. The CCC acquired 793,669 STRV in October as a result of loan forfeitures. Adding this amount to what the CCC had already owned (296,649 STRV) gave them a pre-PIK Diversion dispersal total of 1,090,318 STRV. On December 1, 2000, the CCC transferred title to 277,349 tons of refined crystalline sugar to participating producers, or their assignees. The acres diverted from production by the PIK program represent about 7 percent of acreage planted to sugar beets. Domestic deliveries of HFCS-42 in 2000 are forecast at 3.562 million tons, dry basis, about the same level as last year. Domestic deliveries of HFCS-55 are forecast at 5.603 million tons, dry basis, about 30,000 tons less than last year. These levels represent a major downturn for the industry. If current trends continue, it is likely that total HFCS deliveries for 2001 will be only 1.1 percent higher than levels in 2000. Prospects for HFCS-42 are for little to no growth, while HFCS-55 deliveries may grow as much as 1.7 percent. Much will depend on developments in the soft drink industry where sales growth diminished from the high levels from most of the 1990's. U.S. SUGAR On January 11, 2001, the U.S. Department of Agriculture (USDA) released its latest supply and use estimates for fiscal year (FY) 2000 and projections for FY 2001. Year Ended, FY 2000 Production Crop campaigns ended in Louisiana, Florida, and Texas for a combined total of 3.701 million short tons, raw value (STRV) through August. The Louisiana 2000/01 campaign began in mid-September due to a large projected crop and added an additional 42,520 STRV to the estimate for cane sugar production for FY 2000. The Texas 2000/01 harvest added 460 tons to the September total. The Hawaiian 2000 crop campaign began in February and totaled 250,800 through September. Because the Hawaiian campaign extends through December, the October-December portion of the 1999 campaign, totaling 66,820 STRV, is counted in FY 2000. Together with the 4,240 STRV produced in Puerto Rico, the cane sugar estimate for FY 2000 is a record 4.065 million STRV. Beet sugar production for FY 2000 finished strongly, and is currently estimated at a record 4.976 million STRV. Beet sugar production continued throughout the year due to recovery of sugar from stored thick juice and from molasses. Sugar production from molasses finished a record year more than 35 percent higher than the previous year. The crop year saw 31,586,306 tons of beets sliced, or 94.5 percent of total beets harvested. This percentage is about 1.7 percentage points below the average for the previous 5 years, indicating that a large number of beets harvested did not get sliced. The recovery rate was a record 294 pounds per sliced ton, net of desugaring, indicating high sucrose content and very good quality beets. Also, a large anticipated crop for FY 2001 caused the campaign to start more strongly at an earlier date. September 2000 beet sugar production is estimated at 376,910 STRV, more than 6 percent higher than September 1999. Trade and Deliveries Based on preliminary data, FY 2000 tariff-rate quota (TRQ) imports totaled 1.124 million STRV. The TRQ shortfall is estimated at about 43,994 STRV. Non-TRQ imports are estimated at 512,000 STRV. Non-TRQ components include sugar imported for the Refined Sugar Re-export Program, and the Sugar Containing Products Re-export Program (total of 373,000 STRV), and for the Polyhydric Alcohol Program (15,000 STRV). Sugar extracted from imported sugar syrups classified under HTS 1702.90.40 at a low duty contributed 118,000 STRV. High-tier tariff imports, largely from Mexico, totaled 6,000 STRV. Sugar exports under the Refined Sugar Re-export Program are estimated to have been 124,000 STRV. Total deliveries for domestic food and beverage uses, re-exported products, polyhydric alcohol, and feed uses are estimated at 10.111 million STRV, only marginally above the previous year's total. Compared with the previous year, beet processor deliveries increased 1.2 percent, while cane sugar refiners' deliveries remained about the same. Actual deliveries had been very close to projected levels through June; however, deliveries in the last FY quarter (July-September) were about 86,000 STRV lower than projected levels. During this same period, processor forfeitures of sugar under loan to the Commodity Credit Corporation (CCC) totaled 155,408 STRV (44,940 STRV on August 1 and 110,469 STRV on September 1). Ending Stocks and Miscellaneous Adjustments Ending stocks are estimated at 2.219 million STRV. The CCC held 296,649 STRV, which included the forfeitures in August and September, and sugar purchased by the CCC in June (141,240 STRV). Privately-held stocks constituted the remainder. The ending stocks-to-use ratio is estimated at a historically high level of 22.0 percent. The miscellaneous category proved to be an important component of sugar supply and utilization for FY 2000. Until complete Sweetener Market Data (SMD) become available in November, the ending stocks are estimated as the residual of total supply (beginning stocks, production, and imports) less total use (exports and total deliveries). Once actual stockholding data are available, the difference between calculated stocks and reported stocks is listed under the miscellaneous category in USDA's monthly World Agricultural Supply and Demand Estimates (WASDE) report. The SMD reports a miscellaneous category as well, but it can differ from the WASDE because SMD uses company data for imports rather than the U.S. Customs Service data. The miscellaneous estimate is high at -137,000 STRV in the WASDE and -149,532 in the SMD. These amounts represent additions to sugar supplies or subtractions from sugar use, which causes the estimate for ending stocks to be higher than expected. In the SMD, the miscellaneous category includes refining losses (generally recorded as a positive number), shipments less receipts of sugar by processors and refiners to other processors and refiners, and inventory adjustments. For FY 2000, shipments less receipts equaled -106,453 STRV and refining losses were 120,175 STRV, implying a yearly inventory adjustment of 106,453 STRV. These levels are much higher than the average ending year levels since FY 1995. Shipments, less receipts, for FY 2000 is 2.1 times the average; the refining loss is 2.0 times the average; and the inventory adjustment is 2.7 times as much. Moreover, figure 1 shows the difference between residual and actual stocks summed for the previous 12 months from the dates shown on the horizontal axis since 1993. The expectation is that shipments, less receipts, and inventory adjustments would average out to zero over time, leaving a positively valued refining loss. As can be seen, this expectation was met prior to FY 1997, but only 6 times out of a possible 50 since FY 1996. The value of the miscellaneous category was the lowest (most negative) at the end of FY 2000. Prices U. S. raw sugar prices (nearby futures, C.I.F., duty-paid, Contract No. 14, New York) averaged 20.56 cents per pound during December, down 83 points from the November average, but up 269 points from a year earlier. March No. 14 futures prices rallied to 21.27 cents a pound in mid-January, 0.6 cent above a month earlier, and 3.57 cents above January 2000. As of January 19, March