The National Oil Agency carries out, since 2005, biodiesel auctions. At these auctions, the refineries buy biodiesel to mix it up with the oil-based diesel (ANP 2012). According to the source, the initial objective of such auctions was to gener­ate a market and hence stimulate the biodiesel production in a big enough quantity for the refineries and distributors to compose the mixture, according to the law. Based on the results of these auctions, we obtained the biodiesel production in cubic meters, per state, as shown on Table 1.

We can see on Table 1 that the beginning of the biodiesel production took place in 2005 and that only four states were producing (Minas Gerais, Para, Parana, and Piauf), showing a high concentration, despite the small quantity being pro­duced, when compared to 2011 and 2012. In 2012, the biodiesel was produced in 12 of the 25 Brazilian federal units, where the higher production of the states Rio Grande do Sul, Goias, Mato Grosso, and Sao Paulo, stands out, representing 78.50 % of the national production, showing that this market is highly concen­trated in these states.

Table 1 Biodiesel production in thousand cubic meters per state between January 2005 and December 2012 State 2005 2006 2007 2008 2009 2010 2011 2012 Bahia 0.000 4.238 70.942 65.982 79.941 91.952 131.893 237.520 Ceara 0.000 1.956 47.276 19.208 49.154 66.337 44.524 59.001 Goias 0.000 10.108 110.638 241.364 268.702 442.293 505.586 566.558 Maranhao 0.000 0.000 23.509 36.172 31.195 18.705 0.000 0.000 Mato 0.000 0.013 15.170 284.923 367.009 568.181 499.950 477.008 Grosso Mato 0.000 0.000 0.000 0.000 4.367 7.828 31.023 76.635 Grosso do Sul Minas 0.044 0.311 0.138 0.000 40.271 72.693 76.619 81.313 Gerais Para 0.510 2.421 3.717 2.625 3.494 2.345 0.000 0.000 Parana 0.026 0.100 0.012 7.294 23.681 69.670 114.819 115.709 Piaui 0.156 28.604 30.474 4.548 3.616 0.000 0.000 0.000 Rio de 0.000 0.000 0.000 0.000 8.201 20.177 7.716 16.719 Janeiro Rio Grande 0.000 0.000 42.696 306.056 454.189 605.998 862.110 748.986 do Sul Rondonia 0.000 0.000 0.099 0.228 4.779 6.190 2.264 9.110 Sao Paulo 0.000 21.251 36.885 185.594 236.302 327.458 295.076 154.591 Tocantins 0.000 0.000 22.773 13.135 33.547 86.570 101.182 75.474 Total 0.736 69.002 404.329 1,167.129 1,608.448 2,386.397 2,672.762 2,618.624 Source ANP (2013a)

Table 2 shows that the ethanol production has grown 45.74 % between 2005 and 2012 (from 15.924000 to 23,209000 m3 in 2012). This growth can be partially explained by the increase of demand where, the increase in the internal market has been due to the more favorable price of this fuel, when compared to gas, which forces the consumption of alcohol in biofuel cars (gas and ethanol), which have had more and more representativeness in the national freight of small urban vehi­cles, since, as shown on Fig. 4, in 2005, 7 % of the cars (1.4 million) were using biofuel, and are now 57 % of the national freight. Besides, the mixture level of ethanol in gas in the last decade has varied from 20 to 25 % (according to the gov­ernment decision), thus implying more pressure on ethanol’s demand.

Table 3 shows us that the production of ethanol has increased the concentration in the four most important states (Goias, Mato Grosso do Sul, Minas Gerais, and Sao Paulo) that represented 81 % of the national production of ethanol in 2012. Also, Sao Paulo produced 51 % of the national volume in 2012 (11,830 thousand m3); however, this high concentration has decreased, since in 2005, two-thirds of the national production was centralized in this state.

