2006 Report on Ethanol Market Concentration December…

2006 Report on Ethanol Market Concentration




             December 1, 2006



         Federal Trade Commission
                        2006 Report on Ethanol Market Concentration



       Under Section 1501(a)(2) of the Energy Policy Act of 2005, as codified at 42 U.S.C. §

7545(o), the Federal Trade Commission ("Commission" or "FTC") must annually "perform a

market concentration analysis of the ethanol production industry using the Herfindahl-Hirschman

Index to determine whether there is sufficient competition among industry participants to avoid

price-setting and other anticompetitive behavior." The statute further directs the Commission to

consider marketing arrangements among industry participants in rendering its analysis. The

Commission must report its findings to Congress and to the Administrator of the Environmental

Protection Agency. This Report presents the Commission's concentration analysis of ethanol

production for 2006.

       Based on publicly available data and on interviews with ethanol producers and marketers,

FTC staff concluded that current U.S. ethanol production is not highly concentrated. Assuming

that U.S. fuel ethanol production is a relevant market for competition analysis, our best

measurement of concentration (based on production capacity) yields Herfindahl-Hirschman

Indices ("HHIs") between 326 and 995, depending on the degree to which we attribute individual

ethanol producers' shares to the firms that market their ethanol. These levels represent a drop of

21 to 35 percent from comparable figures presented in last year's report. Under the Horizontal

Merger Guidelines that the Commission and the U.S. Department of Justice use to assess the

competitive effects of mergers, these HHIs indicate an unconcentrated market. Indeed, our

results may overstate these concentration levels, for reasons outlined below. In any event, the

level of concentration in ethanol production would not justify a presumption that a single firm, or
a small group of firms, could wield sufficient market power to set prices or coordinate on prices

or output.

        I.       Background

        This analysis builds upon the factual background contained in the Commission's 2005

Report on Ethanol Market Concentration, which detailed important characteristics of domestic

fuel ethanol production and marketing. 1

        The growth in domestic ethanol production over recent years has been well-documented.

In 2005, U.S. ethanol plants produced 3.9 billion gallons of ethanol, up 15 percent from 3.4

billion gallons in 2004. 2 It is estimated that in 2006 U.S. ethanol production will exceed 4.6

billion gallons, at least 18 percent greater than 2005 production, and at least 188 percent greater

than the 1.6 billion gallons produced in 2000. 3

        Increases in ethanol production reflected comparable growth in U.S. ethanol production

capacity. By the end of 2005, there were 95 facilities with a total of over 4.3 billion gallons per

year of ethanol production capacity. 4 This represents a capacity increase of over 779 million

gallons per year during 2005 alone, as producers built and began producing from 14 new




1
 See FTC, REPORT ON ETHANOL MARKET CONCENTRATION (2005), available at
http://www.ftc.gov/reports/ethanol05/20051202ethanolmarket.pdf.
2
 See Energy Info. Admin. ("EIA"), U.S. Dep't of Energy, Petroleum Navigator ­ U.S. Oxygenate Production,
available at http://tonto.eia.doe.gov/dnav/pet/pet_pnp_oxy_dc_nus_mbbl_m.htm; EIA, EIA-819 Monthly Oxygenate
Report (December 2004), available at
http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/monthly_oxygenate_report/historical/2005/2005_0
2/pdf/819mhilt.pdf.
3
  See EIA, Petroleum Navigator ­ U.S. Oxygenate Production, available at
http://tonto.eia.doe.gov/dnav/pet/pet_pnp_oxy_dc_nus_mbbl_m.htm; EIA, EIA-819M Monthly Oxygenate
Telephone Report (December 2000), available at
http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/monthly_oxygenate_report/historical/2001/2001_0
1/pdf/oxydata.pdf. Our estimate of 2006 ethanol production is a full-year projection based on actual production
from the first seven months of 2006.
4
 See RENEWABLE FUELS ASS'N, FROM NICHE TO NATION: ETHANOL INDUSTRY OUTLOOK 2006, at 2 [hereinafter
FROM NICHE TO NATION], available at http://www.ethanolrfa.org/objects/pdf/outlook/outlook_2006.pdf.


