Prepared by Prudential Securities, Inc.

New-crop production prospects for several major origins, particularly Brazil and Colombia, have brightened appreciably since our last Quarterly Outlook was issued in September, putting coffee prices under considerable downside pressure during September and October. The downmove pushed March futures, which were trading near $1.75 per pound in early September, to the $1.40 area by early November, a loss of roughly 20%. Improved availability prospects also were reflected in the spreads action, with the March premium versus May narrowing to about 2.5 cents from near 7 cents.

Coffee prices firmed in November, reflecting crop uncertainties in parts of Central America. In Guatemala, coffee officials have lowered their earlier 1997/98 export projections by roughly 10% due to adverse weather conditions. Coffee trade representatives in El Salvador expect to see weather-induced quality problems, while growers throughout the region anticipate that harvesting peaks will be delayed by several weeks because of this year's unusual weather patterns.

Changing availability perceptions have had a marked effect on the New York/London price relationship. Whereas arabica tightness concerns had caused the New York premium to widen through much of the summer, the subsequent lessening of those concerns has made for a narrower, more normal premium. Renewed premium strength during November was partly due to the unsettled Central American production outlook as well as expectations of beneficial rains in Indonesia (the leading robusta producer). Tree flowering in most of Brazil has been excellent. Some observers expect Brazil to have its largest harvest in many years, with projections for the 1998/99 crop roughly between 30 million and 35 million bags. This compares with the USDA's estimate of 28 million bags for 1997/98. In Colombia, the National Coffee Growers Federation is tentatively looking for a 1997/98 harvest of 12.0-12.5 million bags, up from the year-ago outturn of about 10.8 million. Colombia's production prospects will remain a pivotal price-making factor for the market, especially given the uncertainty surrounding the El Nino's impact on the country's coffee areas.

El Nino-related crop losses are expected to be most severe in Indonesia, where 1997/98 output could be down by more than 1 million bags from the previous year. (This downturn in robusta output will be partly offset by expanded Vietnamese production.) Other origins whose 1997/98 output is likely to be reduced by El Nino-induced weather anomalies are Papua New Guinea, Philippines and Ecuador. The U.S. Climate Prediction Center expects the current El Nino to persist until March-May 1998, and its impact on global coffee output remains a major market factor as we go to press.

U.S. green coffee imports during the first nine months of the year reached their highest level since 1992, not only facilitating a modest, steady U.S. stocks buildup in 1997, but also seemingly arresting the long-range stocks decline that began in early 1993. This improvement on the supply side appears to be taking place for world import markets as a whole, with the Association of Coffee Producing Countries (ACPC) noting in a recent survey that overall importer availability may have expanded by about 4.5 million bags during the 12 months ended in June.

In June, the USDA forecast global 1997/98 output at 103.7 million bags, up from the previous season's output of 100.7 million. November data released by the Hamburg-based trade firm Bernhard Rothfos and the German commodity publishing firm F.O. Licht had 1997/98 coffee output falling below the year-ago level. Rothfos projected 1997/98 production at 98.5 million bags, down from its figure of 100.2 million in 1996/97, while Licht expects global production of 98.8 million bags versus its year-earlier estimate of 102.4 million. The major reason for the discrepancy between the USDA's data and projections released by the other sources is that the latter anticipate a sharp downturn in Brazil's 1997/98 output. We believe it is likely that the USDA's revised global data (expected to be released in December) will reflect a reduced Brazil production forecast.

Production data for the world's 12 leading origins, accounting for roughly 80% of total coffee output, are shown in Table 1, which includes a comparison of the USDA's June 1997/98 production forecasts with more recent projections released by Rothfos.

For some parts of the world, the impact of the ongoing El Nino phenomenon has heightened the production uncertainty that typically accompanies any growing season. The most extreme impact is likely to be felt in Indonesia, the world's largest robusta producer, where protracted dry conditions are expected to lead to a significant production drop in 1997/98 versus the previous year. Given Indonesia's pivotal role as a robusta producer, the ACPC recently projected global 1997/98 robusta output at 31.5 million bags, down from 1996/97 production of roughly 34 million bags. In the past, some trade sources have accused Indonesian farmers and shippers of "hoarding" coffee, but the country's recent currency crash presumably will result in a more aggressive marketing policy. Other countries in Asia likely to see production fall in 1997/98 as a result of El Nino-related dry conditions include the Philippines and Papua New Guinea.

