Petrobras. Look Beyond The Negatives. Research BRAZIL OIL &GAS. Pre-Salt: The key growth catalyst. M&A could be a Wake-up Call for Investors

Transcription

1 BRAZIL OIL &GAS Initiation of Coverage Oil & Gas Petrobras Look Beyond The Negatives We initiate coverage of Petrobras with a BUY recommendation and fair values of R$26.7 and R$29.8 for PETR4 and PETR3, respectively. The development of the presalt region should enable PETR to more than double its oil output by 2020, on our estimates, mostly through the addition of better-quality oil with a production breakeven that we estimate at around $40/bbl and with the additional advantage of operating in a stable political environment. Despite what we regard as the company s sound growth prospects both in terms of volumes and profitability / barrel Petrobras is trading at a discount to oil majors as investors appear to have focused on its potential weaknesses, namely its public ownership and local supply chain limitations. In our view, all the negatives are more than priced in at the current price, therefore we see PETR as an affordable play on world-class E&P assets. Pre-Salt: The key growth catalyst We expect Petrobras domestic oil and gas production to grow at a CAGR 11E-20E of 10.6%, driven by a fast ramp-up of production from the pre-salt. Our optimism reflects the recent guidance of BG Group, which has increased its estimates for recoverable volumes and highlighted better than initially expected project economics. M&A could be a Wake-up Call for Investors Potential M&A activity involving pre-salt assets could be an important valuation support. After Repsol s sale of a 40% stake in Repsol Brazil to Sinopec (not rated) at a 33% premium to ESIBR s estimate, the next deal scheduled to be announced is the placement of a stake in Galp Brazil. All of Galp s (BUY, FV Eur 18.1) assets in Brazil are in partnership with PETR; therefore, the readacross to PETR should be more immediate. We believe the deal could mean a value readacross to PETR as high as $96bn (57% of the mkt cap), assuming a similar 33% premium, highlighting what we regard as the disparity between PETR s share price and its rich assets portfolio. Validation from an industry player with access to much more detailed data than an equity investor could provide important support for our valuation. Capex Plan: High but Looks Deliverable Along the way from changing from being a small onshore operator to becoming the largest deepwater oil producer in the world, PETR has delivered consistent production growth (CAGR : 8.3%) with a capex pace (CAGR: %) similar to the one we expect until the next peak of investments in 2016 (CAGR 5-year 6.2%). High Oil Price Leverage We see Petrobras as highly leveraged to oil prices. For each $10/bbl increase in our longterm oil price assumption of $80/bbl, our FV increases by 23%. PETR s short-term monetization is not as fast as that of some of its peers due to its policy of not immediately passing through oil price volatility to domestic fuel prices. However, looking at past evidence, we feel comfortable assuming internationally linked prices over the long run. Supply Chain & Local Content Restrictions PETR s high capex plan and local content requirements represent a challenge for the oil equipment industry. We recognize that delays may occur; yet we believe that PETR s plan should be viable due to: i) PETR s efforts to support the supply chain; ii) government commitment to guaranteeing the crucial investment conditions; and iii) a coalition of suppliers attracted by the large capex expected over the next years. We stress that even if the crown jewel projects are delayed, the impact on our FV appears to be limited. Each year of delay to the start up of BMS-11 and BMS-9 would reduce our FV by only c.3%. Valuation: Capitalized Company With Growth at a Discount A recent sell-off has fully adjusted PETR s multiples to account for the 2010 capitalization. We believe it deserves to trade at a premium due to its growth potential; however, PETR4 trades at 7.8x P/E 12E, a slight 1% discount to majors. At this level, we believe that the disadvantages associated with its public ownership are fully priced in. BUY Bovespa, Bloomberg Code PETR4.BZ Share Price (27/Oct/11) R$ 20.9 Market Capitalisation (m) R$ 286,995 Free Float 72% Source: Company data, Espírito Santo Investment Bank, for estimates. Econ. Net Debt/Ebitda (x) 1.1x 1.5x 1.9x 2.1x Source: Company data, Espírito Santo Investment Bank, for estimates. Source: Company data, Espírito Santo Investment Bank for estimates. MARKET RELATIVE PERFORMANCE IBOV PETR4 PETR3 Source: Bloomberg. Analysts Igor Maresti imaresti@bessecurities.com.br [BES Securities do Brasil, S.A CCVM] Filipe Rosa frosa@besinv.pt [Banco Espírito Santo de Investimento, S.A.] 27% upside Fair Value R$ 26.7 Year to Dec, R$ 2010A 2011E 2012E 2013E Revenues (m) 213, , , ,630 EBITDA (m) 60,323 64,734 66,248 74,095 EPS DPS FCF ps EV (m) 339, , , ,244 Economic Net Debt (m) 66,370 95, , ,092 At Current Price: 2010A 2011E 2012E 2013E PE (x) 7.8x 6.8x 7.8x 7.6x EV/Ebitda (x) 5.6x 5.7x 6.0x 5.8x Dividend Yield (%) 4.3% 3.8% 3.8% 3.8% FCF Yield (%) -10.8% -8.7% -6.9% -6.8% At Fair Value: 2010A 2011E 2012E 2013E PE (x) 9.9x 8.6x 10.0x 9.7x EV/Ebitda (x) 6.9x 6.9x 7.2x 6.8x Dividend Yield (%) 3.4% 3.0% 3.0% 3.0% Market Relative Performance Oct-10 Dec-10 Feb-11 Apr-11 Jun-11 Aug-11 Oct-11 FOR IMPORTANT DISCLOSURE INFORMATION, INCLUDING DISCLOSURES RELATED TO THE U.S. DISTRIBUTOR OF THIS REPORT, PLEASE REFER TO THE FINAL PAGES OF THIS REPORT Please refer to page 58 of this report for important disclosures, analyst certifications and additional information. Espirito Santo Investment Bank does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. This research report has been prepared in whole or in part by research analysts based outside the US who are not registered/ qualified as research analysts with FINRA.

2 Summary financial information PETROBRAS Valuation Metrics (Year end Dec) 2010A 2011E 2012E 2013E 2014E 2015E Rating Buy Fair Value (R$ / PETR4): 26.7 Recurrent P/E (x) Fair Value (R$ / PETR3): 29.8 Reported P/E (x) Fair Value (US$ / PBR/A): 31.2 EV / Sales (x) Fair Value (US$ / PBR): 34.8 EV / EBITDA (x) EV / EBIT (x) Share Price (27/10/11 - R$ / PETR4): 20.9 P/BV (x) Share Price (27/10/11 - R$ / PETR3): 22.8 FCF Yield (%) -10.8% -8.7% -6.9% -6.8% -4.8% -2.6% Share Price (27/10/11 - US$ / PBR/A): 25.0 Dividend Yield (%) 4.3% 3.8% 3.8% 3.8% 3.8% 3.8% Share Price (27/10/11 - US$ / PBR): 26.9 Upside / Downside potential (PETR4) 27.5% Key Ratios 2010A 2011E 2012E 2013E 2014E 2015E Upside / Downside potential (PETR3) 30.5% Upside / Downside potential (PBR/A) 25.1% EBITDA margin 28.3% 27.7% 32.6% 34.2% 35.2% 35.4% Upside / Downside potential (PBR) 29.3% EBIT margin 21.3% 20.7% 22.9% 23.8% 24.6% 24.8% Capex / Revenue 35.8% 32.8% 37.2% 35.3% 32.4% 29.2% Bloomberg PETR4.BZ Capex / Depreciation (x) Bovespa PETR4 Net Debt/EBITDA Net Debt / Capital Employed 17.8% 22.1% 25.9% 28.8% 30.4% 30.9% Net Debt/Equity 22% 28% 35% 40% 44% 45% Shares in Issue (Less Treasury) (MM) 13,044 ROE 15% 13% 10% 10% 10% 10% Market Cap (R$mn) 286,995 ROIC 7% 7% 6% 6% 6% 7% Net Debt 2Q11 (R$mn) 73,644 Minority Shareholders & Associates 3,631 Enterprise Value (R$ '000) 364,270 P & L Summary (R$mn, unless stated) E 2012E 2013E 2014E 2015E Net Revenues 213, , , , , ,529 Forthcoming Catalysts % change nm 9.5% -13.0% 6.6% 10.1% 12.2% 3Q11 Earnings Release 11-Nov-11 EBITDA 60,323 64,734 66,248 74,095 84,007 94,703 % change nm 7.3% 2.3% 11.8% 13.4% 12.7% % margin 28.3% 27.7% 32.6% 34.2% 35.2% 35.4% ES Equity Analysts D & A 14,881 16,258 19,780 22,628 25,432 28,292 Igor Maresti Impairment EBIT 45,366 48,472 46,468 51,466 58,576 66,410 imaresti@bessecurities.com.br % change nm 6.8% -4.1% 10.8% 13.8% 13.4% % margin 21.3% 20.7% 22.9% 23.8% 24.6% 24.8% Associates 208 1,073 1,095 1,117 1,139 1,162 Revenue Breakdown Operation Profit 45,575 49,546 47,563 52,583 59,715 67,572 Distribution 18% G&P 4% International 7% Supply 46% E&P 25% Net Financials 2,563 5, ,146-4,248-5,478 Other Pre-tax Income Pre-Tax Profit 48,137 54,579 46,783 49,437 55,466 62,094 Income Tax Expense -12,236-13,356-11,571-12,932-15,310-17,634 Minority Interests Net Income (Controlling Shareholders) 35,189 40,326 34,797 36,074 39,682 43,935 Recurrent Net Income 35,189 40,326 34,797 36,074 39,682 43,935 Reported EPS (R$) Recurrent EPS (R$) DPS (R$) Payout Ratio 33.3% 25.9% 30.0% 28.9% 26.3% 23.8% Shares in Issue (Less Treasury)(m) 13,044 13,044 13,044 13,044 13,044 13,044 Oil Production 4,000 3,500 3,000 2,500 2,000 1, e 2012e 2013e 2014e 2015e Total Production (kboepd) Domestic Oil (kbpd) Cash Flow Summary (R$'000) E 2012E 2013E 2014E 2015E EBITDA (net of provisions) 60,323 64,734 66,248 74,095 84,007 94,703 Taxes Paid -12,236-13,356-11,571-12,932-15,310-17,634 Interest Paid / Received 1,228 4, ,986-2,893-3,978 Change in Working Capital -2,233-2,761 1,675-1,219-1,687-1,849 Associate & Minority Dividends Other Operating Cash Flow 3,199 3, Operating Cash Flow 50,280 56,803 56,754 57,702 63,960 71,009 Capital Expenditure -76,411-76,628-75,588-76,448-77,343-78,238 Free Cash Flow -26,131-19,825-18,834-18,746-13,383-7,229 Acquisitions & Disposals Dividend Paid to Shareholders -10,466-11,422-10,436-10,436-10,436-10,436 Equity Raised / Bought Back 51, Other Financing Cash Flow 12,025 14, ,090 25,090 19,080 Net Cash Flow 27,258-16,813-28, ,271 1,415 EBITDA (R$mn) and EBITDA Margin (%) 34% 33% 28% 28% 74,095 66,248 64,734 60,323 35% 84,007 94,703 35% Balance Sheet Summary (R$'000) E 2012E 2013E 2014E 2015E 40.0% Cash & Equivalents 56,341 39,524 10,710 11,619 12,890 14, % Tangible Fixed Assets 282, , , , , , % Goodwill & Intangibles 83,098 82,679 82,679 82,679 82,679 82,472 Associates & Financial Investments 8,879 9,868 10,962 12,079 13,218 14, % Other Assets 88,815 93,949 85,284 92, , ,414 Total Assets 519, , , , , , % Interest Bearing Debt 122, , , , , ,736 Other Liabilities 87,036 92,316 85,399 91,092 99, , % Total Liabilities 209, , , , , , % Shareholders' Equity 306, , , , , ,639 Minority Interests 3,458 3,851 4,266 4,697 5,171 5, % Total Equity 310, , , , , , e 2012e 2013e 2014e 2015e 0.0% Net Debt 66,370 95, , , , ,431 Source: Company data, Bloomberg, and Espirito Santo Investment Bank for estimates. Page 2 of 61