                                                       2
facilities during the year. 5 Production capacity continued to expand in 2006 as well. As of mid-

October 2006, the Renewable Fuels Association ("RFA") estimated that current domestic ethanol

production capacity exceeded 5 billion gallons per year, a 16 percent increase compared to the

end of 2005. 6 Industry expansion likely will continue into the foreseeable future, as the RFA

estimated that producers are currently adding 3.5 billion gallons per year of production capacity

at 45 new plants and expansions at 7 existing facilities. 7

            The number of participants in ethanol production increased as well. By mid-October

2006, 90 different firms operated ethanol plants, a one-year increase of roughly 15 firms. New

firms are entering ethanol production, and we estimate that 110 firms will operate ethanol plants

by the end of 2007. As new entrants emerge, the largest producer's share of capacity continues

to fall. Currently, the largest producer accounts for 21 percent of domestic ethanol production

capacity, down from 25 percent in 2005 and over 40 percent in 2000.

            II.     Analysis

            Section 1501(a)(2) instructs the Commission to measure concentration in ethanol

production by using HHIs. The Commission and the U.S. Department of Justice use HHIs to

help assess the likely effects of a merger or acquisition on competition in a relevant antitrust

market. 8 The HHI is calculated by summing the squares of the individual market shares of all




5
    See id. at 2.
6
 See Renewable Fuels Ass'n, Industry Statistics ­ Ethanol Biorefinery Locations (estimate as of Oct. 16, 2006),
available at http://www.ethanolrfa.org/industry/locations/.
7
 See id. As described below, our analysis uses a narrower slice of the expected capacity growth identified by RFA.
Our approach considers as competitively significant only those capacity additions that are likely to result in
marketable volumes within the next year. Our estimate (conducted as of mid-June 2006) identified 2.5 billion
gallons per year of additional capacity likely to come on-line by the end of 2007.
8
 See U.S. DEP'T OF JUSTICE & FTC, HORIZONTAL MERGER GUIDELINES (1992, revised 1997) [hereinafter
HORIZONTAL MERGER GUIDELINES], available at http://www.ftc.gov/bc/docs/horizmer.htm.


                                                         3
market participants. 9 The Horizontal Merger Guidelines categorize three levels of market

concentration: markets may be "unconcentrated" (HHI below 1000), "moderately concentrated"

(HHI between 1000 and 1800), or "highly concentrated" (HHI over 1800). 10 The HHI provides

a snapshot of market concentration and, in the context of merger review, the post-merger change

in the HHI suggests the merger's likely effect on market concentration. It must be emphasized,

however, that the HHI is only the starting point for competitive analysis, and the Commission

does not make enforcement decisions based solely on market shares or HHIs. The analytical

significance of the HHI depends on other market factors (such as ease of entry and likely

competitive effects) that require further investigation and market analysis.

          For purposes of calculating the HHIs required by Section 1501(a)(2), we must assume

that U.S. fuel ethanol production is a relevant antitrust market. 11 This assumption precludes

consideration of potentially relevant product and geographic characteristics that would bear on a

complete competitive analysis of the ethanol industry. Indeed, provided that fuel ethanol

production remains above minimum production levels mandated by the Energy Policy Act of

2005, ethanol itself may not be a proper antitrust product market. At levels above the mandatory

minimum, ethanol likely competes with other blending components (alkylate, iso-octane, or

other clean, high-octane blending components) that can be used in gasoline, and refiners and


9
  For example, a four-firm market with market shares of 30 percent, 30 percent, 20 percent, and 20 percent has an
HHI of 2600 [(30 * 30) + (30 * 30) + (20 * 20) + (20 * 20) = 2600]. The HHI ranges from 10,000 (pure monopoly)
to a number approaching 0.
10
     See HORIZONTAL MERGER GUIDELINES § 1.5.
11
   A relevant antitrust market has both product and geographic aspects. A product market is a product or group of
products such that a hypothetical firm that was the only seller of those products would find it profitable to impose at
least a small but significant and nontransitory price increase above the competitive level. If such a price increase
would not be profitable because of the loss of sales to other products, the product or group of products would not be
a relevant product market. Similarly, a geographic market is a region such that a hypothetical firm that was the only
present or future producer of the relevant product in that region would find it profitable to impose at least a small but
significant and nontransitory price increase above the competitive level. If such a price increase would not be
profitable because of the loss of sales to producers outside the region, the region would be too narrowly defined to
be a relevant geographic market. See id. §§ 1.1-1.2.


                                                           4
blenders may choose to use other blending components in their gasoline as ethanol prices

increase. In regions or states that sell E-85 (a blend of gasoline that is 85 percent ethanol),

owners of flexible-fuel vehicles will choose regularly between purchasing E-85 or purchasing

gasoline with no more than 10 percent ethanol. If ethanol is part of an overall gasoline market,

or even a smaller market consisting of other clean-burning blendstocks, the HHIs in this analysis

could overstate concentration in the product market and suggest an exaggerated likelihood of the

potential for ethanol producers to engage profitably in anticompetitive behavior.