Vietnam continues to see a long-term, government-supported production boom that could enable it to overtake Indonesia, its chief regional rival, as the world's leading robusta origin. Traders anticipate that enhanced availability and currency depreciation will boost Vietnam's 1997/98 exports by about 5% over the year-ago level.

Thanks to ideal weather conditions, coffee specialists in India are projecting a bumper crop in 1997/98, which some local sources expect to be up 15%-20% versus a year ago. Roughly 70% of the country's crop originates in the state of Karnataka, where harvesting usually commences in late November. The other important coffee-growing state, Tamil Nadu, opened its harvest in early November.

Production is expected to fall in the 1997/98 season in Ivory Coast and Uganda, two of Africa's (and the world's) leading origins. In Ivory Coast, dry weather in January and February hurt crop development, and low government-controlled producer prices have generally undermined farmer interest in coffee production. Citing the need for enhanced economic efficiency, the World Bank and other foreign aid donors have made "liberalization" of the economy a condition for continued debt relief and other forms of aid. A principal donor demand is that the Ivory Coast government withdraw from commodity marketing, and 1998/99 supposedly will be the first season in which coffee exporting will be controlled by the private sector. In Uganda, according to local sources, drought and coffee wilt disease threaten to push coffee output below the year-ago level, but the downturn may be short-lived given expanded plantings of high-yielding crop varieties.

Hurricane Pauline struck Mexico in October, causing considerable damage to the agricultural sector; Hurricane Rick caused additional damage. Preliminary indications pointed to a loss of 100,000-150,000 bags of coffee in the states of Chiapas and Oaxaca; more recently, producer interests raised their crop-loss estimate to the 400,000 bag level. Mexico's 1996/97 output is pegged near 5.6 million bags, and both the USDA and Rothfos expect output in the new season to come in at about the same level. Mexican governmental sources have released similar forecasts for 1997/98, but they have been rejected as "unrealistic" by the National Coffee Producers' Union, which is looking for a harvest of 5.1 million bags. Given that Mexico is the leading coffee supplier to the U.S. market, any significant change in its production profile can be expected to have an impact on the New York futures market.

Hurricanes Pauline and Rick also caused production worries in Central America, with farmers in some areas reporting high cherry losses due to heavy rains and strong winds. In El Salvador, an unofficial survey found that perhaps 3%-4% of the 1997/98 crop might have been lost, and trade sources warned in mid-November that year-end shipments might be affected by quality problems. In Guatemala, the loss was initially put at about 2%, but has been raised to the 10% area. The rainfall not only ended the protracted El Nino-related dry conditions that had slowed crop maturation but also is likely to postpone harvesting peaks in some areas by several weeks, to January. The high international prices seen earlier this year have eased the financial pressures on the region's exporters, thus, they have the option to export 1997/98 coffee at a more relaxed pace than might otherwise have been the case.

Although three months have elapsed since our last Quarterly Outlook, the market does not seem to have come much closer to reaching a consensus regarding the size of Brazil's 1997/98 crop. The USDA's June forecast of 28 million bags, criticized by Brazilian (and some non-Brazilian) sources as being too high, gained credence in early October when it became known that a former U.S. agricultural attache in Brazil, now working as a private consultant, was projecting the country's 1997/98 crop at 28.2 million bags (down from his earlier forecast of 28.6 million). Other sources, however, have released substantially lower forecasts. The London-based ACPC is looking for a 24-million-bag harvest, while the Brazilian Association of Coffee Exporters (ABECAFE) stated in mid-September that the crop would come in near 22.4 million bags. The vice president of Brazil's National Coffee Council believes the crop will be 19 million bags, which, it was claimed, would make it impossible for Brazil to meet its 1997/98 export quota of 15 million bags without hefty government-controlled stock sales.

Agricultural officials for the state of Espirito Santo, which produces roughly 65% of Brazil's robusta crop, recently estimated the state's 1997/98 overall output at 3.4 million bags, down from 4.9 million in the previous season. The 1997/98 robusta harvest was pegged at 2.4 million bags versus 3.3 million in 1996/97.