3 Table of Contents Section1: Investment Thesis... 4 Section 2: Valuation Section 3: Detailed Valuation: Going Through the Business Units Section 4: Key Investor Themes Section 5: Earnings - Margin Expansion and EPS Growth Section 6: Risk Factors Appendix Appendix 1: All roads lead to... the pre-salt Appendix 2: Regulatory Background Appendix 3: Financials & Detailed Assumptions This report is priced as at 10/27/2011, unless otherwise stated Page 3 of 61

4 Section 1: Investment Thesis Moving to the next level Over the past several years, Petrobras has consistently increased its production growth targets and, consequently, capex estimates as well. The expected cash flow pressure over the next few years due to the company s aggressive capex plan has led some apparent investor concerns about: i) the company s financing ability; and ii) the ability of the supply chain to deliver on time. We acknowledge the risks and recognize that some projects are likely to be delayed. However, in general terms, we believe the plan is achievable. We remind investors that Petrobras has always taken a pioneering approach and is the main company responsible for the establishment of the Brazilian oil industry. Looking back, we can see that Petrobras has already carried out aggressive but successful investment plans in the past. PETROBRAS E&P CAPEX IN REAL USD ($MN) VS. HISTORICAL PRODUCTION (KBPD) 35,000 5,000 USDmn 30,000 25,000 20,000 15,000 10,000 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1, kbpd E&P Capex ($mn) Domestic Oil Production (kbpd) Source: Petrobras and Espirito Santo Investment Bank for estimates between 2011 and 2020 Although we cannot precisely compare the conditions of the past with those that the company will face ahead, we highlight from the chart above that Petrobras aggressive E&P capex spending is nothing new. According to our estimates, PETR s capex should peak in However, we do not believe that the challenges ahead are greater than those that the company faced at the beginning of the 80 s, when a much smaller Petrobras with limited knowhow was making large investments in a much less favourable economic and political environment in order to start developing offshore wells. PETROBRAS E&P CAPEX: HISTORICAL AND FORECAST Period Investments Peak Investments CAGR Domest. Oil Prod. Peak Production CAGR ($mn) Year Period 5-year prior peak kbpd Year Period 5-year prior peak , % 16.4% % 1.1% , % 27.3% % 3.2% , % 9.3% 1, % 12.1% , % 24.1% 2, % 3.5% , % 6.2% 4, % 11.9% Source: Petrobras and Espirito Santo Investment Bank for estimates between 2011 and 2020 Page 4 of 61

5 Strong Production Growth driven by the Pre-Salt The key driver of our production growth expectations is the pre-salt. According to our estimates, oil production from the pre salt should reach 2.5mn boepd in 2020, representing 52% of total domestic oil. The chart below compares our expected production curve to Petrobras guidance. While we are 10% below company guidance in 2015, we are practically in line for However, it is interesting to note the breakdown of our domestic oil production curve. In 2020, we are 20.6% below company guidance for the production in the post salt but 28.4% above for the production from the pre-salt. Our optimism with regard to the pre-salt reflects the recent guidance of BG Group, an important partner of Petrobras in both BMS-11 and BMS-9. On the other hand, our conservative forecast for the post salt solely reflects our expectations for a limited supply of equipment and workforce to develop these resources during this period. We expect Petrobras to prioritize the pre-salt production on the back of higher productivity. PETROBRAS TOTAL DOMESTIC PRODUCTION: ESIBR VS. PETR S GUIDANCE (KBOEPD) 6,500 6,000 5,500 5,000 4,500 4,000 3,500 3,000 2,500 2, ESIBR PETR Source: Petrobras and Espirito Santo Investment Bank for estimates between 2011 and 2020 While Petrobras has kept a low profile by maintaining its guidance unchanged despite the intense and successful exploratory activity in the area, BG has adopted an approach of active dialogue with investors and provided regular updates to its guidance to reflect the new data. As a member of the operating consortium, BG has access to the same information as PETR. We calculate that BG s new guidance implies total resources for Block BM-S-11 of 18bn boe and for BM-S-9 of 3.6bn boe, which compares with PETR s official guidance of 11.8bn boe and 1.55bn boe for BMS-11 and Guara, respectively. Also, BG has also made positive comments on project economics, indicating better well productivity, a faster ramp-up of production and a longer period at plateau level. Thus, despite the significant upward revision of the resources guidance, BG says it believes it will not need to increase the overall capex for the projects, as the already planned facilities (FPSOs and wells) should be able to deliver much higher productivity than initially expected. We note that the pilot project of Lula has shown the highest flow rate per well in Brazil (around 36k boepd) since May/11. M&A: Potential Valuation Support We believe that further M&A activity involving pre-salt assets could be an important catalyst to demonstrate the value of Petro s asset portfolio to investors. In our view, Repsol s sale last year through a capital increase of a 40% stake in Repsol Brazil to Sinopec has provided the best benchmark so far. According to our calculations, the deal valued Repsol s resources at $5.8/boe, a premium of 33% to the fundamental valuation of Repsol s assets ($4.4/boe) of our European oil team (Time for a Breather?, from Feb/23/2011). Moreover, the Portuguese company Galp Energia is about to conclude the placement of a stake in its Brazilian unit. As all of Galp s assets in Brazil are in partnership with Petrobras (including the crown jewel BMS-11), we believe that Page 5 of 61