No. 14 futures prices averaged 20.94 cents per pound, with a maximum and minimum of 21.38 cents and 20.21 cents, respectively. Fears of freeze damage to the Florida sugarcane crop, an 80,000-ton reduction in FY 2001 sugarbeet production, large quantities of sugar in government inventory, and uncertainty about entry procedures for Mexico's 116,000-ton TRQ allocation have boosted futures prices in the past month. World raw sugar prices, which have been in steady decline since mid-1998, averaged 10.23 cents per pound during December, 4.23 cents a pound above a year ago. As of January 19, the world raw sugar price averaged 10.73 cents a pound, with a maximum of 10.99 cents a pound and a minimum of 10.54 cents a pound. Although world prices have shown strength recently, futures prices are likely to decline in the next few months due to a cutback in Russian sugar imports, an expected rebound in 2001/02 world sugar production, slackening world sugar demand, increasing global sugar export availabilities, lower sugar imports by the United States, and large 2000/01 world sugar carryover stocks. Current Year, FY 2001 Cane Sugar Production Cane sugar production for FY 2001 is projected at 4.168 million STRV, about 2.5 percent above the estimated total for FY 2000. Production increases in Florida, Texas, and Puerto Rico are expected to more than offset declines in Louisiana and Hawaii. Florida cane sugar production is projected at 2.130 million STRV. The National Agricultural Statistics Service (NASS) estimates Florida sugarcane acreage harvested for sugar and seed at 455,000 acres, a decrease of 5,000 acres from last year. Of this total, NASS estimates that 437,000 acres are for sugar. NASS estimates sugarcane for sugar yield at 37.0 tons per acre, and sugarcane for sugar production at 16.169 million tons. This year's estimate is 664,000 tons more than last year. Sugar recovery to date has been at a record high, implying that the season-average recovery rate will likely approach 13.2 percent. Louisiana cane sugar production is projected at 1.570 million STRV. NASS estimates Louisiana sugarcane area harvested for sugar and seed at 500,000 acres, an increase of 35,000 acres over last year. Of this total, NASS estimates that 465,000 acres are for sugar. This expansion has occurred in northern and western producing areas and has been induced by continuing low prices for alternative crops such as soybeans, corn, and rice. NASS estimates Louisiana sugarcane yields of 30.0 tons per acre, a decrease of 2.7 tons per acre from last year. Sugarcane production is estimated at 13.950 million tons, a decrease of 275,000 tons from last year. Crop growing conditions were difficult due to a continuing lack of adequate moisture. Although the sugarcane harvest beginning in September has progressed at a normal pace, muddy conditions since mid-October have made progress difficult. Sugar recovery is expected to be 5 percent lower than last year. Texas cane sugar production is projected at 180,000 STRV. NASS estimates Texas area of sugarcane for sugar and seed at 46,600 acres, up 15,600 acres from last year. Of this total, 46,000 acres are estimated for sugar. Yield is estimated at 37.7 tons per acre and the crop is estimated at 1.734 million STRV. Hawaii cane sugar production for FY 2001 is projected at 265,000 STRV, a projected decrease of more than 50,000 STRV from FY 2000. In September 2000, AMFAC/JMB announced that it plans to close its remaining sugar plantation at Lihue by the end of the year and sell its 18,000 acres on Kauai to pay its debts and focus on other business opportunities. All its agricultural operations on Kauai were to cease by December 31, 2000. With the departure of AMFAC/JMB, Gay and Robinson is the sole sugar producer on Kauai. In October, the Hawaii Board of Land and Resources issued temporary permits allowing a consortium of farm companies, including Gay and Robinson, to take control of 27,000 acres of State land that AMFAC/JMB had been leasing. Gay and Robinson hopes to farm about 7,000 acres of the land that had been in sugarcane for AMFAC/JMB's Kekaha Plantation. Although the Kekaha Plantation was closed last year, sugarcane formerly milled there was being trucked to the Lihue Plantation on the other side of the island. The Hawaii Sugar and Commercial Company (HSC) closed its Pa'ia Mill on Maui in the fall. HSC cited depressed sugar prices and an extended drought as determining factors in the decision to close the facility. HSC continues processing at its larger Pu'unene Mill, in which substantial investments have been made during the last 2 years. The drought that is now in its third year may affect production in the coming year. Puerto Rico sugar production for FY 2001 is projected at 22,500 STRV. It is expected that the two sugar mills on the island will be back in production this year. Sugarcane harvested area is expected to be about 11,900 acres, with plans for as much as 300,000 tons of sugarcane. Beet Sugar Production and the PIK Diversion Program On August 1, 2000, USDA announced the Payment-in-Kind (PIK) Diversion Program to alleviate the excess supply of sugar and reduce CCC program costs. Producers were offered the choice of not harvesting some of their crop in exchange for sugar held by the CCC. USDA gave farmers a 2-week sign-up period beginning August 21 for the PIK. Growers put in bids as to how many tons of refined sugar they would accept from government stocks in return for not harvesting a certain number of acres. The USDA announced that it would not accept bids that represented more than the farm's average sugar production over the past 3 years. Working with processors, producers wishing to participate in the program submitted a bid to CCC for acreage under contract for delivery to a processor. Acreage accepted into the program could not be harvested for sugar or used for any other commercial purposes in return for payments in sugar from the CCC's inventory of refined sugar. The CCC ranked the bids on the basis of the bid amount as a percentage of the bid cap for those acres, with the lowest percentages to be selected first if CCC stocks were less than sugar wanted by bidders. The PIK Diversion Program succeeded in diverting 101,832.9 acres of sugarbeets from 2000-crop production. The CCC transferred title to 277,349 tons of refined crystalline sugar (296,763 STRV) to participating producers, or their assignees on December 1. The acres diverted from production equaled about 7 percent of acreage planted to sugar beets. Transfer of this sugar resulted in about a $555,000-reduction in monthly CCC storage-related outlays. The sugar transferred from CCC inventory also represented about 7 percent of the expected FY 2001 domestic sugar production from sugarbeets. The USDA made its forecast of sugarbeet and beet sugar production taking into account the PIK Diversion Program in the October 2000 Crop Production report and the October WASDE. NASS reduced its forecast of sugarbeet area harvested by 101,600 acres, or about 7 percent less than the September 1 forecast; and NASS cut its production forecast by 2.030 million tons, or about 6 percent lower than its September 1 forecast. National yield was forecast upward by 0.2 ton per acre. NASS forecast the largest production reduction in Minnesota