While the market's focus is shifting to 1998/99 production prospects, the USDA's release of a 1997/98 update (expected in December) will be awaited with keen interest by numerous market participants. For the time being, we are using 26 million bags as our working figure for the current season.

Although parts of Brazil have been facing El Nino-related weather problems (excessive rainfall to the south of the coffee belt and very dry conditions in northeastern areas of the country), these anomalies appear not to have affected most parts of the coffee belt, and overall growing conditions generally have been favorable. Indeed, weather conditions in some areas have been so beneficial that agronomists working in south Minas Gerais reported that tree flowering was the best ever seen by local growers. These initial indications point toward a substantially improved production scenario, reinforced by the fact that 1998/99 will be the "strong" year in coffee's two-year bearing cycle. Recently, a representative of the Brazilian National Coffee Council unofficially projected the country's 1998/99 (May/April) crop at around 30 million bags, while a speaker at a Licht-sponsored international coffee seminar stated the crop could be as high as 35 million bags. Brazil has not harvested more than 30 million bags of coffee since 1990/91. A major production boost would be expected to lead to a more aggressive marketing posture, which would exert a negative price influence on the world coffee market.

Indications of a potential bumper crop in 1998/99 also come from another quarter. Manufacturers of agricultural machinery have seen sales pick up sharply earlier than usual, which they attribute to farmers' expectation of a very large coffee harvest. For instance, sales of certain types of drying equipment usually do not rise until December or January, but this year manufacturers saw sales going up in September and October. The fact that this machinery entails substantial expenditures may be regarded as a measure of farmer confidence in the size of the next crop.

One likely exception to Brazil's improved production prospects is the state of Espirito Santo, which accounts for about 65% of the country's robusta harvest. Severe September drought conditions have prompted local coffee officials to lower their earlier 1998/99 robusta projection of about 4 million bags by approximately 30%.

Presumably because of the controversy that has frequently surrounded Brazil's coffee production forecasts, the Brazilian government has agreed to coordinate an effort (via Embrapa, a governmental research entity) to produce an "official" 1998/99 crop forecast. The forecast will incorporate the work of various participants, including university personnel and state agricultural officials. In order to lend the forecast as much credibility as possible, the research methodology is to be released to the press. We have not heard when the forecast will be released.

During the first half of 1997, exports were running strongly ahead of last year's pace, but since then export performance has become more mixed, with shipments falling below the year-ago level in three out of four months. We anticipate that exports for November and December will be reported near last year's levels.

Brazil's ABECAFE, which until recently was projecting 1997 exports at 16.5 million bags, has lowered this figure to 15 million, which we view as a realistic target and would represent an increase of roughly 18% versus 1996 exports. ABECAFE's reduced export projection reflects the group's expectation of a relatively small 1997/98 production of 22.4 million bags. According to ABECAFE's calculations, global demand for Brazilian coffee exceeds current-season output by about 5 million bags, and the group has pressed the government to increase the frequency of authorized stock auctions in order to help boost availability.

In September, ABECAFE requested that the New York Coffee, Sugar & Cocoa Exchange (CSCE) modify its coffee contract so that washed arabicas from Brazil would be accepted for delivery against futures. This matter is apparently under review. ABECAFE officials have noted that Brazil currently produces about 1 million bags of washed arabicas, making it a larger producer than several other origins recognized in the contract. Should the exchange act in the affirmative, it would be the first time that Brazilian coffee has been included in the New York "C" coffee futures contract.

The USDA's forecast of Colombian 1997/98 (October-September) production is 11.3 million bags, about 1 million bags above the previous season's output. However, these production numbers are below those released by Colombia's National Coffee Growers Federation, which has pegged the 1996/97 crop at 10.8 million bags and has tentatively projected 1997/98 output at 12.0-12.5 million.

It is interesting to note that, while various Colombian coffee specialists anticipate that the 1997/98 harvest will be adversely affected by El Nino-related weather anomalies, both the USDA and the Federation are projecting 1997/98 output to exceed the year-ago level. The reason is that more attractive farmer prices and various governmental incentives are expected to lead to more efficient farming and harvesting practices, which will tend to offset some of the potential weather-related damage.