6 Galp s deal could be an important catalyst for PETR, as it should support our view that PETR s share prices are discounted in comparison with its rich assets portfolio. We also think that it should increase investor confidence, as any buyer will have scrutinized all data available in much more detail than is available to an equity investor. We illustrate below the potential readacross of the value of deal to Petrobras. Our European Oil&Gas team s EV for Galp Brazil is US$13bn. Using our EV breakdown, we calculate the implied value for Petrobras and compare this with its market cap. At our base case valuation, Petrobras implied value would reach 43% of its current market cap. GALP S BRAZIL EV BREAKDOWN VS. IMPLIED VALUATION FOR PETROBRAS % Galp BZ EV-ESIBR NPV / boe Gross Volumes PETR's Blocks (US$mn) ESIBR ($/boe) Total 27 years stake Volume NPV (US$mn) % current mt cap BMS-11 9, % ,000 17, % 11,082 59, % BMS-24 / Jupiter 2, % 2.0 6,500 6, % 4,896 9, % BMS-21 / Caramba % % 507 2, % BMS-8 / Bem-te-vi % % 405 1, % Sub -Total 25,800 24,417 72, % Net risked exploration resources % Total 13,026 Source: Espirito Santo Investment Bank for estimates We highlight that our SoTP valuation for PETR values BMS-11 at the same US$59,164m NPV implied from this base case Galp Brazil readacross. Therefore, this deal could be an important support to our estimates. Moreover, we should bear in mind that, at our estimated EV, BMS-11 represents 20% of the total, while it would represent 35% of the current market cap, reflecting the valuation gap implied by the current share price. We also provide below a sensitivity analysis of how this readacross changes as a function of the deal s value. If the premium to our valuation reached the 33% of Repsol s, the readacross to PETR would reach 58% of its current market cap. The last column shows the potential impact on our FV if we were to include the implied value of BMS-11 in our SoTP. We note that a lower WACC (vs. our used 10%) or higher oil prices (vs. our flat curve at US$80/bbl) could add upside risk to our numbers. GALP S BRAZIL EV BREAKDOWN VS. IMPLIED VALUATION TO PETROBRAS Galp's Deal Premium to ESIBR Implied PETR's % of PETR Impact on FV Value ($mn) Valuation (%) assets value (US$mn) mkt cap R$/sh % 18,236 40% 102, % % 17,289 33% 96, % % 16,934 30% 94, % % 15,631 20% 87, % % 14,328 10% 80, % % 13,026 0% 72, % % 11,723-10% 65, % % 10,421-20% 58, % % 9,118-30% 51, % % 7,816-40% 43, % % Source: Espirito Santo Investment Bank for estimates Declaration of Commerciality as a Catalyst Another momentum generator that we will watch is the deadline for the declaration of commerciality of key fields. We point out that the deadline for Carioca is Nov/11/2011. We expect Petrobras to provide guidance on recoverable volumes together with this declaration by this date. So far, the company has only disclosed volumes guidance for Guara at the BMS-9 of between 1.1bn boe and 2.0bn boe. We estimate 3.61bn for the entire BMS-9; therefore, we expect this announcement to offer important confirmation of our estimates. Page 6 of 61

7 We note, however, that we see risks of this deadline being postponed. According to a report in Brasil Energia on Oct/25/11, ANP has required Petrobras to build a new well to be drilled at Carioca to approve the revision of its evaluation plan. Therefore, this may postpone the deadline to Jun/2013. However, Petrobras has confirmed to us that so far the deadline is still Nov/11/2011. DEADLINE FOR DECLARATION OF COMMERCIALITY OF KEY BLOCKS Block Evaluation Area Deadline BMS-9 Carioca 11-Nov-11 BMS-10 Parati 12-Apr-12 BMS-8 Bem-te-Vi 31-Dec-12 BMS-9 Guara 31-Dec-12 BMS-11 Iara 31-Dec-12 BMS-21 Caramba 30-Apr-15 BMS-24 Jupiter 28-Feb-16 Source: Espirito Santo Investment Bank for estimates A State-Owned Company: Discount Looks Too High One cannot ignore the potential risks associated with public ownership of the company. We note, for example, value erosion of US$8.4bn from the oil-for-shares agreement carried out in The company acquired producing rights for 5bn boe in the pre-salt area through new shares for US$8.51/boe (or US$10.08 at 2012 value vs. our valuation of US$8.39). We believe, however, that our valuation is ahead of market consensus as we add US$4.6bn of fiscal benefits to our valuation. The R$74.8bn paid to the government was registered as an intangible at PETR s 3Q10 results. This amount will be amortized as per the unit-for-production method. This means that each boe produced from these fields will generate an amortization of R$14.96 (the price per boe in BRL). Petrobras has confirmed to us that this amortization will be deducted from the taxable profit, thus generating a fiscal benefit. Out of the US$8.4bn of value erosion, we break this down into: i) US$7bn from a different WACC assumption (8.83% for PETR vs. 10% for ESIBR) not recoverable; and ii) US$1.4bn due to different operating assumptions likely recoverable. Petrobras will renegotiate the acquisition conditions after all fields are declared commercially viable and any overly optimistic assumptions will be reviewed to reflect reality. Therefore, we believe that the company may be able to recover this US$1.4bn if the agreed assumptions prove to be too optimistic. Moreover, another upside risk to our valuation is that operating figures might prove to be better after the contract revision, as Petrobras should be able to retain these gains. However, for the time being, we are not including either of these points in our valuation. We also believe that we are being very conservative by not attributing any value to the company s exploration potential. We value the E&P business only through the full consumption of its current total reserves base (3P) plus the resources of the main pre-salt blocks (BMS-11 & BMS-9) and the Onerous Assignment. We believe that this is a very conservative assumption as the company in addition to its rich exploration potential of around 190 exploratory blocks in Brazil should have a competitive advantage in the next bid rounds due to its superior knowledge. However, we reiterate that Petrobras is a state-owned company for better or for worse. On the bright side, we believe that this gives the company privileged access to resources, scale gains, greater geological knowledge and, consequently, better access for future exploration activity. Moreover, we believe that it is important to make it clear that we do not see the company as a direct peer with other National Oil Companies (NOC) rather we see it as more comparable to an Integrated Oil Company (IOC). Unlike a typical NOC, PETR has vast expertise and is fully able to develop its own resources independently. Moreover, PETR is currently the world s largest deepwater oil production player DEEPWATER PRODUCTION Anadarko, 3% BG Group, 3% Reliance, 4% Total, 8% Chevron, 8% BP, 9% ExxonMobil, 10% Source: Petrobras Others, 10% PETR, 22% Statoil, 12% Shell, 11% Page 7 of 61

8 We See High Oil Price Leverage We see Petrobras as a highly leveraged player to oil prices. For each $10/bbl increase in the oil price (from our base case assumption of $80/bbl), our FV increases by 23%. We welcome this exposure given oil s limited availability, especially taking into consideration that higher exploratory and production marginal costs should provide an important support to its long-term price. However, the company s short-term monetization potential is not as clear. Local prices in Brazil for gasoline and diesel (around 60% of total oil products production) only reflect international oil prices with a delay. PETR s pricing policy states that any local price readjustment (whether to the company s advantage or disadvantage) will only occur when international prices stabilize at a new level. This policy is meant to prevent Petrobras from transferring short-term volatility in international oil markets to final customers. However, the policy is not straightforward for investors, as the company has not defined when it considers the market to have stabilized at a new level. During periods when PETR s sale price is below international markets as is currently the case, this policy becomes intensively scrutinized with suggestions of unfair interference by the government featuring widely in the press. Allowing a fuel price increase this year would put even more pressure on inflation (which is already close to the upper limit of the government s target); and thus we believe would not be in line with the government s best interest. Hence, it is hard to argue that there is no political influence over Petrobras policies at all. We recognize that this proximity to the government may delay the immediate full monetization of high oil prices; however, we believe that such an effect has a limited impact on PETR s valuation due to: Based on past evidence (as per the chart on page 26), we believe that the company works with prices linked to international markets in the long term. We believe that the average period of adjustment is around 3 quarters, but this may vary in the event of extraordinary shocks. As local prices decoupled from international prices in 1Q11, we believe the company would be close to pursuing a price increase if markets were more stable. However, we would not be surprised if, on the back of a highly uncertain international scenario, the government and the company prefer to wait a little bit more before announcing any concrete measure; This policy could be viewed as a sign of political interference, but it was implemented several years ago and investors are already aware of it; therefore we believe that it is fully priced in. Moreover, when oil prices fall heavily, this policy usually goes in favour of the company, enabling it to keep prices higher for longer, as occurred in 2009; The company s growth, and thus a meaningful part of its EV, is driven by strong E&P expansion. We estimate Petrobras production should exceed domestic demand by c.1.8mn bpd in Therefore, this oil should be directed to international markets and thus be fully aligned to market prices. We believe that by 2020, exports should account for a much larger share of the company s sales. We estimate exports to reach around 33% of net revenues in 2020 (from the current 17%). High-Quality Workforce Over recent years, Petrobras has invested heavily in developing a high-quality workforce, which has been a key component of its exploration success. The company has its own training center (the so-called Petrobras University ), where some new employees study for more than a year after joining the company. Petrobras success in this area can also be seen in the workforce of other major oil players in Brazil, many of which have PETR s former employees in their top management. The generally acknowledged high quality of its employees, together with the limited remuneration ability of a state-owned company, raises the risk of a talent run to its competitors. While we acknowledge that this could happen, we believe that Petrobras is conscious of this risk and has taken steps to avoid it. Recently, Valor newspaper quoted OSX CEO Mr. Luiz Carneiro, who stated that OSX does not recruit any employee from Petrobras, Page 8 of 61