of 1.237 million tons, or 60.9 percent of the total reduction. Other major producing States in which production forecasts were reduced include North Dakota (418,000 tons), Idaho (214,000 tons), and Colorado (111,000 tons). The production of beet sugar projected in the October WASDE was reduced by 320,000 STRV to 4.350 million STRV from the September WASDE, a drop of 6.9 percent. Beet Sugar Production: Current Projection Beet sugar production for FY 2001 is currently projected at 4.370 million STRV. NASS estimates sugarbeet production at 32.521 million tons, about 3 percent lower than last year. NASS estimates area harvested at 1.378 million acres, a 10-percent decrease from last year largely attributable to the PIK Diversion Program. The yield estimate is 23.6 tons per acre, a record. Good yields were due to a long growing season with above-normal temperatures and adequate water supplies. Yields in Idaho and California reached record levels, although harvests progressed more slowly due to the size of the crops and, in Idaho, bad weather in October and November. Good growing and harvesting conditions prevailed in the Red River Valley. Dry weather and favorable temperatures have left the crop in generally good condition for stockpiling prior to processing, but some frozen beets were stockpiled that may rot before they can be processed. NASS sugarbeet estimates relate to the year of intended harvest except for overwintered spring-planted beets in California. Implications of the NASS sugarbeet estimate for FY 2001 beet sugar production have to be adjusted because California beet processing plants in Woodland and Tracy are out of operation in 2001. The USDA Interagency Commodity Estimates Committee (ICEC) for sugar has projected that an additional 29,000 acres will be lost for sugarbeet production in California in FY 2001. Applying the 1997-99 average sugar yield for these areas of 3.8 tons per acre, FY 2001 beet sugar production should be reduced 110,000 STRV beyond what otherwise would have been the case. Based on historical relationships, it is projected that beets sliced during the current crop year will exceed 30.4 million tons or about 3.7 percent less than last year. Beets sliced September through December equaled 13,273,323 tons, for a rate that is 2.8 percentage points above the trend for the crop year 1995 through 2000. Based on to-date sugar recovery rates and their average relationship with respect to the final sugar recovery rate, it is projected that the final recovery rate net of sugar from molasses will equal 270 pounds per sliced ton, down 24 pounds from last year. This rate, along with a projection for sugar recovered from molasses, implies sugar production for FY 2001 of 4.37 million STRV or about 12 percent less than last year. Non-TRQ Imports Sugar imports outside the sugar TRQ for FY 2001 are projected to total 515,000 STRV, including 365,000 STRV under the combined Refined Sugar Re-export Program, the Sugar-Containing Products Program, and the Polyhydric Alcohol Program. This projection includes 50,000 tons granted to the C&H Sugar Company for FY 2001 as a result of a Settlement Waiver that was meant to facilitate the company's removal of 100,000 tons of sugar from the U.S. market. The company is permitted to import up to 50,000 tons of raw sugar during each of FY 2001 and 2002, provided it first surrenders Certificates for Quota Eligibility (CQE) to the USDA for an equivalent amount of sugar. This requirement implies that projections of sugar entering under the sugar TRQ are reduced by the same amount of sugar that C&H Sugar Company is authorized to import (up to 50,000 tons) under the Refined Sugar Re-export Program. The USDA projects an increase in U.S. sugar supply of 125,000 STRV as a result of sugar syrup imports under HTS 1702.90.40, up 7,000 STRV from FY 2000. High-tier tariff sugar imports for FY 2001 are projected to increase to 25,000 STRV, up from 6,000 STRV in FY 2000. Most of this sugar is expected to be Mexican refined sugar presently held in U.S. Customs bond. On January 1, 2001, the high-tier North American Free Trade Agreement (NAFTA) tariff for refined sugar dropped 1.60 cents to 11.21 cents a pound. TRQ Imports On September 15, 2000, the USDA established the FY 2001 TRQ for imports of sugar at 1,360,983 metric tons, raw value (MTRV), or 1,500,227 STRV. The total includes a quantity for raw sugar of 1,117,195 MTRV, the minimum level to which the United States is committed under the World Trade Organization (WTO); a quantity for refined sugar of 38,000 MTRV; a required quantity of 105,788 MTRV for Mexico under the NAFTA Side-Letter Agreement that may be shipped as raw or refined sugar; and a reserve of 100,000 MTRV, to be allocated, if needed, contingent on developments in international markets. USDA has made all except the contingent quantity available to the U.S. Trade Representative (USTR) for allocation for import into the U.S. customs territory. Pursuant to the Harmonized Tariff Schedule of the United States, USDA has established the FY 2001 raw sugar TRQ at 1,222,983 MTRV, or 1,348,108 STRV. This sum is equal to the WTO minimum access quantity of 1,117,195 MTRV plus Mexico's residual allocation of 105,788 MTRV. Mexico's total raw sugar allocation is equal to 113,046 MTRV, the sum of its WTO raw sugar minimum access (7,258 MTRV) and the residual allocation quantity. USDA established the FY 2001 refined sugar TRQ at 143,788 MTRV, or 158,499 STRV. This sum is equal to Mexico's NAFTA commitment of 105,788 MTRV plus 38,000 MTRV. The refined sugar TRQ includes the specialty sugar allocation, a subset of the refined sugar TRQ, at 17,656 MTRV. Canada is allocated 10,300 MTRV as a result of an agreement reached with that country. In addition, 2,954 MTRV of refined sugar is allocated to Mexico. Therefore, Mexico's total share of the refined sugar TRQ is 108,742 MTRV. The remainder of the refined sugar TRQ is made available on a first-come, first-served basis. TRQ sugar imports are currently projected at 1.275 million STRV. The sum of the raw and refined sugar TRQs adjusted to eliminate double counting of NAFTA sugar that can be imported as either raw or refined sugar is 1.390 million STRV. The shortfall from countries that are unlikely to be able to ship their allocations to the United States is projected at 65,000 STRV. As described above, the C&H Sugar Company is projected to surrender CQE equivalent to 50,000 STRV to the USDA. Subtracting these sums from 1.390 million STRV produces the current projection of 1.275 million STRV. As of January 8, 2001, sugar imports under the TRQ have amounted to 317,703 STRV, or about 25 percent of the amount projected to enter for FY 2001. Refined in-quota imports from Mexico have totaled 3,751 STRV. Entry procedures for Mexico's NAFTA allocation have not yet been determined. Exports and Deliveries Sugar exports reflect shipments to foreign destinations made under the Refined Sugar Re-export Program. These exports for FY 2001 are projected at 175,000 STRV. Deliveries to domestic food and other products manufacturers under the Sugar-Containing Products Re-export Program are projected at 125,000 STRV. Deliveries for the Polyhydric Alcohol Program are projected at 15,000 STRV. Total deliveries for FY 2001 are projected at 10.385 