          Although this analysis is limited to U.S. ethanol production, imported ethanol also may

be analytically significant. The RFA reports that over 136 million gallons of ethanol were

imported into the U.S. in 2005, down from 149 million gallons in 2004. 12 Yet imports increased

significantly from January through July 2006, totaling 309 million gallons in that period alone

and increasing domestic ethanol supplies by 12 percent. 13 Of course, the annual variation in

ethanol imports is related to changes in the relative prices of ethanol in the U.S. and other

countries. As relative U.S. ethanol prices increase (as they did in 2006), selling ethanol into the

U.S. becomes more attractive to foreign producers, despite federal tariffs. Although imports

remain a small share of U.S. consumption, their presence suggests that HHIs for domestic

ethanol production tend to overstate concentration in the ethanol industry.

          Staff used three different methods of calculating HHIs in ethanol production and

marketing. As described below, staff first calculated HHIs based on the ethanol production

capacity of each individual producer. Staff then calculated HHIs that attributed each producer's


12
     See FROM NICHE TO NATION at 17.
13
  See EIA, Petroleum Navigator ­ Imports by Area of Entry, available at
http://tonto.eia.doe.gov/dnav/pet/pet_move_imp_a_EPOOXE_IM0_mbbl_m.htm. To calculate this figure, we
divided the volume of fuel ethanol imports during the first seven months of 2006 by domestic fuel ethanol
production during the same time. See EIA, Petroleum Navigator ­ Oxygenate Production, available at
http://tonto.eia.doe.gov/dnav/pet/pet_pnp_oxy_dc_nus_mbbl_m.htm.


                                                      5
capacity to the firm responsible for marketing the producer's ethanol. Finally, by way of

confirming these calculations, staff determined the HHIs on the basis of actual production, rather

than capacity.

            A.         Producer-Based Allocation, Using Capacity

            Staff first calculated market shares of producers based on their fuel ethanol production

capacity. Although market shares may be measured in other ways, such as by total dollar sales,

production capacity provides a useful and easily confirmable indicator of a producer's

competitive significance. 14

            To determine the production capacity of each ethanol plant, staff relied on publicly

available information supplemented by interviews with industry participants. RFA publishes and

frequently updates data regarding ethanol capacity and announced capacity additions. Many

producers publicly disclose existing plant capacity or future construction plans. Marketers also

may announce new agreements with producers. Staff interviewed producers, marketers, and

other industry participants to confirm public data.

            In attributing capacities to individual producers, staff included additional capacity from

new plant construction or expansion, provided that the construction or expansion had sufficiently

progressed such that the extra capacity could yield marketable volumes within one year. This is

consistent with the approach adopted in the Horizontal Merger Guidelines. 15 Staff attributed

additional capacity to the firm only if the firm had finished its expansion plans, received

necessary financing for the construction, and begun physical construction or expansion.

Although a producer may plan on expanding capacity substantially over the next few years, staff



14
  See HORIZONTAL MERGER GUIDELINES § 1.41. A firm's capacity likely is the best measure of its competitiveness,
because ethanol is an undifferentiated product (i.e., producers manufacture chemically identical ethanol).
15
     See id. § 1.32.


                                                      6
deemed these plans to be too speculative for this analysis until the producer has secured

financing and begun actual construction. 16

          Using this approach, if each U.S. ethanol-producing firm is allocated market share based

on its capacity, staff determined that the HHI would be 326, which is deemed an unconcentrated

market under the Horizontal Merger Guidelines. 17 Staff's similar calculation of this figure in last

year's report yielded an HHI of 499. Thus, there has been a reduction in concentration in ethanol

production since 2005. 18

          B.      Marketer-Based Allocation, Using Capacity

          Marketing agreements add complexity to a competitive analysis of the ethanol industry.

Producers must reach oil companies and others that ultimately blend ethanol with gasoline for

sale to consumers. Some producers market their own ethanol by entering into sale agreements

with oil companies, blenders, or brokers, and by arranging for truck or rail transportation to

storage facilities. Other producers, however, rely on third-party ethanol marketers to make these

arrangements. Because marketers often represent more than one producer, a marketer can

amalgamate volumes from multiple ethanol production facilities to provide purchasers with a

single source for volumes of ethanol that would exceed what any individual producer can

provide. Marketers also may negotiate more favorable transportation or storage rates that can

broaden the geographic area to which a plant's output can be supplied economically.