9 unless the employee has retired or is close to retirement. In our view, this is an indication that other companies in Brazil believe that a talent hunt at Petrobras could have commercial implications and, therefore, they would not be willing to take the chance. Local Content & Brazilian Industry PETR s strong demand for oil equipments and services coupled with the local content requirements is a significant challenge for the company, in our view. In addition to taking on much larger-scale projects, the company is facing stricter local content requirements (LC) than in the past. All concession contracts signed in Brazil require that a minimum level of goods and services from Brazilian providers are used in the exploration, development and production phases. We think the LC requirement is meant to take the place of an industrial policy for the sector. The LC guardian is the Brazilian Oil Agency (ANP). In our recent meetings with ANP, our perception was that, although they will not relax the LC, they will adopt a constructive and active dialogue with the market to make these requirements achievable. Also, we believe that the current methodology is flexible and may be changed in order to better adapt to reality, while keeping its primary target of encouraging the establishment of a solid local industry. However, it is very important to bear in mind that the large fields from the pre-salt were auctioned during the 2 nd and 3 rd Bid Rounds, when LC was below 50%. Therefore, if needed, we think PETR could adopt lower levels for these fields while directing domestic production to newer concessions. Analysis carried out by the Brazilian National Oil industry Organization (ONIP) in Aug/2010 estimated that the oil sector will make investments of US$324bn (and US$76bn of opex) by If we put that together with the LC, it is easy to see that there should be very high demand for oil equipment with a high level of LC over the coming years. Therefore, with an eye on this captive market, domestic and international equipment suppliers have announced several new greenfield and brownfield projects targeting to meet this demand. However, we estimate that the local industry will take at least 5 to 10 years to reach productivity levels anywhere close to those currently seen in Asia. In the meantime, we believe that some projects are likely to be delayed. It is important to note that such delays would not significantly affect our investment case due to: Limited impact on FV: Even if the crown jewel projects are delayed, the impact to our fair value is limited. Each year of delay to the start up of BM-S-11 and BM-S-9 would reduce our FV by around only 3%. We see subsequent fields as less likely to be delayed as we would expect the Brazilian oil equipment industry to be more mature in the second half of this decade; Better productivity than planned: We expect the company to concentrate on the highly productive fields of the pre-salt area, achieving higher productivity levels per producing units than currently guided. Tax Discussions: No Impact on Valuation The Brazilian regulatory framework has been in the spotlight in recent years due to the legal changes that have taken place since the pre-salt discovery. However, independent of the noise around this subject, we highlight that none of the changes has resulted in any contract breach or changes to agreements already in force. The federal government has stated that it is committed to maintaining all conditions of signed contracts and fully respecting their economics. Over recent months, Congress has intensively discussed changes to the current methodology of distributing the government take between various levels of government, as the oil non-producing regions want to receive a higher stake from the oil-related taxes. One of the discussions included a proposal to increase the taxation on oil companies in order to increase the transfer to nonproducing regions while keeping the participation of the producing areas unchanged. However, the federal government has voiced its commitment to respecting the status quo of current concessions and this proposal has not gathered much support thus far. We do not see these discussions as a material risk to our investment case, as they refer only to the tax sharing at government levels. We believe that there is no legal support for altering the Page 9 of 61

10 current taxation environment for oil companies. We base our view on the fact that the concession contracts, as per the draft version available at the time of the 10 th Bid Round, through its Annex V, specifically define the government special participation rates based on decree #2,705 (Aug/03/1998). In contrast with some other countries (e.g. the UK), we believe that these contracts do not foresee any changes to accommodate new legislation. Therefore, even if any changes are implemented, we believe that they would be overruled in the courts. Moreover, as we do not assume any new concessions in our model, any changes to taxation on new areas would have a neutral impact on our valuation. Section 2: Valuation Comparison to Peers In 2010, Petrobras carried out a R$120bn capital increase that led to a 33% earnings dilution. However, this capitalization had no positive effect on the company s results, as it was mostly used to finance its future growth and acquire more producing rights, which will take around five years to mature and start showing the first fruits. Therefore, PETR s short-term valuation (based on P/E or EV/EBITDA), rose far above the sector s average at the time. Since Sep/10, PETR s share price has fallen c.21%, adjusting fully for the capital increase. We estimate that the company is currently trading at a slight 1% discount to the sector average on 2012E P/E. Last year s capital increase has not impacted PETR s earnings yet and the first material fruits should start flowing in from 2014E, therefore we deem the current discount as very appealing with an eye on the forecast strong growth ahead. We also like the higher ratio of oil to gas in PETR s reserves. At YE10, oil represented 84.4% of its SEC proved reserves, vs. an average of 51.1% for Exxon (not rated), BP (not rated), Chevron (not rated), Shell (not rated) and Total (not rated). PRODUCTION COMPARISON (KBOEPD) 7,000 6,000 Petrobras 5,000 Exxon 4,000 BP 3,000 Shell Chevron Total 2,000 1, Source: Bloomberg, Company data and Espirito Santo Investment Bank for Petrobras estimates between Page 10 of 61

11 Methodology We calculate fair values of R$26.7 and R$29.8 for PETR4 and PETR3, respectively, based on a sum-of-the-parts Enterprise Value with DCF valuations for Upstream, Downstream, Gas & Power, Distribution, International Operations and Corporate Overheads. We value the company s stake in Braskem (BRKM5, NR) at the 90-day trading average of the non-voting shares due to higher liquidity and we value other non-consolidated investments at book value. We value the E&P business at a WACC of 10% (nominal USD). We do not assign any perpetuity value to the E&P business as we carry our valuation until the consumption of current 3P reserves (proved + probable + possible reserves) of 25.8bn boe. We also carry out a separate valuation for the 5bn barrels from the Onerous Assignment, block BMS-11 (Lula / Cernambi and Iara) and block BMS-9 (Guara and Carioca). For the remaining business units, we use a WACC of 11.3% and a nominal perpetuity rate of 4.5% (in line with long-term expected inflation). We deduct Net Debt and minorities at FY11 book value to reach an equity valuation for Petrobras. We attribute no value to the tag-along rights of the voting shares as the company is state-owned and we do not expect it to be privatised in the foreseeable future. Moreover, both classes of shares have historically received the same amount of dividends and interest on capital. However, in order to attribute a liquidity premium to the voting shares, we have considered the 1-year average premium of 11.5% of the voting shares over the nonvoting shares to calculate the fair value of the different classes of shares. Cost of Capital (Ke) 14.1% Risk Free Rate 3.0% Equity Premium 5.5% Beta (levered) 1.11 Country Risk (bp) 250 Differential over inflation 2.5 Cost of Debt 7.3% Avg. Cost of Debt 7.3% Tax Shield 34% Debt Ratio 30% Equity Ratio 70% WACC 11.3% Source: Bloomberg and Espirito Santo Investment Bank for estimates. Sum-of-the-Parts (SoTP) and Fair Value The key value driver for Petrobras should be its E&P business, representing 90% of its EV, on our estimates. We note that, after initial successful drilling activity in the pre-salt, the separate valuation of BMS-11, BMS-9 and the Onerous Assignment already represents 38% of our EV. SOTP VALUATION FOR PETROBRAS BRL mn Value % of EV BRL/sh Methodology Implied EV / EBIT 12e MM boe NPV ($/boe) E&P 421,627 90% 32.3 DCF (WACC 10%) , Current Assets (3P reserves ex Santos) 242,375 52% , BMS-11 96,257 20% , BMS-9 14,711 3% 1.1 1, Onerous Assignment 68,284 15% 5.2 5, Dow nstream 34,650 7% 2.7 DCF (WACC 11.3%, g 4.5%) 3.4 G&P 10,289 2% 0.8 DCF (WACC 11.3%, g 4.5%) 4.0 Distribution 20,842 4% 1.6 DCF (WACC 11.3%, g 4.5%) 12.0 International Operations 11,565 2% 0.9 DCF (WACC 11.3%, g 4.5%) 6.8 Corporate Overhead (39,489) -8% -3.0 DCF (WACC 11.3%, g 4.5%) 5.2 Consolidated Assets 459, Braskem 4,900 1% day average of BRKM5 Other Non consolidated Investments 5,829 1% book value EV 470, Net Debt 2011E (95,372) -20% -7.3 Minorities (3,851) -1% book value Equity 370, Fair Value (BRL) - PETR Fair Value (BRL) - PETR # PETR4 shares 5,602 # PETR3 shares 7,442 Total Shares 13,044 Source: Espirito Santo Investment Bank for estimates Page 11 of 61

12 PETR4 FAIR VALUE BREAKDOWN E&P Current Assets BMS-11 BMS-9 Onerous Assignment Downstream G&P Distribution International Overhead Braskem Others R$ / sh Net Debt 2011 Minorities Total Source: Espirito Santo Investment Bank for estimates Sensitivity to Oil Prices We see Petrobras as highly leveraged to oil prices. According to our sensitivity analysis, for each $10/bbl increase in our long-term oil price assumption of $80/bbl, our FV increases by 23%. We note that this analysis measures the sensitivity of the company s E&P business to changes in oil prices while our valuation for the remaining business units remains unchanged. We believe that a change in oil prices is unlikely to have a strong effect on the other parts of the business, which should maintain their cash flow and valuation reasonably unchanged while passing through oil price changes to their clients. We also keep our valuation for the 5bn boe from the Onerous Assignment unchanged; this assumption will be revisited when the government and the company renegotiate overall conditions after all fields are declared commercially viable. OIL PRICE SENSITIVITY ANALYSIS Oil price NPV Current Mkt Cap Upside PETR4 FV Upside PETR3 FV Upside ($/bbl) (R$mn) (R$mn) (%) (R$/share) (%) (R$/share) (%) , , % % % , , % % % , , % % % , , % % % , , % % % , , % % % , , % % % Source: Bloomberg, Espirito Santo Investment Bank for estimates Page 12 of 61