In October, the Colombian National Coffee Growers Federation, which had been looking for October-December production to reach 4.5 million bags, lowered this figure to 4 million following its reassessment of El Nino-related dry conditions in July and August. Meanwhile, a spokesman for the private exporters has pointed out that in many areas the main El Nino impact will not be on production, but on quality and marketing conditions. Local weather watchers believe that insufficient rainfall remains the chief threat to the coffee crop, and that El Nino's most severe impact will not be felt until the first few months of 1998. There can be little doubt that the El Nino's impact on Colombia's crop outlook will be one of the most closely watched market factors over the next few months and has the potential to induce substantial price volatility over an extended period.

Colombia's National Coffee Growers Federation has reported 1996/97 exports at 11.2 million bags, about 4.4% above the year-ago figure. (End-season stocks were reported at 4.3 million bags, representing an inventory drawdown of 1.8 million bags from a year ago.) Preliminary data released by a Federation spokesman showed October exports at 740,000 bags, well below September shipments of 930,000 and the year-ago level of 1.03 million. In a move intended to make Colombian coffee exports "more competitive.in a market that is no longer a sellers market," the 30-cent-per-pound premium added to the official export registration price was reduced to 16 cents on October 6 (December futures settled 7.80 cents lower the next day). Government officials emphasized that this step would not decrease coffee revenues because the exporters' enhanced competitiveness would enable them to reach the 3.5-million-bag export target for the October-December quarter.

The ACPC, whose membership accounts for about 75% of global coffee output, decided in September to extend its export restriction program for two more years, i.e., the 1997/98 and 1998/99 seasons that begin in July; the existing 12-month export target of 52.75 million bags was left unchanged. Given the Association's incomplete membership, voluntary compliance, lack of an independent verification mechanism and absence of penalties for cheating, traders can be forgiven if they regard the ACPC's export plan with some skepticism. It can be argued that the existence of the ACPC's export guidelines provides some psychological support of the futures market. Ultimately, however, we believe that export levels will be dictated by coffee availability and the self-interest of the individual origins rather than by an international export agreement.

Table 3 gives the Association's export breakdown for 1997/98, with the largest allotment (15 million bags) going to Brazil; Colombia is in second place, with 10.65 million bags. Representatives of several origins, however, maintained that availability problems made it unlikely that specified export limits would be met, countering with reduced "expected" export targets. Indonesia, a participant in the 1997/98 export plan, has declined to commit itself with regard to 1998/99 membership, but presumably will clarify its position at the ACPC's April 1998 meeting at which details regarding 1998/99 exports are to be worked out.

Click here to view Table 3
1997/98 Export Limits As Proposed By The ACPC

At an October meeting in El Salvador, the ACPC's Central American group (consisting of Costa Rica, El Salvador, Honduras, Nicaragua and Ecuador) declared that El Nino-related crop problems in the region might make it difficult to honor all contractual export commitments. Insisting that the region's harvest "is not going to be what the market wants," and that a "message" had to be sent to the market, the group reduced its overall export allocation of 8.92 million bags, to 8.57 million. Of this, 3.45 million bags are to be exported during July-December, with the individual country allocations as follows: Costa Rica, 1.2 million bags; El Salvador, 826,000 bags; Honduras, 506,000 bags; Nicaragua, 142,000 bags; and Ecuador, 780,000 bags. However, several delegates noted that even the reduced export target might not be met given slow crop progress in some member countries (e.g., Honduras and Nicaragua).

At its September meeting, the ACPC presented an analysis of its 1996/97 export targets versus actual shipments. The overall target of 52.75 million bags had been exceeded by 3.78 million bags, with African members the chief violators (accounting for about 51% of total overshipments); Brazil was close behind. The ACPC's secretary general noted that member nations had overshot their export ceilings "without any harm to the price levels (and) that governments decided to do this at a moment that the market needed this coffee." We assume that future overshipments will be justified or explained in similar terms.