16
  The increased amount of ethanol plant construction and expansion seems to have lengthened the time needed for
builders to complete these projects. For example, according to industry participants, a new ethanol plant ordinarily
requires 15 to 18 months of construction time before volumes can be produced at the plant, a slight increase from
last year's findings. These considerations justify caution in estimating how much of announced capacity additions
will actually come to pass within one year.
17
  This number suggests an analytic precision that does not reflect the rate of change in this industry, particularly as
producers announce capacity additions seemingly on a weekly or even daily basis. Staff's HHI calculations
represent staff's best estimate as of June 30, 2006, the cut-off date for our analysis unless otherwise indicated. In
this regard, staff's approach excludes some capacity expansions identified in RFA data.
18
     See FTC, REPORT ON ETHANOL MARKET CONCENTRATION 9 (2005).


                                                           7
         There is no standard marketing agreement in the industry, and marketing agreements vary

in length. Some marketers maintain an equity ownership interest in their producers' facilities. In

virtually all instances, however, the ethanol producer determines its own output level.

         To analyze market concentration in light of these marketing relationships, staff evaluated

whether the capacities used to produce the ethanol marketed by one company should be

attributed to that single ethanol marketer rather than to the individual ethanol producers

represented by the marketer. A producer's ability to adjust its own output in response to

changing prices suggests that staff should treat each producer individually. Marketing "pools,"

however, may warrant a different approach. In a pooling arrangement, the marketer treats all of

its producers' volumes in common, makes sales to accounts, and decides which plant is best

situated to service the account. Each producer is allocated a prorated share from the common

revenue pool, based on the volume it contributes. 19 As a result, each producer within the

marketing pool receives an identical netback (e.g., the sale price less the cost of transportation

from the ethanol plant), regardless of where its production was actually delivered. Because a

customer receives a single offer from a marketer representing a pool of numerous producers,

pooling arrangements effectively reduce the number of bidders that compete to supply a given

customer. This characteristic might suggest that volumes from all producers sharing a common

marketer should be attributed to the single marketer. 20




19
  Under a non-pool marketing arrangement, a marketer sells its producers' volumes on a plant-specific basis. The
marketer typically presents sales opportunities to each plant, leaving it to the plant's management to decide whether
to accept the offer. Some ethanol also is sold to gasoline suppliers by firms acting essentially as brokers or resellers
that take ownership of the ethanol and sell for their own benefit.
20
  On the other hand, because marketers have no control over a producer's output decisions, a producer may have an
incentive to boost production in the event of an ethanol price increase. Increased production could undercut the pool
price and force the marketer to find additional buyers at potentially lower prices. Thus, even in the pooling context,
producers might be treated more appropriately as independent firms.


                                                           8
               Given the fact-specific nature of market analysis, staff cannot determine with certainty

the effect of each marketing agreement on the industry. Staff therefore calculated HHIs by

attributing all producers' shares to their marketers, regardless of whether the marketing

agreement involves pooling volumes. This approach yields an HHI of 995. This figure falls

within the range for unconcentrated markets under the Horizontal Merger Guidelines and is

down from last year's HHI of 1259 using the same allocation method. 21

               Staff alternatively calculated an HHI that attributed shares to marketers only when they

had pooling arrangements with their producers. For producers that did not market through a

pooling arrangement, staff attributed the market shares to the producers themselves. This

approach yields an HHI of 635 ­ also an unconcentrated market, and down from last year's HHI

of 813 using the same methodology.22

               C.     EIA Production Data

               Although capacity-based data provide a good indication of ethanol industry

concentration, calculation of HHIs on this basis is limited by difficulties in measuring ethanol

production capacity with precision. Most industry participants report capacity based on

"guaranteed" or "name-plate" capacity. Typically, builders and designers guarantee that a newly

constructed plant (or expansion) will produce a certain volume of ethanol. In this industry, the

plant often can produce more than the guaranteed capacity. Moreover, a plant will tend to

exceed its rated capacity as the producer improves the production process and gains expertise in

the plant's operation. It is not uncommon for ethanol plants to run 10 to 15 percent higher than

their stated capacities. 23


21
     See FTC, REPORT ON ETHANOL MARKET CONCENTRATION 11 (2005).
22
     See id.
23
     In addition, some industries demonstrate significant differences among competitors' capacity utilization rates.