13 Petrobras vs. Peers MULTIPLES Bloomberg Last Market P/E EV/EBIT Div Yield Company Ticker Currency Price Cap (USD m) 2011E 2012E 2013E 2011E 2012E 2013E 2011E 2012E 2013E Shell RDSA NA EUR , % 4.8% 5.0% BP BP/ LN GBp , % 4.7% 5.3% Total FP FP EUR , % 6.0% 6.2% Exxon XOM US USD , % 1.7% 1.8% Chevron CVX US USD , % 2.1% 2.2% Eni ENI IM EUR , % 6.4% 6.5% Majors % 3.9% 4.1% BG Group BG/ LN GBp 1, , % 1.3% 1.5% Repsol YPF REP SQ EUR , % 5.1% 5.6% Galp Energia GALP PL EUR , % 1.3% 1.3% European w ith BR exposure % 2.6% 2.8% Gazprom GAZP RU USD , % 1.9% 2.2% Rosneft ROSN LI USD , % 1.2% 1.6% Petrochina 857 HK HKD , % 4.0% 4.3% China Petroleum 386 HK HKD , % 3.5% 3.7% CNOOC 883 HK HKD , % 3.6% 3.9% Lukoil LKOD LI USD , % 2.9% 2.8% Oil & Natural Gas Corp ONGC IN INR , % 4.4% 4.0% Novatek NVTK LI USD , % 1.2% 1.6% Emerging Markets % 2.8% 3.0% Petrobras PBR/A US USD , % 3.8% 3.8% Source: Bloomberg for prices and consensus for not rated stocks and Espirito Santo Investment Bank for estimates for Petrobras, Repsol YPF and Galp Energia, priced as at 27 October PROVEN RESERVES SEC (BN BOES) 2010 OIL AND GAS PRODUCTION (MM BOEPD) Exxon BP Shell Petrobras Total Chevron 0.0 Exxon BP Shell Petrobras Total Chevron Oil Gas Oil Gas Source: Company data Source: Company data Page 13 of 61

14 CAGR COMPARISON EBITDA EPS 14.0% 12.0% 10.0% 8.0% 6.0% 6.7% 8.2% 8.1% 12.1% 14.0% 12.0% 10.0% 8.0% 6.0% 9.4% 10.2% 6.2% 10.7% 12.2% 4.0% 2.0% 2.4% 4.0% 2.0% 0.0% Exxon Chevron Shell Total Petrobras Source: Petrobras 0.0% Exxon Source: Petrobras Chevron Shell Total Petrobras PRODUCTION 6.0% 5.1% 5.0% 4.0% 3.0% 2.0% 1.0% 1.0% 0.4% 0.1% 0.0% 1.0% 1.0% 2.0% Exxon Chevron Shell Total Petrobras Source: Petrobras RESERVES 3.0% 2.7% 2.5% 2.0% 1.5% 1.5% 1.0% 0.5% 0.0% 0.5% 0.2% 0.1% 1.0% 0.9% 1.5% Exxon Source: Petrobras Chevron Shell Total Petrobras SHARE PRICE PERFORMANCE - CAGR 00-CURRENT SHARE PRICE PERFORMANCE YTD (%) 25.0% 20.0% 20.7% 30.0% 20.0% 14.0% 22.2% 15.0% 13.9% 10.0% 0.0% 8.3% 7.2% 10.0% 9.1% 8.2% 10.0% 5.0% 2.4% 20.0% 0.0% XOM US CVX US RDSA NA TOT US PBR US Source: Bloomberg. * RDSA NA refers to trading prices in Netherlands as ADR was listed only in % 27.3% XOM US CVX US RDSA NA TOT US PBR US Source: Bloomberg. * RDSA NA refers to trading prices in Netherlands as ADR was listed only in We believe that Petrobras historical outperformance reflects its superior growth to oil majors. In our view, the stock s underperformance YTD fully prices in last year s capitalization. Therefore, at the current level, we see a good opportunity to buy shares post the dilution effect, with a rich assets portfolio that should allow the company to continue posting above-average growth and, consequently, superior relative share-price outperformance to peers. Page 14 of 61

15 Section 3: Detailed Valuation: Going Through the Business Units Exploration & Production (E&P) Petrobras Brazilian E&P operations are its key value driver, in our view. We value the company s E&P unit at R$422bn, which represents 90% of our EV. The implied EV / EBIT 12E for E&P at 10.4x may appear high at a first glance; however, it is important to highlight that this reflects the EV contribution from the BMS-11, BM-9 and the Onerous Assignment, as these assets are still in the exploration/development phases with no material earnings contribution expected before If we adjust the multiple to consider an EV/EBIT of the current assets that are currently generating cash flow, we reach a more appealing figure of 6.3x. We carried out a separate valuation for the company s largest projects: BMS-11, BM-9 and the government s 5bn boe Onerous Assignment. We assume oil prices converging to $80/bbl in 1Q12 (average Brent of $107.3/bbl in 2011, $81.3/bbl in 2012 and flat $80/bbl from 2013 on). E&P EV BREAKDOWN Current Assets 57% BMS-11 23% Source: Espirito Santo Investment Bank for estimates BMS-9 4% Onerous Assignment 16% E&P CONSOLIDATED FREE CASH FLOW (US$MN) $mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 24,814 26,202 28,932 32,912 39,895 48,103 54,540 59,799 62,874 Taxes on EBIT 8,437 8,909 9,837 11,190 13,564 16,355 18,543 20,332 21,377 NOPLAT 16,377 17,293 19,095 21,722 26,331 31,748 35,996 39,467 41,497 Depreciation & Amortization 7,274 8,116 8,893 9,676 10,944 12,666 14,007 15,027 15,731 Capex (24,336) (23,907) (23,486) (23,088) (33,867) (31,229) (27,507) (24,072) (24,170) in Working Capital 265 (186) (254) (279) (288) (377) (360) (332) (207) Free Cash Flow (419) 1,316 4,248 8,031 3,120 12,808 22,136 30,091 32,852 Source: Espirito Santo Investment Bank for estimates ANNUAL DOMESTIC OIL PRODUCTION (MMBBL) E&P ANNUAL FREE CASH FLOW & CAPEX ($BN) 2,000 80,000 1,800 1,600 60,000 1,400 1,200 40,000 1, , , ,000 Current Assets BMS 11 BMS 9 Onerous Assignment (OA) Source: Espirito Santo Investment Bank for estimates Source: Espirito Santo Investment Bank for estimates E&P Current Assets We value E&P Current Assets at US$149bn (52% of our EV). We calculate PETR s current assets as its total reserves base (3P) deducted from the reserves already incorporated from the BMS-11, which was already declared commercially viable in Dec/2010. We estimate that its total reserves portfolio excluding Santos Pre-Salt is 25.8bn boes. We do not attribute any perpetuity value to this business unit. We reach an EV for E&P of US$149bn through the full production of its reserves portfolio by It is important to highlight that we are not attributing any value to the company s exploration potential as we value only the existing reserves base. We consider Page 15 of 61

16 that all necessary capex to develop the non-developed reserves will be disbursed by We apply a development capex of US$6/boe. E&P CURRENT ASSETS CASH FLOW (US$MN) $mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 23,780 24,296 26,086 27,721 29,305 30,160 30,127 29,771 29,193 Taxes on EBIT 8,085 8,261 8,869 9,425 9,964 10,254 10,243 10,122 9,926 NOPLAT 15,694 16,035 17,217 18,296 19,341 19,906 19,884 19,649 19,268 Depreciation & Amortization 7,035 7,548 7,928 7,960 7,449 7,295 7,124 6,895 6,964 Capex (23,431) (19,784) (13,435) (6,911) (16,331) (16,732) (17,073) (17,345) (17,498) in Working Capital 265 (186) (254) (279) (288) (377) (360) (332) (207) Free Cash Flow (436) 3,614 11,456 19,065 10,171 10,092 9,575 8,867 8,526 Source: Espirito Santo Investment Bank for estimates E&P CURRENT ASSETS EV BREAKDOWN US$mn R$mn ,787 87, , ,865 E&P Current Assets Value 148, ,375 Source: Espirito Santo Investment Bank for estimates PETROBRAS DOMESTIC RESERVES (SPE / ANP CRITERIA) Mmboe Oil 12,726 12,745 13,508 14,419 15,573 17,413 21,079 19,634 20,926 24,857 YoY - 0.1% 6.0% 6.7% 8.0% 11.8% 21.1% -6.9% 6.6% 18.8% Proved Reserves 8,322 9,557 10,613 10,977 11,365 11,671 11,802 11,900 12,057 12,909 Probable + Possible Reserves 4,405 3,189 2,895 3,442 4,209 5,742 9,277 7,735 8,870 11,948 Natural Gas 2,030 2,099 2,850 3,125 2,770 3,523 4,493 3,896 3,548 4,571 YoY - 3.4% 35.8% 9.6% -11.3% 27.2% 27.5% -13.3% -8.9% 28.8% Proved Reserves 1,349 1,452 1,989 2,045 1,868 2,082 2,118 2,193 2,112 2,374 Probable + Possible Reserves , ,441 2,375 1,703 1,436 2,197 Total 14,756 14,845 16,358 17,544 18,343 20,936 25,572 23,530 24,474 29,428 YoY - 0.6% 10.2% 7.2% 4.6% 14.1% 22.1% -8.0% 4.0% 20.2% Proved Reserves 9,670 11,009 12,602 13,023 13,232 13,753 13,920 14,093 14,169 15,283 Probable + Possible Reserves 5,086 3,836 3,757 4,521 5,111 7,183 11,652 9,437 10,305 14,145 Source: Petrobras for proved reserves and Espirito Santo Investment Bank estimates based on ANP data for Probable and Possible reserves Page 16 of 61