million STRV. After netting out deliveries made for the Sugar-Containing Products and Polyhydric Alcohol Programs, along with deliveries for livestock feeding (20,000 STRV), domestic food and beverage deliveries are projected at 10.225 million, about 2.3 percent higher than FY 2000. Deliveries for October and November totaled 1.852 million STRV. Although deliveries for the first two months of FY 2001 are only 0.6 percent higher than the same period last year, last year's deliveries for October and November were much higher than had been expected. The early pace of deliveries this year is in the higher range predicted by the USDA. Cane sugar deliveries are 5 percent above last year's to-date levels, while beet sugar deliveries are 5 percent below. End use delivery data indicate strong growth in non-industrial user demand, about 6.5 percent higher than the same period last year. Non-industrial users tend to demand cane sugar over beet sugar, thereby explaining most of the growth in cane sugar deliveries. Although industrial demand for sugar has tended to grow more strongly than non-industrial demand in the past, the pattern in the first two months of FY 2001 is the opposite: industrial demand is 3.7 percent lower than in last year's first two months. Because industrial demand is more evenly split between beet and cane sugar, its negative growth mostly accounts for lackluster beet sugar deliveries. Ending Stocks Ending stocks are currently projected at 1.987 million STRV, for an ending stocks-to-use ratio of 18.8 percent. Of the total, the CCC owns 793,555 STRV or 39.9 percent of total projected ending stocks. The ratio of privately-held ending stocks-to-use is projected at 12.5 percent. The CCC acquired 793,669 STRV in October as a result of loan forfeitures by beet processors in several States (499,322 tons actual weight or 534,274 STRV) and by Florida cane processors (247,145 tons actual weight or 259,395 STRV). Adding this amount to what the CCC had already owned (296,649 STRV) gave them a pre-PIK Diversion dispersal total of 1,090,318 STRV. On December 1, 2000, the CCC transferred refined beet sugar amounting to 296,763 STRV under the PIK Diversion Program. Their remaining inventory includes 483,878 STRV of refined beet sugar and 309,676 STRV of cane sugar. Box Text IMPORTS OF CERTAIN SUGAR SYRUPS RISE DRAMATICALLY IN THE 1990'S Most imported raw cane sugar, refined sugar, and sugar syrups enter the United States under either the raw cane sugar or refined sugar tariff-rate quotas (TRQs). Since 1995, however, certain sugar syrups containing molasses have entered the United States under the HTS 1702.90.40 tariff line, which is outside the sugar TRQs. Imports of these sugar syrups have contributed to greater U.S. supplies of sugar and have pressured TRQ levels. In May 1995, the U.S. Customs Service, in response to a request from a sugar refiner in Michigan, determined that a type of sugar/molasses syrup should be classified under the tariff category, HTS 1702.90.40. The syrup is produced outside the United States by adding raw granular sugar to molasses and water and heating the mixture to dissolve the sugar. Once in the United States, the syrup is refined into liquid sugar, which is used in breakfast cereals and other food products. Imports of the sugar syrup described in the previous paragraph began to increase after the Michigan refinery came on line in mid-1997. Prior to the May 1995 ruling, imports under the 1702.90.40 tariff line were of juices derived from raw sugarcane with 6 percent or less molasses. Imports under the tariff category, which came from Central American or Caribbean countries, were less than 20,000 tons annually. Imports under the 1702.90.40 category rose from less than 1,000 short tons raw value (STRV) in FY 1994/95 to 118,000 STRV in 1999/2000 -- the equivalent of 10.5 percent of imports under the raw and refined sugar TRQs in 1999/2000. Imports of the sugar/molasses syrup are assessed much lower duties than are imports of sugar syrups subject to the high-tier (over-quota) tariff of the refined sugar TRQ. Most of the sugar syrup entering the United States under 1702.90.40 is processed in a plant in Ontario, Canada, from imported raw sugar. Since 1995, raw sugar imports from a number of countries have been used to make the molasses/sugar syrup. Examples include Australia, Brazil, the Dominican Republic, and, most recently, Colombia. Some of the syrup also is produced in Mexico. In FY 1997/98 and FY 1999/2000, between 3,000 STRV and 4,000 STRV of the syrup were imported from Mexico. The same quantity has been imported from Mexico in the first two months of FY 2001 alone. As imports of the sugar/molasses syrup increased through the late 1990's, they began to expand U.S. sugar supplies. When imports under tariff line 1702.90.40 topped 83,000 STRV in FY 1997/98, the U.S. Department of Agriculture (USDA) began to include the product in its estimates of non-TRQ sugar imports. USDA forecasts FY 2001 imports of the sugar/molasses syrup at 125,000 STRV, up 7,000 STRV from actual FY 2000 imports of 118,000 STRV. Sugar producers are concerned that other firms will begin producing and refining the sugar/molasses syrup, further adding to already large U.S. sugar supplies. Some countries that hold shares of the U.S. raw cane sugar TRQ agree with the U.S. producers' concerns because increasing imports of the sugar/molasses syrup expand U.S. sugar supplies, erode quota rents, and reduce the size of the raw cane sugar TRQ when import needs exceed the World Trade Organization (WTO) minimum. In January 1998 and again in January 1999, the United Sugar Cane Refiners Association and the United States Sugar Beet Association jointly petitioned the U.S. Customs Service to revoke its May 1995 classification ruling for the sugar/molasses syrup. Following a lengthy administrative process which allowed comments from all interested individuals and companies, the Customs Service announced that it was revoking the May 1995 ruling on September 8, 1999 for implementation in November 1999. By revoking its original ruling, the Customs Service recognized the product as a sugar syrup which would be subject to the quota restrictions and higher over-quota tariff rates of the refined sugar TRQ. The U.S. company affected by the U.S. Customs Service ruling filed a complaint against the Customs Service in the U.S. Court of International Trade (C.I.T.). On October 19, 1999, the U.S. C.I.T. decided in favor of the company and blocked the Customs Service's September 8 ruling. The Customs Service appealed the U.S. C.I.T. decision to a higher court on March 31, 2000, and it is expected that the case will be heard shortly. In addition, Canada initiated consultations with the United States under the WTO dispute settlement procedures in October 1999, but stopped the dispute settlement procedure after the U.S. C.I.T. blocked the U.S. Customs Service September 8, 1999, ruling. In 2000, legislators in the U.S. House and Senate introduced bills that would add the HTS 1702.90.40 tariff line to the list of tariff lines included under the refined sugar TRQ and would extend the list to any other products from which sugar can be extracted or which can be used to circumvent the sugar TRQs. The legislation was not approved in 2000, but it is