Table 3 shows that both indexes (HHI and Cr(4)) point to a high concentration of the biodiesel production in Goias, Mato Grosso, Rio Grande do Sul, and Sao Paulo, these four states accounted for 87.22 % of the total biodiesel production in

State	2005	2006	2007	2008	2009	2010	2011	2012
Acre	0	0	0	0	0	1	3	4
Alagoas	546	604	853	845	626	716	673	541
Amazonas	6	6	8	8	5	7	6	4
Bahia	103	94	141	141	118	127	118	155
Ceara	1	1	1	9	11	3	8	4
Espirito Santo	235	173	252	275	238	187	224	178
Goias	729	822	1,214	1,726	2,196	2,895	2,677	3,130
Maranhao	139	128	170	182	168	182	177	160
Mato Grosso	771	749	894	952	826	857	844	975
Mato Grosso do Sul	496	641	877	1,076	1,261	1,849	1,631	1,917
Minas Gerais	959	1,291	1,777	2,168	2,255	2,558	2,084	1,994
Para	43	52	36	45	38	23	39	33
Paraiba	268	315	337	391	389	298	357	305
Parana	1,040	1,319	1,859	2,049	1,885	1,619	1,402	1,299
Pernambuco	328	319	417	530	400	385	358	272
Piaui	35	51	36	45	41	35	37	33
Rio de Janeiro	136	87	120	128	113	61	76	37
Rio Grande do Norte	74	78	49	115	122	83	106	72
Rio Grande do Sul	3	6	7	6	2	6	7	2
Rondonia	0	0	0	7	9	11	12	9
Sao Paulo	9,963	10,910	13,325	16,722	14,912	15,354	11,598	11,830
Sergipe	48	54	48	90	77	103	133	98
Tocantins	4	12	0	3	2	16	111	157
Total	15,924	17,710	22,422	27,513	25,694	27,376	22,682	23,209

Table 2 Ethanol production in thousand cubic meters per state between January 2005 and December 2012

Source ANP (2013b)

Table 3 Herfindahl-Hirschman index and Cr(4) for the biofuel production in the four biggest states producers Biofuel Index 2005 2006 2007 2008 2009 2010 2011 2012 Biodiesel Cr(4) (%) 100.00 93.04 67.16 87.22 82.45 81.46 80.92 78.50 HHI 5,305.72 2,939.78 1,525.30 2,010.28 1,864.03 1,798.57 1,939.58 1,850.44 Ethanol Cr(4) (%) 76.27 77.15 76.68 78.84 80.27 82.76 79.32 81.31 HHI 2,982.13 2,958.64 3,446.59 3,619.37 3,895.83 3,746.89 3,976.42 4,070.33

2008, remaining at a level close to 80 % in the following years. Notably, this result is influenced by the main raw material used: soybeans—Goias, Mato Grosso, and Rio Grande do Sul are the main producers. Sao Paulo, on the other hand, stands out due to the usage of beef fat and soybeans. In order to verify this high concen­tration is present, we initially identified the market participation of the 16 biggest companies producing biodiesel, as shown on Table 4.

We can see on Table 4 that for 2005, none of these companies produced biodiesel, whereas only in 2006 did Granol start its activities. In 2012, the 16 analyzed companies kept 80.19 % of market participation, thus indicating that this industry has characteristics of a Cr(16), because these companies keep at least 70 % of the Brazilian biodiesel production since 2007. Do also note that the majority of these companies are located in the south and central west, cor­roborating Picture 2. In order to confirm the installed capacity concentration, we calculated the HHI for the daily installed capacity per Brazilian region, as per Table 5.

Considering the increase in demand for ethanol, as well as the strategies of mergers and acquisitions among suppliers and distribution channels recently observed in Brazil, much has been discussed about the power of the market that may be taken by those agents who are involved in the product chain. Discussions on a possible cause for the increase of the product prices have risen interest on the existence of the market power by the ethanol producers, and/or by fuel distribution channels (Beiral 2011).

The analysis by companies (distilleries and/or sugar and ethanol plants) reveals a scattered environment, granted the great number of registered units. There are now 432 operational plants in Brazil, and 83 of these plants are located in the north-northeastern region and 349 are in the central-south region (MA 2012). Besides the great number of plants, this industry shows traits of a decentralized market, since in 2012 the five largest economic groups responded for 20 % of all the grinded cane in the country (UNICA 2010).

Despite the traits indicating a low concentration level in the ethanol industry in Brazil, it is impossible to calculate it, since the Instrugao Normativa No 52, of November 12, 2009, on the Diario Oficial da Uniao of November 13, 2009, does not allow the communication of all the necessary information, since according to the Article No. 5 “the information received from the legal entities will be classi­fied, and can only be communicated in an aggregated manner per state, region, or national total” (MA 2009).