17 PETR S RESERVES BASE WITHOUT SANTOS PRE-SALT (MN BOES) PETR's Proved Reserves (SPE)-BRAZIL 15,283 a PETR proved reserves / BZ Proved reserves 92.9% b PETR's Proved Reserves from the Santos Pre-Salt* 1,071 c PETR's Proved Reserves ex Santos Pre-Salt 14,212 d = a - c Brazil Total Reserves (SPE) by ANP 33,333 e Total reserves from Lula / Cernambi (based on Galp) 5,540 f PETR's stake in Lula / Cernambi 65% g Brazil Total Reserves ex-santos Pre-Salt 27,793 h = e - f PETR's Total Reserves PETR ex-santos Pre-Salt 25,827 i = h x b PETR Reserves Lula / Cermanbi 3,601 j = f x g PETR's Total Reserves 29,428 k = i + j Source: Petrobras Galp, ANP and Espirito Santo Investment Bank *This figure is available in the company's annual reserves publication; E&P CURRENT ASSETS RESERVES CONSUMPTION (MM BOE) 30,000 25,000 20,000 15,000 10,000 5, Source: Espirito Santo Investment Bank for estimates In terms of the oil realization price, we assume a discount to Brent in line with the historical average. We apply a US$9.52 / bbl discount to Brent prices, 11.9% below our assumption of Brent at US$80 / bbl. For the gas realization price, we are adopting a conservative approach due to a lack of visibility, leaving prices in the long run at US$13.31 / boe (c.us$2.29 / MM BTU). This is 16% above the 2Q11 price but 15.3% below the 2010 average. The gas realization price dropped 55.2% between 2Q10 and 1Q11. According to the company, this partially reflects the need to remunerate the G&P division for past investments; however, there is no guidance for future performance. We may revisit this assumption based on future performance or clear guidance from the company. We estimate lifting costs of US$10-11/boe. Historically, these costs have been between 10% and 15% of the oil realization price. We estimate lifting costs to stay at the higher end of this range in our projections. Over recent quarters, ANP has tightened inspections and regulatory requirements after the oil spill in the Gulf of Mexico. These tighter requirements have been pushing lifting costs up; nevertheless, we expect to see a learning curve and an LIFTING COSTS (US$/BOE) Lifting Cost ($/bbl) Lifting Cost / Oil Realization Price Source: Espirito Santo Investment Bank for estimates 16% 15% 14% 13% 12% 11% 10% Page 17 of 61

18 accommodation below the recent levels of US$13.12/boe and US$11.38/boe in 2Q11 and 1Q11, respectively (but no return to the previous levels of US$8-9/boe). BM-S-11 & BM-S-9 These blocks are the key value drivers in our valuation, adding US$68bn to our EV (24% of the total). We have carried out a separate DCF valuation for each block in order to gain a better insight into their economics. Our analysis is largely based on BG Group s recent guidance. (BG is an important partner in both blocks, as can be seen in the charts to the right.) As a partner in the operating consortium, BG has access to the same information as Petrobras. Although PETR is the operator of both blocks, it has maintained its guidance virtually unchanged even after intense and successful exploration activity, while its partners have adopted a more active dialogue with the market, upgrading their estimates in order to more accurately reflect the data gathered so far. Therefore, we see BG s latest guidance as providing the most up-to-date figures for both blocks. BMS-11 OWNERSHIP BG, 25% Galp, 10% PETR, 65% KEY HIGHLIGHTS FROM BM-S-11 AND BM-S-9 Resources (M boe) Liquids Gross Petrobras net NPV NPV weight First # # Blocks Total 27 years Stake (M boe) $/boe Usd m output Oil FPSO wells BM-S-11 18,000 17,049 65% 11, ,231 83% BM-S-9 3,610 3,610 45% 1, ,032 83% Total 12, , Source: Espirito Santo Investment Bank for estimates Source: Espirito Santo Investment Bank BMS-9 OWNERSHIP Repsol, 25% BG, 30% PETR, 45% Following the extensive appraisal activity carried out in blocks BM-S-11 and BM-S-9, BG has doubled its guidance for mean total reserves and resources (P50) in Brazil to 6bn boe, with 95% of this total accounted for by the blocks BM-S-11 (Lula, Cernambi, Iara) and BM-S-9 (Guara and Carioca). According to our calculations, BG s new guidance implies total resources for Block BM-S-11 of 18bn boe and for BM-S-9 of 3.6bn boe, which compares with PETR s official guidance of 11.8bn boe and 1.55bn boe for BM-S-11 and Guara, respectively. Source: Espirito Santo Investment Bank RESOURCES BASE: ESIB VS. PETR VS. BG ESIBR PETR BG Group Implied bn boes Stake SoP Guidance (Mid) Guidance (Mid) Tupi / Cernambi 100% Iara 100% Total Block BM-S % Block BM-S-9* 100% * Petrobras discloses guidance only for the reservoir of Guara, while BG includes Carioca. We include both reservoirs. Source: BG Group, Petrobras and Espirito Santo Investment Bank for SOTP BG and Galp have also made very positive comments on project economics. Both companies have said that the data gathered from Lula s pilot production system, the EWTs at Lula and Guara, and from additional appraisal activity have highlighted that the reservoirs have a deeper oil water contact and a better reservoir connectivity and continuity than initially expected, especially in the flanks for Lula. The latter should allow for increased well productivity, a faster ramp-up of production and a longer period at plateau level. Thus, despite the significant upward revision of the resources guidance for blocks BM-S-11 and BM-S-9, BG says it believes it will not need to increase the overall capex of the projects, as the already planned facilities (FPSOs and wells) should be able to deliver much higher productivity. This should have very positive implications in terms of capex/boe and opex/boe. We highlight that initial production at the Lula pilot project looks very promising, with the well 9BRSA716RJS showing the highest production in Brazil since May/11, after the company dealt with the gas flaring restrictions. Page 18 of 61

19 PRODUCTION AT WELL 9BRSA716RJS (LULA S PILOT PRODUCTION SYSTEM) 40,000 35,000 30,000 25,000 boepd 20,000 15,000 10,000 5,000 0 Nov-10 Dec-10 Jan-11 Feb-11 Mar-11 Apr-11 May-11 Jun-11 Jul-11 Aug-11 Total Oil Gas Source: ANP Therefore, we have a resources estimate for Block BM-S-11 (Lula, Cernambi and Iara) of 18bn boe, from which we expect PETR to recover c.17bn boe during the 27-year concession period. We estimate peak well flow rates of 30kboepd, opex/boe (ex-fpso) of US$5/boe, a production plateau period of 8 years and a declining rate of 7%. We have not built in any potential de-bottlenecking of the FPSOs, which BG estimates could enable them to produce between 10% and 20% above their nominal capacity at a later stage of development. Thus, we are working with 17 FPSOs to support the estimated production volumes. Overall, our capex/boe (ex-fpso) for this block is at US$3.7/boe. We apply a discount to Brent prices of 5% (vs. a historical average of around 12%), as we expect a better quality oil of between 28º and 30º (vs. the current average for Brazil of between 23º and 24º). Moreover, we expect natural gas realization prices of US$6/ MM BTU over the long term. As the sale of this gas will be through new sales agreements, we expect sales prices to be closer to international levels. Petrobras has already secured 9 FPSOs for the development of Lula and Cernambi and therefore we believe that visibility over the production ramp-up of these two fields is quite high. The pilot production system has been in operation for close to a year, the second FPSO (Lula Northeast) was already chartered and should start operations in 2013, and the contract for the third vessel (Cernambi South) was just signed in time to allow operations to start in The remaining 6 units will come from the order of 8 replica FPSOs whose hulls are already being built in Brazil by local engineering company Engevix. The contracts for the construction and integration of the topside modules for these 8 FPSOs should be awarded gradually before the end of 2012 so that the vessels can be delivered starting between 2015 and We believe that given the lack of track record of Brazilian shipyards in executing such a large order there are some risks of delays in the delivery of the 8 FPSOs and of some cost overruns. However, we highlight that the construction of the hulls is so far on time and according to Engevix the US$3.5bn contract is not subject to revisions due to cost overruns. We have also built in a buffer on top of the overall FPSO cost guided by Petrobras of US$1.5bn per unit (we assume a total cost of US$1.8bn). LULA/CERNAMBI DEVELOPMENT Source: Galp IARA S WELL LOCATIONS Source: Galp Regarding Iara, visibility on the development of this block is lower, although we believe that the recent successful appraisal results of the Iara Horst prospect have put it in a good position in the pre-salt merit order. According to the partners, Iara Horst had better reservoir characteristics in terms of permeability and porosity than those found with the original Iara wildcat drilled in 2008, pointing to very positive well productivity. Another appraisal well should be drilled to the north-west of Iara in early 2012 and an EWT is planned for 2013, ahead of the deadline for the declaration of commerciality in December We have assumed that the first FPSO at Iara will only be deployed in Iara is not part of Petrobras investment plan up to Page 19 of 61