likely that the same legislators will introduce similar legislation in 2001. U.S. imports of sugar/molasses syrup under HTS 1702.90.40 Import volume metric tons Import value Import volume Year commercial weight $1,000 STRV 1994/95 1,864 295.2 1,025 1995/96 14,648 1,137.9 8,056 1996/97 38,324 11,048.8 21,078 1997/98 151,382 33,540.9 83,260 1998/99 208,537 38,415.2 114,695 1999/00 214,735 34,002.1 118,105 Source: U.S. Census Bureau import data. Import volumes in metric tons, commercial weight, are multiplied by 0.55 to derive imports of sugar in short tons, raw value. HIGH FRUCTOSE CORN SYRUP (HFCS) Domestic deliveries of HFCS-42 in 2000 are forecast at 3.562 million tons, dry basis, about the same level as last year. Domestic deliveries of HFCS-55 are forecast at 5.603 million tons, dry basis, about 30,000 tons less than last year. These levels represent a major downturn for the industry. Yearly growth of HFCS-55 deliveries from 1993 to 1999 averaged about 280,000 tons. Deliveries of HFCS-55 began leveling off last year when growth was only 35,000 tons above 1998 levels. Yearly growth of HFCS-42 deliveries from 1993 to 1999 averaged 107,000 tons. Although HFCS-42 grew at an accelerated pace in 1999, most of this growth was price-induced substitution away from relatively higher priced glucose and HFCS-55. The same factors affecting HFCS-55, and perhaps sugar as well, caught up with HFCS-42 this year. Figure 2 shows consumption growth of HFCS and sugar since 1993. HFCS growth has consistently outpaced sugar growth until 2000 when growth in both sectors was negative for the calendar year. While current projections of fiscal year growth for domestic sugar deliveries are above 2 percent, growth in HFCS may not rebound as strongly. If current trends continue, it is likely that total HFCS deliveries for 2001 will be only 1.1 percent higher than levels in 2000. Prospects for HFCS-42 are for little to no growth, while HFCS-55 deliveries may grow as much as 1.7 percent. Much will depend on developments in the soft drink industry where sales growth diminished from the high levels from most of the 1990's. The soft drink industry accounts for about 90 percent of domestic demand for HFCS-55 and about 44 percent of domestic demand for HFCS-42. The industry has reportedly abandoned volume marketing in favor of concentrating on higher price margins. Efforts of the HFCS industry to negotiate higher sales prices appear not to have been successful. Figure 3 shows a comparison of producer price indices for the soft drink and HFCS industries. The soft drink index increased 5.6 percentage points in 2000 for 4.3 percent growth, which is about twice the average yearly growth since 1982. Although the HFCS index has increased marginally in both 1999 (6.4 percent) and 2000 (2.7 percent), these increases are very small relative to the declines in 1997 (12.4 percent) and 1998 (30.0 percent). The HFCS price decline shown in the figure is largely attributable to increases in HFCS producing capacity. Entry of new firms into the industry and the expansion of facilities by established producers increased capacity from 8.1 million tons, dry basis in 1994 to 12.4 million tons, dry basis in 1997, an increase of 54 percent. Capacity usage fell from 96 percent in 1994 to under 75 percent in 1997. Figure 4 shows estimates of capacity usage since 1994. Domestic demand growth and industry rationalization have improved capacity usage since 1997. Although HFCS production in 2001 is not expected to be strong, the capacity usage ratio may reach 86 percent, as reports indicate a major HFCS producer in North Dakota is ceasing production as of mid-January. HFCS and Mexico In June 1997, Mexico enacted a provisional antidumping measure on U.S. exports of HFCS. The measure was made final in January 1998. After consultations with Mexican officials in May 1998, the United States referred its complaint to a World Trade Organization (WTO) dispute settlement panel in October 1998. On January 20, 2000, the WTO dispute settlement panel determined that Mexico's antidumping measure on U.S. exports of HFCS was inconsistent with the WTO Antidumping Agreement. The panel agreed with the U.S. position that Mexico did not properly establish injury as required in the Antidumping Agreement. Although Mexico had the right to appeal the ruling, it did not do so. However, on September 20, 2000, Mexico released its revision of the 'original' final resolution. The revision contains an analysis of economic factors that was missing in the previous analysis, and confirmed that there was a threat to the Mexican sugar industry from the importation of HFCS from the United States. The same final countervailing duties from January 1998 were confirmed in the analysis. The compensatory duties are listed in table T-1. The only bright note for the United States was that preliminary duties paid from June 1997 to January 1998, prior to the 'original' final resolution, were to be returned to the U.S. companies. U.S. exports of HFCS-55 and above (HTS 1702.60) have been between 75 and 78 percent of the previous year's level. The U.S. Customs Service reports a January - September level of 126,974 metric tons (mt), down from 162,143 mt last year for the same period. Mexico's Secretariat of Commerce and Industrial Development (SECOFI) reports a higher level of HFCS-55 shipments but reduced from last year's levels through September: 200,461 mt compared with 268,428 mt last year. As can be seen in figure 5, discrepancies between U.S. and Mexican sources of estimates of the level of HFCS-55 shipments continue to persist. Estimates of export levels from the U.S. Customs Service are 44 to 65 percent lower than the corresponding estimates from SECOFI. Estimates of HFCS production in Mexico range between 300,000 and 360,000 mt, dry basis. Current estimates of HFCS consumption as a percentage of total sweetener consumption are in the 10 to 13 percent range. It is not likely that the HFCS share of the Mexican sweetener market will grow until the various sweetener disputes with the United States are settled. SPECIAL ARTICLE U.S. Sugar Price Indices and Stocks-to-Use Ratios: Accounting for A Volatile Fiscal Year 2000 Stephen Haley 1/ ----- ----- 1/Agricultural economist in the Market and Trade Economics Division, Economic Research Service, USDA. ----- Abstract: Fiscal year 2000 was a very volatile year for the U.S. sugar market. Pricing relationships were called into question as U.S. sugar supplies reached high levels that eventually resulted in record forfeitures of sugar under loan to the Commodity Credit Corporation. This article examines the effect of the year's events on the relationship between stocks-to-use ratios and sugar pricing. The article extends sugar pricing analysis to the refined beet sugar price published by the Milling and Baking News and several sugar producer price indices published by the Bureau of Labor Statistics. The article also extends pricing analysis to all quarters of the fiscal year instead of focusing solely on the fourth quarter. Raw sugar price results indicate that a 1-standard deviation change in quarterly stocks-to-use is accompanied by a -0.76 standard deviation change in the raw sugar price. The effect of FY 2000 has been to increase the response coefficient from the -0.51 estimated through FY 1999 by 44 percent. Results indicate that the refined cane sugar response has been unaffected by FY 2000. Refined beet sugar pricing response has increased by 32 percent to -0.67. Ongoing analysis involves whether the relationships relevant for FY 2000 will continue on into FY 2001 and beyond. Keywords: Sugar, sugar prices, stocks-to-use ratio, baseline. Economists in the U.S. Department of Agriculture (USDA) are concerned with the relationship between sugar prices and sugar stocks-to-use ratios at the end of the fiscal year (FY). Excluding the period since the third quarter of FY 1999, economic analysis of times series of stocks and price data has revealed a more or less constant correlation between the ending stocks-to-use ratio and the average FY fourth-quarter nearby futures price of the raw sugar No.14 Contract listed on the New York Coffee, Sugar, and Cocoa Exchange (CSCE). This relationship has been useful in assessing the probability of forfeitures to the Commodity Credit Corporation (CCC) of sugar pledged in the USDA sugar loan program. The relationship has also been a key component of USDA's longterm projection of sugar prices that are part of the USDA sugar baseline. The correlation between the ending stocks-to-use and the No.14 raw sugar price became suspect starting in the fourth quarter of FY 1999. According to an econometric model used up until that time, an ending stocks-to-use ratio of 16 percent would have implied an average fourth-quarter No. 14 raw sugar price of 22.19 cents a pound. Although the actual average was 21.32 cents a pound (only 0.87 cent a pound below the predicted value), the within-quarter monthly averages were decreasing from 22.61 cents a pound in July, to 21.24 cents a pound in August, and 20.10 cents a pound in September. Average monthly prices dropped to the 16-17 cents a pound range from November 1999 to February 2000, although the same-period projected ending stocks-to-use ratios in USDA's World Agricultural Supply and Demand Estimates (WASDE) report were in the 16.0-17.6 percent range. The FY 2000 ending stocks-to-use ratio, including stocks owned by the Commodity Credit Corporation (CCC), ended up at 22.0 percent. The model would have predicted an average fourth-quarter price of 19.96 cents a pound, but the actual average price was 18.24 cents a pound. In retrospect, FY 2000 was a very volatile year for the U.S. sugar market. Many pricing relationships were called into question and the year saw record forfeitures of sugar under loan to the CCC. The goal of this article is to measure the effect of the year's events on the relationship between stocks-to-use ratios and sugar pricing. The article extends analysis to additional sugar price indices besides the nearby No.14 futures price of U.S. raw sugar. These include the refined beet sugar price published by the Milling and Baking News and several sugar producer price indices (PPI) published by the Bureau of Labor Statistics (BLS). The article also extends pricing analysis to quarterly stocks-to-use ratios. A part of this analysis examines relationships between sugar prices and deseasonalized sugar stock levels. An ongoing part of the analysis started here will involve whether the relationships relevant for FY 2000 will continue on into FY 2001 and beyond. Fourth-Quarter Sugar Prices and Ending Stocks-to-Use Ratios Analysts at the USDA's Economic Research Service (ERS) reported in the March 1996 edition of the Sugar and Sweetener Situation and Outlook that there was a measurable relationship between the ending stocks-to-use ratio reported in the WASDE and the nearby No. 14 raw sugar futures price for the fourth quarter of the fiscal year (July-September). The reported coefficient was -0.361, indicating that a 1-point increase in the ending stocks-to-use ratio is typically accompanied by a 0.361 cent decrease in the average No.14 raw sugar price. 2/------- 2/------- This equation was re-estimated for this article, and the coefficient value was estimated at -0.357, which is slightly but not significantly, different than the value originally published. -------- The relationship was estimated covering FY 1986 through FY 1995, and over 60 percent of the variation in the fourth quarter FY raw sugar price was accounted for. This analysis was updated in the May 1998 edition, covering the period from 1982 through 1997. The price response coefficient was estimated at -0.374, very close to the earlier value. A difference from the earlier estimation was that indicator variables were included for the years 1992 and 1994. The indicator variables adjusted the estimated equation's intercept coefficient, meaning that the ending stocks-to-use ratios for 1992 and 1994 were excluded in estimating the price responsiveness (or slope) coefficient. The primary effect of including the indicator variables was to increase the percentage of explained variance in the raw sugar price accountable to changes in the ending stocks-to-use ratio to 87.7 percent. Similar but less successful efforts have been made to determine a relationship between refined beet sugar prices and ending stocks-to-use ratios. The refined beet sugar price used in the analysis is not a futures price like the raw sugar price but is the low end of a spot price range for the U.S. Midwest reported in the publication Milling and Baking News (M&B). Analysis covering FY 1982 through FY 1999 indicated a statistically significant response coefficient of -0.474. However, only 20.7 percent of the variation in the fourth quarter refined beet sugar price can be explained by the relationship. Extending the analysis into FY 2000 shows that historical relationships were apparently no longer relevant. The raw sugar price began its descent in August 1999, i.e., during the fourth quarter of FY 1999. The refined beet sugar price, on the other hand, did not show a significant decrease until the first quarter of FY 2000 in December. Table 1 shows several equations that include and exclude observations for FY 2000. (Equations 3 and 4 show results discussed above for the raw sugar price, and equation 7 shows results discussed above for the refined beet sugar price.) Equations 1 and 2 provide alternative raw sugar price specifications for FY 1999 and 2000. The specification in the first equation forces the effects of 1999 and 2000 into a slope coefficient adjustment, whereas the specification of the second equation forces the effects into an intercept adjustment, similar to what was done for 1992 and 1994 in equation 3, discussed above. The value of having two years that are different from the other years (as opposed to just one year) is that there are criteria to judge which specification (slope versus intercept) is better. One indicator is the adjusted R-squared (R2), which measures the percentage of explained variation in the dependent variable (i.e., the raw sugar price). Two other indicators are the Akaike Indicator Criterion (AIC) and Schwarz Indicator Criterion (SIC). Both of these indicators are measures of one-step ahead prediction error variances. In judging which of two (or more) specifications has greater predictive power, equations with the lower AIC or SIC are considered better predictors than equations with higher AIC or SIC. Comparing equations 1 