Table 5 compares global exports for the 1995/96 and 1996/97 seasons as well as September 1996 and 1997 exports. For two coffee categories, Robustas and Brazil & Other Arabicas, the 1996/97 export level was well up on the previous season. A decline in shipments for the other two categories (Colombian Milds and Other Milds) in 1996/97 versus the previous season was an important supportive factor for futures prices during 1997.

Click here to view Table 5
World Coffee Exports

Brazil, the world's largest producer, was able to boost 1996/97 shipments by almost 6 million bags over the previous season, whereas Colombian exports increased by only about 400,000 bags. Mexico, which has displaced Brazil as the leading source of coffee for U.S. importers, saw 1996/97 exports fall versus 1995/96. Indonesia, the world's foremost robusta exporter, reduced shipments last year (a supportive factor for London futures), although this was offset by Vietnam's increased marketings. Monthly exports, particularly by Other Milds producers in Latin America, will be among the market's most closely watched statistics over the next several months. The alarm is already being sounded in one important growing region, Central America, where adverse weather conditions have slowed bean maturation and harvesting; it is expected that harvesting peaks in some areas could be delayed several weeks. This condition suggests relatively low export availability into early 1998, which is likely to act as a price-supportive factor.

Following the high annual import levels of the early 1990s, green coffee imports into the United States crashed in 1994, falling almost 20% below the previous year's level. The following year saw a moderate recovery, which picked up momentum in 1996 and remains on track in 1997, with imports 11.1% ahead of last year's pace through the first three quarters and at their highest level since 1992. The uptrend reflects greater export availability following two seasons (1994/95 and 1995/96) of exceptionally low global shipments, and has enabled the U.S. coffee trade to begin rebuilding inventories, albeit modestly, from their low point in 1996. Roasted coffee imports were up sharply in 1996 versus 1995, and continue to trend upward, with January-September shipments reported at 19 million kilograms, about 25% ahead of last year's comparable level.

Table 8 shows the disposition of coffee stocks controlled by consumers versus producers since 1995 as well as a forecast of end- 1997/98 stocks distribution. Consumer-held stocks have risen modestly for three years to 12.4 million bags in 1997 from 10.1 million in 1995. Rothfos is forecasting a continuation of this trend, with consumer-held stocks expected to expand to 13.5 million bags by October 1998.

Click here to view Table 8
Distribution Of Global Coffee Stocks

The growth in consumer inventories has been uneven, with most of the expansion taking place in robustas (from about 1 million bags in 1995 to a projected 4.1 million in 1998). Mild arabica supplies (which are of far greater relevance to the New York futures market) are projected at 6.6 million bags in 1998, roughly 10% below the year-ago level and 5% less than in 1995. The continued relative tightness in consumer-held Milds types can be viewed as a long-range supportive factor for the futures market.

Conversely, total producer stocks have fallen nearly 40% since 1995, and are projected to decline further in 1998, to 17.8 million bags, or 13% below the 1997 level. Most of this reduction is attributable to the drawdown in the unwashed arabica category.

Consumers have gained increasing control over stocks. In 1995, roughly 23% of total world stocks were in their hands. By 1997, the consumer share of overall stocks grew to 38% and the figure is expected to rise to 43% in 1998. While producer nations obviously have an inherent interest in selling their coffee, it can be assumed they do not welcome a trend in which a relatively larger proportion of global stocks comes under the control of consumer interests.

After peaking at 10.0 million bags in March 1993, U.S. monthly coffee stocks fell with little interruption until the end of 1996. This year has seen a modest pick up, with inventories first pulling ahead of last year's level in July. End-October stocks were reported at 2 million bags, up about 60% from the year-earlier level. We anticipate that monthly inventories will remain above the year-ago level for the next three to four months, but do not expect to see any major stocks buildup on an absolute basis.

Average monthly U.S. retail prices for roasted coffee began advancing sharply in March, overtaking year-ago values in April and continuing to surge through August. The upward trend stalled in September, with retail prices finally coming under the influence of the sharp drop in green coffee prices that began in May. We look for further price softening at the retail level, but anticipate that average monthly values will remain above year-ago prices through the first quarter of 1998.