Table 4 Market participation of the 16 largest biodiesel companies Company City Region 2005 (%) 2006 (%) 2007 (%) 2008 (%) 2009 (%) 2010 (%) 2011(%) 2012 (%) Adm Rondonopolis Central west 0.00 0.00 0.37 15.54 10.67 10.34 5.78 5.61 Biocamp Campo Verde Central west 0.00 0.00 0.00 1.07 1.73 2.08 2.05 2.49 Bsbios Passo Fundo South 0.00 0.00 3.58 7.91 7.02 5.63 4.82 4.91 Bsbios Marialva South 0.00 0.00 0.00 0.00 0.00 1.97 3.54 3.98 Camera Ijuf South 0.00 0.00 0.00 0.00 0.00 0.25 4.11 6.26 Caramuru Sao Simao Central west 0.00 0.00 11.43 9.80 7.62 6.70 5.46 5.54 Caramuru Ipameri Central west 0.00 0.00 0.00 0.00 0.00 1.96 3.78 4.59 Fiagril Fucas Rio Verde Central west 0.00 0.00 0.00 6.25 5.72 4.76 5.40 4.62 Granol Anapolis Central west 0.00 14.75 18.19 11.95 8.39 7.68 6.76 8.59 Grand Cachoeira do Sul South 0.00 0.00 0.00 7.71 7.54 6.92 7.91 4.72 JBS Colide Central west 0.00 0.00 0.31 6.26 5.43 5.22 3.83 2.44 Oleoplan Veranopolis South 0.00 0.00 2.08 8.66 11.13 8.54 9.15 8.70 Olfar Erechim South 0.00 0.00 0.00 0.00 0.00 2.28 4.59 5.05 Petrobras Candeias Northeast 0.00 0.00 0.00 0.87 2.54 2.86 4.19 5.75 Petrobras Montes Claro Southeast 0.00 0.00 0.00 0.00 2.50 3.06 2.80 2.95 V-Biodiesel Iraquara Northeast 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.98 Total 0.00 14.75 35.97 76.01 70.29 70.24 74.19 80.19 Source ANP (2013a)

Table 5 Market participation of the 18 largest distribution firms for ethanol in the country, from 2005 to 2012 Firm 2005 (%) 2006 (%) 2007 (%) 2008 (%) 2009 (%) 2010 (%) 2011 (%) 2012 (%) BR 15.22 17.44 15.07 17.89 19.29 22.25 21.24 21.30 Shell 8.07 10.17 8.39 10.61 12.40 13.09 13.22 17.99 Ipiranga 12.03 13.91 10.64 12.49 11.48 17.02 16.46 16.37 Subtotal 35.32 41.52 34.09 40.99 43.17 52.36 50.92 55.66 Cosan/Esso 5.29 5.39 3.91 5.01 4.89 5.05 5.06 2.61 Alesat 0.00 0.00 0.86 2.06 2.22 2.59 2.20 2.25 Quality 0.00 0.00 0.00 0.00 0.00 0.00 0.04 2.05 Eldorado 0.00 0.00 0.00 0.00 0.00 0.18 1.81 1.99 Euro 0.00 0.00 0.00 2.31 2.59 2.16 2.01 1.93 MM original 0.70 0.00 0.57 1.38 1.24 1.38 1.53 1.80 Petroluna 0.00 0.00 0.07 0.76 1.25 0.00 1.38 1.58 Brasil oil 0.00 0.00 0.00 1.65 2.66 1.84 2.46 1.54 Gigante 0.00 0.00 0.00 0.00 0.00 0.00 0.36 1.48 Petrosol 0.00 0.00 0.00 0.00 0.00 0.94 1.15 1.41 Gpetro 0.91 0.48 0.37 0.05 0.01 0.00 0.80 1.39 Manguinhos 0.00 0.00 0.00 0.00 0.00 0.00 0.43 1.36 Petromais 0.00 0.00 0.11 0.58 0.61 0.74 1.74 1.32 Tube toy’s 0.00 0.00 0.00 0.00 0.00 0.03 0.25 1.10 Fera 0.00 0.00 0.00 0.00 1.00 0.00 0.75 1.09 Subtotal 6.89 5.88 5.89 13.80 16.47 14.90 21.96 24.91 Total 42.21 47.40 39.98 54.79 59.64 67.26 72.88 80.57 Source ANP (2013c)

The perspectives of the Brazilian and American governments are not the same, since according to the Section 1501 of the Energy Policy Act of 2005, the Federal Trade Commission must analyze the market concentration of the production of etha­nol, using the HHI to determine whether or not there is enough competition among the participants of this industry so as to avoid fixing prices and other anticompetition behaviors. According the US Federal Trade Commission (2012), the American etha­nol industry is not deconcentrated (HHI equals 244 in 2010, and 284 in 2011), thus suggesting that an attempt to exert market power by any agent is unlikely.