20 Our development model for Iara assumes the deployment of 5 FPSOs between 2016 and 2020 and incorporates operating metrics similar to Lula/Cernambi. The lower NPV/boe compared with Lula reflects our expectation that the Iara development will start two years later. BMS-9 is the second block in line to start production in the pre-salt of the Santos Basin. This block is divided into the evaluation areas of Carioca and Guara and their deadlines for declaration of commerciality are 11/Nov/2011 and 31/Dec/2012, respectively. There is no official guidance for the overall resources potential of this block, but both Repsol and BG Group have provided information in the past that leads us to estimate overall recoverable resources at 3.6bn boe: The official guidance for the Guara structure is that it holds between 1.1bn boe and 2.0bn boe (1.55bn boe at the mid-point). In November 2010, BG disclosed an internal resource forecast narrowing the range to between 1.5bn boe and 2.0bn boe (1.75bn boe at the mid-point), but neither Repsol nor Petrobras has confirmed this new guidance; BLOCK BM-S-9 Source: Repsol Repsol initially guided for the Carioca discovery alone to have between 650mn boe and 850mn boe. Repsol has never provided guidance for the Abaré West and Iguaçu Norte discoveries; According to the appraisal report carried out by DeGolyer and MacNaughton (D&M) for Repsol Brasil, the two structures within the Block BM-S-9 had overall 3C resources (P10) of 3.6bn boe as of June 2010; According to our estimates, BG s new net resources guidance for Brazil of 6.0bn boe, announced at the end of June 2011, implies a P50 resource estimate for Block BM-S- 9 of 3.7bn boe, with P90 at 2.6bn boe and P10 at 4.7bn. The latter is above the 3.6bn boe certified by DeMac one year earlier, which in our view reflects the extensive appraisal activity carried out both at Guara and Carioca in the past twelve months. Visibility over the development calendar of the Guara reservoir is high as the consortium has already chartered the two FPSOs that will be deployed in this block. The first FPSO (150k boepd) has been chartered from Schahin/Modec and will be located in the original Guara Sul discovery. The EWT at Guara South has recently been concluded and production is expected to start in 1Q13. The second FPSO (200kboepd) has been chartered from Q.Galvão/SBM and is expected to start production in 3Q14 at the Guara Norte discovery. The formation tests at both Guara Sul and Guara Norte have pointed to high productivity levels for this structure, anticipating peak well flow rates of up to 50k boepd. The extension well drilled 5.7km south of the wildcat discovery at Guara Sul also confirmed the hydraulic communication of the reservoir between both wells, which demonstrates excellent connectivity in the reservoir and enhances the estimates of hydrocarbon volumes connected to the Guara Sul. The Carioca evaluation area encompasses three separate structures that have already been successfully drilled, namely Carioca (Sep07), Iguaçu Norte (Apr09) and Abaré West (Sep09), and two additional prospects that have not been drilled yet, namely Abaré and Iguaçu Complex. The first FPSO at Carioca is scheduled for the end of 2015 and is one of the 8 clone FPSOs with a capacity of 180k boepd. Following the recent decision to charter the second FPSO for Guara, we believe the Carioca evaluation area is also likely to receive the second clone FPSO that was attributed to Block BM-S-9, although this will likely depend on the results of the appraisal to be carried out ahead of the declaration of commerciality in Early in 2011, Petrobras and its partners announced that they had made another discovery in the evaluation area of Carioca well, 9 kms north-east of the original discovery. Preliminary analysis confirmed that Carioca Nordeste yielded better results than the original Carioca find, with light oil being found in 250 metres of good quality reservoir rock. In 4Q11, Petrobras plans to spud the Carioca Sela appraisal well in the middle of the Carioca formation to provide further clarity on this structure. The first EWT should start before the end of 2011 using the same small FPSO used at the EWT of Guara. Petrobras is also drilling the Abare prospect to test the extension to the east of the original Abare West discovery. The results should be known before the end of the year and Page 20 of 61

21 should provide more visibility over the resources potential of the southern section of the Carioca evaluation area. Overall, we value 3.6bn boe of resources on Block BM-S-9 at an average US$5.6/boe, slightly above the US$5.3/boe of BM-S-11. However, bear in mind that Block BM-S-11 is more critical for Petrobras as it has larger recoverable resources (17bn boe vs 3.6bn boe) and moreover Petrobras has a bigger stake (65% vs 45%). Onerous Assignment We value the 5bn boe producing rights received during last year s Onerous Assignment at US$42bn (15% of our EV). As we will detail below, we arrive at our base-case valuation through two key value drivers: i) Separate DCF analysis for each block assigned during the deal; ii) Fiscal benefits from the R$74.8bn intangible that was recorded on the company s balance sheet due to the producing rights acquisition. ONEROUS ASSIGNMENT VALUATION (US$) BASE CASE NPV Resources NPV $/boe M boe $mn Total Producing Rights ,000 37,295 Fiscal Benefits ,000 4,676 ESIBR Base Case Valuation ,000 41,970 Source: Espirito Santo Investment Bank for estimates In 2010, the company closed an oil-for-shares agreement with the Federal government, issuing the equivalent to R$74.8bn ($42.5bn) in new shares for rights to produce 5bn boes in specified pre-salt areas close to current concessions. The agreement grants rights similar to those of a regular concession, but with some key differences. Beyond being a contract with guaranteed volumes, it has a price adjustment mechanism. The price was negotiated taking into consideration several assumptions (oil price, production curve, costs, discount rate). After the declaration of commerciality of the last block (exploration phase of 4 years, extendable for 2 additional years), the government and the company, with more data at their disposal, will renegotiate the price. All assumptions are subject to revision with exception of the discount rate of 8.83% (vs. our discount rate for the E&P division of 10%). If the value is higher, Petrobras might pay the difference or ask for a reduction in the volume. If the value is lower, the government will pay the difference to Petrobras. Other differences include: Production is exempt from the Special Participation Tax (SPT). We note that this tax can reach up to 40% of a field s EBIT, so the cash flow generation of this agreement is higher than a regular concession. This exemption was defined during the negotiations between the company and the government; therefore, this difference was already taken into account in the price paid by Petrobras. In practical terms, it is as if the company had paid the SPT upfront; Exploration program: If the minimum exploration program is not concluded within the stipulated period, the penalty should be restricted to a fine without any risks of losing the blocks; The term of the agreement is 40 years, which may be extended for an additional 5 years. A regular concession term is the exploration phase (variable) plus 27 years for production. We have evaluated these 5bn boe production rights with a discounted cash flow per block, largely based on the information publicly available for each block from the consulting companies Page 21 of 61

22 involved at the time of negotiations between the government and PETR (DeGolyer and MacNaughton, and Gaffney, Cline & Associates). We note that we are using a discount rate (10%) that we understand to be in line with the industry standard. Moreover, we consider that all FPSOs will be acquired ($1.8bn per unit), treating the related disbursements as Capex. We arrive at a NPV/boe of $7.46. NPV OF 5BN BOE TRANSFERRED FROM BRAZILIAN GOVERNMENT TO PETROBRAS NPV/boe $ Resources NPV Output Mix First Weighted Peak flow rate Output/FPSO Output/well capex opex M boe $mn oil gas oil First oil (kboepd) (M boe) (M boe) $/boe $/boe Guara South ,904 80% 20% Florim ,486 80% 20% Tupi NE ,641 80% 20% Tupi South ,036 80% 20% Tie Back Tupi Exploration ,342 11,066 80% 20% Franco ,058 22,916 80% 20% , Iara Surround ,313 80% 20% Contingent ,658 26,228 80% 20% Total ,000 37,295 80% 20% Source: Petrobras and Espirito Santo Investment Bank for estimates Our NPV/boe is 12.3% below the official price of US$8.51. However, if we adjust this figure by the time value of money using the official discount rate of 8.83%, we reach an adjusted figure of US$10.08/boe for 2012 (bringing the difference to 26%). This suggests that the deal would have eroded value to the tune of US$13.1bn. Official price vs. ESIBR valuation Official Price Official Price ESIBR Difference Value value $/bbl % ($bn) Guara South % (1,096) Florim % (498) Tupi NE % (688) Tupi South % (154) Exploration % (2,437) Franco % (9,826) Iara Surround % (823) Contingent % (10,649) Total % (13,086) Source: Petrobras and Espirito Santo Investment Bank for estimates We have also carried out an alternative scenario analysis to determine how our valuation would compare with the official price if we were to use the same discount rate of 8.83% used by Petrobras. Page 22 of 61