Figures 4-7 illustrate movements in the cash/futures differential for four coffees available for "prompt shipment" in the New York market. Two of these, Prime (washed) Mexicans (Figure 4) and Prime (washed) Guatemalas (Figure 5) are among the deliverable growths used by the CSCE to establish the "basis." Also, because these coffees come from adjacent countries, they frequently are affected by the same set of weather conditions. As one might expect, their cash/futures spreads tend to move in similar fashion. This year, both spreads moved to a discount during June-September, but firmed in October, perhaps due to lower availability, concern for the new crop outlook and the possibility of harvesting delays. The Colombian MAMs differential (Figure 6) remained on the plus side during the summer, with the cash premium widening in September and October. The "odd man out" is Santos 4 from Brazil (which is not recognized as a deliverable growth by CSCE), whose cash price has traded at a discount to futures throughout the year, although it narrowed slightly in September and October (Figure 7).

Click here to view Figure 5
New York Guatemala (Prime Washed) Differential

Click here to view Figure 6
New York Columbia (MAMs) Differential

Click here to view Figure 7
New York Brazil (Santos 4) Differential

Futures prices tend to advance from late January or early February into May as traders position themselves in anticipation of crop-threatening frost conditions in the Brazilian coffee belt. (Brazil's winter runs June-August, with July the most frost-prone month.) Contrary to what one might expect, futures prices tend to decline through the Brazilian winter, with traders increasingly discounting the weather threat. (This assumes there is no frost. In the event of adverse weather conditions, and depending on the extent of the crop loss, values will move higher.) After moving to a seasonal low near the end of July, prices tend to show a modest recovery and then move sideways for the rest of the year.

Coffee futures prices have recovered from their October lows and remain firm as we go to press. The interrelated supportive factors include: (1) forecasts by several sources that world 1997/98 production will fall below the year-ago level; (2) concern that several Latin American arabica origins (which are of particular relevance to the New York futures market) could see weather-related crop losses exceed earlier expectations; (3) a sluggish Brazilian export pace that has seen shipments below the year-ago level in three of the last four months; (4) very slow stock accumulation in the consumer markets, with overall consumer stock levels still low by the standards of several years ago (U.S. end-October stocks were reported at 2 million bags, which compares with 2.2 million in 1995 and 4.8 million in 1994) and (5) seasonal considerations. We believe that these factors are not yet discounted and that they will provide market support over the next three to four months.

The ongoing El Nino, and its final impact on global coffee output, remains an important market factor. We expect that El Nino-related 1997/98 crop losses will be relatively greater for the robusta sector than for arabicas, with Indonesia the prime candidate for substantial crop damage. (This factor has been a supportive for London coffee futures.) Two other important robusta areas expected to see output down sharply in 1997/98 (although not necessarily due to El Nino) are Ivory Coast and the Brazilian state of Espirito Santo.

El Nino-induced dry conditions in Colombia have caused officials with the country's National Coffee Growers Federation to lower their October-December crop expectations by 0.5 million bags to 4.0 million. Overall 1997/98 output is tentatively forecast at 12.0-12.5 million bags versus the year-ago production of 10.8 million. Any further decline in 1997/98 crop prospects would be a bullish signal for coffee futures prices.

Central America is another very sensitive region for New York coffee futures. The El Nino phenomenon has made for very unusual weather patterns this year, and it now appears that the region's 1997/98 production will fall short of earlier expectations, with some countries also experiencing quality problems. It seems generally expected that harvesting peaks will be postponed by several weeks, implying that there could be some export disruptions, which is potentially supportive for coffee futures.

With March coffee futures near $1.63, they likely will attract increased aggressive origin activity, making upside price progress more labored. However, in view of the various bullish factors acting on the market, we look for prices to remain generally firm over the next three to four months. We anticipate that any price setbacks will hold at about the $1.40 level, and look for an upside target in the $1.80-$1.90 area.

December 4, 1997Arthur Stevenson

Prudential Securities, Inc.

One New York Plaza, New York, New York

Consensus National Futures and Financial On Line Index

Added to the WWW 12-12-97
Last updated on 12-12-97

Hosted by:
One Crossroads Place
610 West Maple Ave, Suite WWW
Independence, MO 64050
(816) 252-4080
sysop@kcmo.com

wmeubank@ocp.kcmo.com