and 2 reveals that equation 1 with the slope specification has a higher adjusted R2 of 0.866 than the 0.839 of equation 2. Comparing the AIC and SIC suggests that the slope specification is better than the intercept specification: 1.397 versus 1.584 for the AIC, and 1.645 versus 1.833 for the SIC. Accepting these results implies that the price responsiveness has increased from -0.402 for the period prior to FY 1999 to -0.514 for the period after FY 1998, an increase of 28 percent. Equation 5 shows results for the slope specification for the refined beet sugar price. Because this price did not show a decrease until FY 2000, there is no way to discern whether a slope or intercept specification produces greater predictive power. In both cases, the adjusted R2, the AIC, and the SIC will be exactly the same. In the alternative specification, the intercept coefficient would be equal to the slope coefficient (-0.293) times the ending stocks-to-use ratio for FY 2000. If the slope specification is assumed correct, however, then the response coefficient grew from -0.474 prior to FY 2000 to -0.767 after FY 1999, an increase of 62 percent. In any event, FY 2000 made a difference for refined sugar pricing. Equation no. 6 shows a specification where no special account is made for FY 2000. The estimate for the slope coefficient for the entire sample period jumps from -0.474 (equations 5 and 7) to -0.654. Only 27.7 percent of refined beet price variation is explained, compared with 49.9 percent in equation 5. The AIC and SIC of equation 5 are much smaller as well: 4.103 and 4.252 compared with 4.424 and 4.524. The analysis presented thus far relies on only one observation being available for each fiscal year. If quarterly stocks-to-use ratios were used, one would have four times the number of observations from which to estimate relationships. A problem is that stock levels vary from one quarter to the next. The fourth quarter is the one immediately prior to the start of the new harvesting and processing season when stocks are at their lowest level for the year. Second-quarter stocks (measured at the end of March) are the highest because, with minor exceptions, the sugarbeet harvest is over and the sugarcane harvest nearly so. Use of differing stocks-to-use ratios within the same year require some analytical modifications. Two approaches are attempted. Both involve transforming price and stocks-to-use ratios into units that strip away the units in which they are measured. Price and stocks-to-use ratios used in both cases are adjusted by their standard deviations before the estimation of relationships. The estimated coefficients are technically referred to as 'beta' coefficients, whose values measure the standard deviation change of the dependent (price) variable due to a 1-standard deviation change in an independent variable (stocks-to-use ratio). The first approach involves the separate estimation of relationships for each quarter, and then tests the equality of corresponding beta coefficients across quarters. The hypothesis being tested is that differences in corresponding beta coefficients across equations cannot be significantly distinguished from zero. The second approach analyzes price changes as a function of current and lagged responses to changes in deseasonalized stocks-to-use ratio series. Deseasonalization involves standardization of quarterly stock levels in terms comparable with fourth-quarter levels. Results are presented in terms of beta coefficients to permit comparison with the first approach. Producer Price Indices for Sugar The Bureau of Labor Statistics publishes producer price indices for raw cane sugar, refined cane sugar, and refined beet sugar. These series are compiled on the basis of surveys completed by firms engaged in sugar milling, processing, and refining. These series permit additional testing of hypothesized relationships between sugar pricing changes and changes in sugar stock-to-use ratios. In the case of refined cane sugar, there are no agreed-upon, market-generated price series available to analysts. Also, because of the tenuous ('not quoted on a futures exchange') nature of the widely used Milling and Baking News refined beet sugar spot price, the beet sugar PPI provides a valuable alternative for testing of pricing relationships. Figure 1 shows a charting of the refined beet sugar PPI and the Milling and Baking News spot price indexed the same way as the PPI (second-quarter 1982=100.0). The series are plainly vertically displaced but seem to follow a similar pattern over time. The correlation between the series is 0.861. Figure 2 shows a similar pairing of the raw cane sugar PPI and the nearby No.14 raw sugar futures price. Displacement seems less of an issue and the correlation between these series is somewhat higher at 0.909. Single-Quarter Analysis Table 2 shows regression results organized first by the price variable and then by the fiscal quarter. Initially, each equation specified the price as a function of the quarter's stocks-to-use ratio, a slope coefficient adjustment to account for FY 2000, a constant, and a trend. In cases where serial correlation was a problem for estimation, a first-order autoregressive term was added. Underneath the coefficients are t-statistics -- values above 1.96 indicate a statistically significant relationship. T-statistics below 1.96 mean that the coefficient measuring the strength of the hypothesized relationship cannot be distinguished from zero. Coefficients on the stocks-to-use variable are almost all significantly different from zero, the only exceptions being for the second and fourth quarters of the refined beet sugar PPI equations. The adjustment coefficients for FY 2000 are all statistically significant for the raw sugar price equations in all quarters, but none of the adjustment coefficients for the refined sugar PPI are significant. Only the fourth quarter adjustment coefficient in the refined beet sugar equations are significant. Adjusted R2 are fairly high for the raw and refined cane equations, but not nearly as high for almost all the refined beet sugar equations. Within each price grouping, Wald statistical tests are performed to test the joint hypothesis that stocks-to-use coefficients are equal across the four quarters, and that the FY 2000 adjustment coefficients are equal across the four quarters. 