On the other hand, part of the ethanol marketing in Brazil is made via market­ing groups, the structure of the ethanol market is much more concentrated (Beiral 2011). Thus, point to a concentration of the ethanol market as a trend. For exam­ple, the purchase of the Esso by Cosan (the largest ethanol producer in Brazil); the purchase of the Ipiranga network by Ultra (the second largest distribution channel after the purchase of Texaco, only after BR Distribuidora), along with Petrobras and Braskem (Beiral 2011).

Considering the possible increase of concentration among the ethanol distributors, we sought to evaluate the market participation of the 18 largest distribution firms for ethanol in the country, from 2005 to 2012, as shown on Table 5. Table 5 shows that in 2012, 18 firms kept 80.57 % of the ethanol distribution market. Also, note that

Table 6 Daily installed capacity, usage of installed capacity per Brazilian region of Brazilian biodiesel industry Region Daily installed % in Brazil # of companies Annual plant usage capacity Considering 264 days (%) Considering 365 days (%) Central west 6,689.55 42 31 59 43 Northeast 2,058.13 13 06 32 24 North 620.00 4 05 63 46 Southeast 2,367.33 15 12 61 45 South 4,367.38 27 10 85 62 Source Conab (2013)

only three companies (BR Distribuidora, Shell, and Ipiranga) held in 2012 55.66 % of this market, showing a high market concentration, increasing since 2005, since these companies held 35.32 % market share (a 57.58 % increase for this period).

On Table 6, we can see the higher installed capacity in central west region, with 42 % of all authorized capacity in Brazil, having 31 of those 64 companies authorized to produce biodiesel. When analyzing the usage of the authorized capacity of plants installed in each one of those five regions in Brazil, and considering a 365 day year, we can see that the south region has 62 % of usage, and that if we consider 264 work­ing days in a year, the south region holds 62 % of usage and that if we consider 264 working days in a year (22 days in a month and 12 months in a year), the south region points to the limit of its capacity, with 85 % usage of its capacity in 2011. On the other hand, the northeast region only has 32 % usage of its capacity, whereas the total num­ber of companies has used 63 % of its authorized capacity for producing biodiesel.

Table 7 shows that the largest capacity installed for cane grinding and sugar and ethanol production is in the southeast, with 89, 65, and 90 % of all the Brazilian productive capacity, respectively. By analyzing the use of the average installed capacity of plants, for each of the five Brazilian regions, we can see that the main productive constraint is grinding, where the central west, northeast, and southeast regions show an over 70 % usage. Note that the usage of the installed capacity of grinding in the southeastern region is 75 %, a concerning fact since two-thirds of the installed national capacity is in this region.

On Table 8, we can see that the industrial concentration level has significantly decreased since the National Biodiesel Production Program was implemented, where in the first years of this program (2005-2006), the market showed to be highly concentrated, significantly decreasing the companies power in the market, since in 2012, the four main companies (Granol in Anapolis; Oleoplan in Veranopolis; Petrobras in Candeias and Camera in Ijuf) represented around 30 % of the volume produced in the last three years (2010, 2011, and 2012). In order to verify the impact of this low industrial concentration, we analyzed the weighted average cost of the 16 main companies and compared it to the other companies in the same sector.

As it is commented, the negotiation process for biodiesel is performed accord­ing to auctions. In order to verify whether or not this low industrial concentration

Table 7 Daily installed capacity, usage of installed capacity per Brazilian region of Brazilian cane grinding, sugar and ethanol industry Daily installed capacity % in Brazil Annual plant usage (Considering 365 days) Region Cane grinding (thousand of tons) Sugar (thousand of tons) Ethanol (thousand of tons) Cane grinding Sugar Ethanol Cane grinding (%) Sugar (%) Ethanol (%) Central west 356.78 26.83 44,070,00 15 11 18 72 60 62 Northeast 205.50 40.90 22,511,00 9 16 9 79 49 46 North 26.84 1.18 6,559,00 1 0 3 44 38 28 Southeast 1,543.64 165.43 152,490,00 66 65 63 75 52 52 South 189.42 20.00 17,760,00 8 8 7 63 38 39 Source Conab (2013)