                                                             9
               To address the measurement issues in the capacity-based HHI calculations, staff

performed a parallel analysis using ethanol production data. Every month, EIA collects

confidential information on the production of oxygenates such as ethanol and methyl tertiary-

butyl ether. Firms that produce over 8 million gallons of oxygenates per year must report to EIA

their monthly production volumes by product. EIA agreed to calculate the HHI data based on

annual production from July 2005 through June 2006, following the same attribution methods

outlined in the previous sections. To maintain its confidentiality obligations, EIA reported only

the final HHI numbers and did not disclose to us the volumes of ethanol attributed to each

producer or marketer.

               The two right-hand columns of Figure 1 show HHIs using EIA production data. If all

shares are attributed to the individual producers, the HHI is 683, which is unconcentrated under

the Horizontal Merger Guidelines and shows deconcentration in the market when compared to

the HHI of 929 from the 2005 report.24 If we allocate producers' shares to their marketers, the

HHI is 1345. This would be moderately concentrated under the definitions in the Horizontal

Merger Guidelines, although it represents deconcentration when compared to the HHI of 1613

from 2005. If we allocate producers' shares to their marketers for pooling agreements only, the

production-based HHI is 981, down from the HHI of 1221 from the 2005 report.25




Calculating HHIs by using actual production data controls for these differences.
24
     See FTC, REPORT ON ETHANOL MARKET CONCENTRATION 13 (2005).
25
     See id.


                                                         10
                                  Figure 1: Domestic Fuel Ethanol Concentration



                                                                          HHI Based on                HHI Based on
                                                                           Capacity                    Production

             Treatment of Marketing Agreements                         2006(P)       2007(P)         2005         2006

 Attribute shares to each producer                                           499           326           929           683

 Attribute shares to marketer only for pool marketing                        813           635          1221           982
 agreements

 Attribute shares to marketer for all marketing agreements                  1259           995          1613         1345
Source: RFA, EIA
Note: Capacity for 2006(P) includes capacity additions that, as of the FTC's 2005 report, were anticipated to be
completed by the end of 2006. Capacity for 2007(P) includes new capacity additions that, as of July 2006, are
expected to be completed by the end of 2007. Production data for 2005 are from July 2004 to June 2005, and
production data for 2006 are from July 2005 to June 2006.



         Because they are based on historical data, the production-based HHIs likely overstate the

concentration levels that will prevail in the near future in a deconcentrating industry. Production

data do not fully account for entrants that began producing ethanol during the period measured

by EIA, and should now be considered as full market participants. Production data also do not

account for capacity expansions or additions that will produce marketable volumes within the

next year.

         Also, year to year comparisons show a substantial amount of entry into ethanol

production. Such entry can deter anticompetitive conduct by reducing the likelihood that one

firm (or a group acting together) profitably could raise prices above competitive levels. 26




26
  See HORIZONTAL MERGER GUIDELINES § 3.0 ("A merger is not likely to create or enhance market power or to
facilitate its exercise, if entry into the market is so easy that market participants, after the merger, either collectively
or unilaterally could not profitably maintain a price increase above premerger levels."). For a more complete
discussion of entry and ethanol production, see FTC, REPORT ON ETHANOL MARKET CONCENTRATION 14-16 (2005).


                                                            11
                                          III.    Conclusion

                                          Our HHI analysis of market concentration shows that U.S. ethanol production is

unconcentrated or, at most (using actual production data), only moderately concentrated under

the Horizontal Merger Guidelines, revealing little incentive or ability for one or more firms to act

anticompetitively. New entry and other market factors reduce the significance of these figures,

and, as shown in Figure 2, concentration has fallen as production capacity has increased.

Nevertheless, given the highly fact-intensive nature of antitrust analysis, staff cannot exclude the

abstract possibility that future mergers may potentially create anticompetitive effects in a given

segment of the industry, or that industry participants may engage in anticompetitive conduct.


                                                        Figure 2: Historical Fuel Ethanol Capacity and HHIs
                                         8,000                                                                          2,500

                                         7,000
   Capacity (Million Gallons per Year)




                                                                                                                        2,000




                                                                                                                                Herfindahl-Hirschman Index
                                         6,000

                                         5,000
                                                                                                                        1,500

                                         4,000

                                                                                                                        1,000
                                         3,000

                                         2,000
                                                                                                                        500
                                         1,000

                                            0                                                                           0
                                                 1998     2000   2001   2002   2003        2004   2005   2006 2007(P)
                                     Capacity (Left Axis)        HHI (Right Axis)
 Source: RFA
 Note: Annual figures are for year-end for 1998 to 2004, and October for 2005 to 2006. 2007(P) is projected
 capacity for late 2007, which adds construction of new plants and expansions as of July 2006.




                                                                                      12