3/------ ------ 3/ The Wald test computes a test statistic by estimating regression equations without imposing coefficient restrictions. The test statistic measures how close the unrestricted estimates come to satisfying the hypothesized restrictions on the coefficients. ------ None of these hypotheses could be rejected. The equations were re-estimated with equality restrictions placed on the stocks-to-use and adjustment coefficients within each price index grouping across all quarters. Table 3 shows the coefficient and t-statistic results. The raw sugar price results imply that a 1-standard deviation change in a quarterly stocks-to-use ratio relative to the value in the same period of the previous year is accompanied by a -0.70 standard deviation change in the raw sugar price relative to the price in the same period of the previous year. The effect of FY 2000 has been to increase response coefficient from the -0.50 estimated through FY 1999 by 40 percent. Results indicate that the refined cane sugar response has been unaffected by FY 2000. A 1-standard deviation change in stocks-to-use is accompanied on average by a -0.22 standard deviation change in the refined cane sugar price. The refined beet sugar price is affected by stocks-to-use changes in the response range of -0.59 and -0.67. The effect of FY 2000 has been modest (10-15 percent). Times Series Analysis An advantage to time series analysis is the ability to test relationships from quarter-to-quarter rather than year-to-year for individual quarters. A problem, already addressed above, is that quarterly stock series display a cyclical pattern. The remedy pursued above was standardize by transforming variables with reference to their standard deviations. Although that procedure can be followed here as well, further corrective adjustments deseasonalizing the stocks data may be desirable, given the expectation of lagged adjustments of prices to stocks-to-use changes. Procedure for Deseasonalizing Quarterly Sugar Stocks The goal is to achieve a degree of uniformity across quarterly stock levels that allows for direct comparisons of stock levels. The deseasonalization process adjusts stock levels in the first three quarters of the fiscal year that make them directly comparable with the fourth-quarter stocks level. The fourth quarter is the period where stock levels for the year are at their lowest, immediately prior to the next year's harvesting and processing season. The fourth-quarter stocks level (typically expressed as a ratio relative to total sugar use) is used as a measure of the sugar supply-use balance for the year and is well accepted as a measure inversely related with sugar prices. The first step to the procedure is to remove stocks owned by the CCC. Because government-owned stocks are not immediately available to the market without explicit policy intervention, they form an exogenous component that is unlikely correlated with a seasonal pattern for which the procedure is trying to adjust. At the last step in the procedure, the CCC stocks are added back to the deseasonalized market stocks series to assure that all stocks are being accounted for. The second step is to calculate a yearly averaged stock level centered about each observed stock level. This calculation consists in averaging each observed market stock level with market stock levels for the preceding two quarters and for the succeeding quarter. The third step is to compute the ratio of each observed market stock level to the yearly average market stock level centered about it. In the fourth step, the average value of the ratios is calculated for each quarterly grouping. The resulting first-quarter average for market sugar stocks for the period 1982-2000 is 1.160. The interpretation is that average first-quarter market stock levels are on average 16 percent higher than corresponding yearly stock levels. The ratios for the second, third, and fourth quarters are 1.334, 0.961, and 0.549, respectively. At this point, one could adjust the stock levels relative to the averaged yearly levels in order to have direct comparability. However, because the goal is to express comparability in terms of the fourth quarter, the next step is to express the ratios of the first, second, and third quarters relative to the ratio of the fourth quarter. This is accomplished by dividing the first, second, and third quarter ratios (i.e.,1.160, 1.334, and 0.961. respectively) by the fourth quarter ratio (0.549). These values are 2.113 for the first quarter, 2.430 for the second quarter, 1.750 for the third quarter, and 1.000 for the fourth quarter. These coefficients are divided into the appropriate quarterly stocks series. After adding back the CCC stocks, the result is the desired deseasonalized quarterly stock series. Figure 3 shows actual and deseasonalized quarterly stock levels for FY 1982 through FY 2000. The figure highlights especially the growth in stocks since the FY 1999 fourth quarter. Also evident has been an upward trend in stocks beginning in the third quarter of FY 1996: quarterly deseasonalized stock levels have exceeded the previous quarter's level in 13 of the 17 quarters. Time Series Results The time series specification allows price changes to be functions of current and lagged values of quarterly stocks-to-use ratios, FY 2000 adjustment factors, and trends. The length of each equation's lag structure is made on the basis of which specification has the lowest SIC, indicating which has the best predictive ability among alternative lag lengths. Results are shown in table 4 for the 5 price series. Statistical results are strong for all equations. The adjusted R2 are all above 0.80, and the sums of the current period and lagged stocks-to-use coefficients are statistically significant. The raw sugar equations support the hypothesis that events in FY 2000 have changed the relationship between the raw sugar price and stocks-to-use ratios. The adjustment coefficients are strongly significant and somewhat higher in value than the single-quarter results. The responsiveness range that includes FY 2000 is between -0.62 and -0.86, which are wider than the single-quarter range (-0.68 to -0.73). The refined cane sugar equation produces a result similar to the single-quarter result. The adjustment coefficients are insignificantly different from zero, while the sum of stocks-to-use coefficients (lagged 3 quarters) equals -0.24, which is very close to the -0.22 of the single-quarter analysis. The beet sugar equations indicate strong FY 2000 effects, with stocks-to-use effects being 28 to 47 percent higher than the pre-FY 2000 effects of about -0.50. Unlike the other equations, lagged FY 2000 effects are significant. Summary Figure 4 illustrates raw sugar price results from the three approaches. The first is the traditional FY fourth-quarter analysis, presented in terms of beta coefficients to allow comparison with the other two approaches. As can be seen, the fourth-quarter analysis produces the largest adjustment coefficients for pre-FY 2000 and post-FY 1999 periods. In all cases, the absolute value of the increase in the stocks-to-use adjustment coefficient is 0.20 or above. The average raw sugar price standard deviation adjustment across all equations due to stocks-to-use changes is -0.76, up an average of 44 percent over the pre-FY 2000 period. Figure 5 shows the refined beet sugar price results. The average price standard deviation adjustment across all equations due to stocks-to-use changes is -0.67, up an average of 32 percent over the pre-FY 2000 period. Knowledge of these coefficient values will be of assistance to USDA economists concerned with policy tradeoffs and longterm baseline projections. Given the likely volatility of sugar markets in the future, the relationship between sugar pricing and stocks-to-use ratios will continue to be of research interest. One implication of current research is that sugar prices are seemingly more flexible with respect to stocks-to-use ratio changes. This increased flexibility might imply greater uncertainty regarding predictions of forfeitures of sugar under loan to the CCC because small stocks-to-use changes may become associated with large swings in sugar prices. END_OF_FILE