Table 8 HHI and Cr(4) to identify the industrial concentration of the biodiesel production in Brazil between 2005 and December 2012 Index 2005 2006 2007 2008 2009 2010 2011 2012 HHI 5,306 2,906 1,147 812 605 508 465 475 Cr(4) (%) 100 92 60 46 38 33 30 29

Table 9 Difference between the weighted average cost per liter sold in 2011 and 2012 the main biodiesel producing companies and the other companies in the market	among
Auction	16 largest companies	Other companies	Difference	t Test	sig.
22	R$2.32	R$2.33	R$(0.01)	0.075	0.943
23	R$2.45	R$2.51	R$(0.06)
24	R$2.46	R$2.48	R$(0.01)
25	R$2.12	R$2.07	R$0.05
26	R$2.49	R$2.44	R$0.05
27	R$2.67	R$2.67	R$(0.00)

would interfere in the pricing at these auctions, we calculated the average cost per liter of the 16 main biodiesel producing companies in Brazil. The choice of these companies was due to the fact that they represent around 80 % of the national production. As a result, we observed that in 2011 (Auctions, 22, 23, and 24) these companies used a lower price than the other companies; however, in 2012, (Auctions, 25, 26, and 27) the situation was inverse: in two out the three auctions, these companies used a price lower than the others.

In order to verify whether or not the weighted average costs of these two cat­egories are statistically significant, we calculated the average difference t test (Table 9), where we accepted the null hypothesis of equal averages, thus indicating that such companies, despite the fluctuations, do not price in a differentiated man­ner in the long term. This result can be explained by the hypothesis that companies would not have significant gains, considering the sector’s low concentration that makes a significant price reduction impossible at auctions, thus indicating some homogeneity of prices practiced in the biodiesel industry in Brazil.

On the other hand, it is not possible to calculate the concentration level of the ethanol production in Brazil, considering the Instrugao Normativa Number 52 of the Ministry of Agriculture, prohibiting the communication of ethanol production per productive unit, only allowing its communication in an aggregated manner, per state, region, or national total. However, there is another type of strength in this system: the distribution channels. Thus, we sought to evaluate the evolution of industrial concentration of the ethanol distribution in Brazil, in the last 8 years, where we calculated the HHI and the Cr(4), as per Table 10.

Table 10 shows that the industrial concentration level has grown significantly, according to Bain and Qualls (1968) and Usdoj (1997), pointing to a situation where this market moved from a weak-concentration oligopoly, to a moderate concentration, especially after 2009. In order to verify the impact of this increase of industrial concentration on this industry’s profitability, we simulated the

Table 10 Herfindahl-Hirschman index (HHI) and Cr(4) to identify the industrial concentration of the ethanol distribution in Brazil between 2005 and December 2012 Index 2005 2006 2007 2008 2009 2010 2011 2012 HHI 746 624 744 795 1,048 980 1,098 nda Cr(4) (%) 49 41 48 50 57 56 58 nda and = no data available

Fig. 5 Contribution margin for the ethanol distribution channels (adapted from Beiral 2011)

Table 11 Contribution margin, on monthly basis, of distributors of ethanol, Herfindahl — Hirschman index between 2005 and 2011 and Pearson’s product-moment correlation coefficient Year Contribution margin (on monthly basis) (%) HHI Year Contribution margin (on monthly basis) HHI Correlation Sig 2005 1.46 746 2009 6.65 % 1,048 0.443 0.320 2006 1.81 624 2010 5.36 % 980 2007 6.17 744 2011 3.77 % 1,098 2008 6.06 795 2012 nd* nda and = no data available

contribution margin for the ethanol distribution channels, as suggested by Beiral (2011). Results are shown on Fig. 5.

Figure 5 shows that in 2005 and 2006, the ethanol distribution channels showed, for several months, a negative contribution margin, and it became positive as of January 2007. Note that as of now, the concentration level of the distribution industry has grown, showing a positive correlation among variables. In order to confirm this supposition, we calculated the Pearson’s product-moment correlation coefficient between these two variables, as shown on Table 11.

As it is shown on Table 11, the Pearson’s product-moment correlation coefficient between the distributors contribution margin, and the HHI, for this industry, was 0.433 (sig equals 0.320) pointing to a positive correlation, although it is relatively weak, between the variables, i. e., as one of the variables increases (industrial concentration level), the other one increases too (distribu­tors contribution margin), noting that this increase of concentration has made it easier for pricing, thus implying a profitability increase for this industry, in det­riment of society.