23 OFFICIAL PRICE VS. ESIBR AT 8.83% WACC Official Price Official Price 2012 Difference Value value value (8.83% w acc) $/bbl % ($bn) Guara South % (713) Florim % 110 Tupi NE % (144) Tupi South % 12 Exploration % (735) Franco % (5,788) Iara Surround % (112) Contingent % (5,899) Total % (6,635) Source: Petrobras and Espirito Santo Investment Bank for estimates In this alternative scenario, our NPV/boe would be US$8.75, still 13.2% below the official price, but reducing the value eroded from US$13.1bn to US$6.6bn. However, we note that the R$74.8bn paid to the government for these rights was registered on the company s balance sheet in 3Q10 as an intangible to be amortized as per the unit-forproduction method. This means that each boe produced from these fields will generate amortization of R$14.96 (the price per boe in BRL). Petrobras has confirmed to us that this amortization will be deducted from the taxable profit, thus generating a fiscal benefit. In order to measure the value generation of the fiscal benefit, we have carried out a separate DCF analysis over the production period of the 5bn boes (2015 to 2051). We assume: i) amortization per barrel of R$14.96; ii) a tax rate of 34%; and iii) WACC of 10%. The value contribution is the present value of 34% of the annual amortization. We arrive at a present value for the fiscal benefit of R$7.6bn (c. $4.7bn). PRESENT VALUE WITH FISCAL BENEFIT FROM AMORTIZATION BASE CASE NPV Resources NPV/boe $mn M boe $/boe ESIBR Valuation 37,295 5, Fiscal Benefits 4,676 5, Adj Valuation 41,970 5, Official 2012 value 50,380 5, Value Erosion -8,410 5, Source: Petrobras and Espirito Santo Investment Bank for estimates The fiscal benefit increases our NPV/boe by 12.5%, reducing the value eroded to $8.4bn in our base case scenario. In order to understand the reason for this value erosion better, we carried out an alternative scenario, applying the company s WACC and including the fiscal benefit in our calculation. PRESENT VALUE WITH FISCAL BENEFIT FROM 8.83% WACC NPV Resources NPV/boe $mn M boe $/boe ESIBR 8.83% w acc 43,746 5, Fiscal 8.83% w acc 5,285 5, Adj 8.83% w acc 49,031 5, Official 2012 value 50,380 5, Value Erosion -1,350 5, Source: Petrobras and Espirito Santo Investment Bank for estimates Page 23 of 61

24 Under this alternative scenario the valuation gap narrows materially, reducing the value eroded by $7bn to $1.4bn. Therefore, the US$8.4bn value eroded considered in our base case scenario can be explained by two factors: i) US$7bn due to different discount rate assumptions; ii) $1.4bn due to different operating assumptions. As we do not have access to the assumptions used by the company and the government to reach the final price, we cannot determine the main reason for this difference as there are too many factors involved (e.g., we use a capex per well of $200mn vs. $160mn considered by D&M and $180mn by Gaffney). However, it is very important to bear in mind that if the initial assumptions prove to be too optimistic, they will be adjusted when conditions are renegotiated in a few years. Four to five years from now, both parties will likely have a much clearer view of actual opex / capex as by this time there will be some fields operating in the pre salt. Consequently, we think it is highly likely that the $1.4bn gap we calculate due to different operational assumptions would not be an issue, as our assumptions may prove be too conservative or the company could renegotiate with the government. ONEROUS ASSIGNMENT BLOCKS Source: Petrobras We have valued all projects on a standalone basis Tupi South was the only project that assumed a tieback to a FPSO in the southern corner of Tupi. However, as we can see in the map above, the areas are adjacent to other concessions of PETR; therefore, we think it is highly likely that infrastructure optimization could lead to better-than-forecast economics. We expect Petrobras to try to maintain individual valuations at the time of renegotiation, which would unlock value when the development of the areas starts and the company is able to benefit from the infrastructure in place. Nevertheless, we do not expect to have visibility on this potential upside risk anytime soon, as the company will likely pursue further infrastructure optimization only after the definitive conditions are set, allowing Petrobras to keep these gains. Therefore, we prefer not to assume such gains in our valuation for the time being. Downstream We value PETR s downstream business at R$35bn, which represents 7% of our EV and an implied 2012E EV/EBIT of 3.4x. We believe that the downstream valuation is affected by its high capex needs over the coming several years. We forecast the company to turn cash flow positive only in 2019, when its last planned new refinery, Premium I, is scheduled to start up. We believe this segment s main value driver is the company s ability to remain fully aligned to international prices and maintain its margins. Page 24 of 61

25 DOWNSTREAM CASH FLOW R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 10,314 12,097 13,626 13,578 13,049 15,183 17,186 19,075 21,930 Taxes on EBIT 3,507 4,113 4,633 4,617 4,437 5,162 5,843 6,485 7,456 NOPLAT 6,807 7,984 8,993 8,961 8,612 10,021 11,343 12,589 14,474 Depreciation & Amortization 3,435 4,108 4,765 5,412 6,029 6,620 7,241 7,725 7,863 Capex (24,624) (24,904) (25,196) (25,488) (24,207) (25,827) (27,314) (14,483) (7,957) in Working Capital 775 (562) (768) (834) (847) (1,113) (1,107) (1,082) (722) Free Cash Flow (13,607) (13,375) (12,206) (11,949) (10,413) (10,300) (9,837) 4,750 13,657 Discounted Cash Flow (13,607) (12,015) (9,850) (8,661) (6,781) (6,025) (5,169) 2,242 5,791 Source: Espirito Santo Investment Bank for estimates DOWNSTREAM EV BREAKDOWN (R$MN) (54,075) Terminal Value 88,724 Dow nstream Value 34,650 Source: Espirito Santo Investment Bank for estimates Fuel sales volume has grown at a higher rate than GDP over recent quarters. We attribute this trend to an increase in the average purchasing power of Brazilian consumers over the past years. Based on our expectations for still-strong personal income performance, higher penetration of car ownership and a historically low unemployment level, we estimate that sales volume should continue to slightly outperform Brazilian GDP growth until We expect sales volume growth to gradually decelerate from 10.7% in 2010 to 4.5% in 2013E, remaining close to this level until 2015E and stabilising after that at around 3.8% (0.85x our GDP growth forecast for the same period). In terms of average realization prices for oil products, we expect Petrobras to follow international prices starting from It is important to note, however, that we are expecting oil prices to converge to US$80/bbl already next year. Therefore, aligning domestic and international prices could actually mean a 7.9% and 12.0% reduction for gasoline and diesel prices, respectively. We believe that this is a conservative assumption, as in the past the company has been able to keep prices unchanged for several months before aligning them with international markets. However, for the sake of conservativeness and owing to lack of full visibility on the company s pricing policy, we prefer to use this assumption. Around 35% of Petrobras production (jet fuel, naphtha, fuel oil and others) is instantly linked to international prices. However, it is for the remaining 65% (mainly diesel and gasoline) that there is more uncertainty. The company says that it only passes through price changes to these products when international prices are stabilised at a new level. However, the definition is not precise. We ran an exercise to check if our assumption of international prices looks reasonable. In the chart below, we compare the behaviour of oil prices and the spread between Petrobras oil products realization price and prices in the reference market of the Gulf of Mexico (GoM). In order to calculate this spread we add the crack spread 321 to WTI prices, reaching the average realization price in GoM. Then we subtract the reported realization price of Petrobras, arriving at the spread. TOTAL REFINING CAPACITY (KBPD) 3,500 3,300 3,100 2,900 2,700 2,500 2,300 2,100 1,900 1,700 1, Source: Petrobras and ESIBR estimates for 2011 to 2020 OIL PRODUCTS SALES (KBPD) 3,100 2,900 2,700 2,500 2,300 2,100 1,900 1,700 1, Source: Petrobras and ESIBR estimates for 2011 to 2020 We believe that the results of this simulation confirm our assumption of internationally linked prices in the long run. During normal times (oil prices between US$70 and US$90/bbl), the spread was reasonably stable, averaging $2.5/bbl. Petrobras prices underperformed when oil prices were at their peak; however, we note that the periods of high prices have historically not lasted long and have generally been caused by some sort of shock, going against PETR s policy of passing through international prices only when they stabilize at a new level. Page 25 of 61

26 On the one hand, this fuel price policy reduces PETR s short-term leverage to oil prices. On the other hand, if oil prices fall heavily, the company may be able to keep prices unchanged for longer, which we see as an upside risk to our valuation of this business unit. SPREAD (GOM - PBR) VS. OIL PRICES Mar-08 Jun-08 Sep-08 Dec-08 Mar-09 Jun-09 Sep-09 Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Brent ($/bbl)-average GoM -PBR-Spread Average Spread with oil between Source: Petrobras, Bloomberg and Espirito Santo Investment Bank estimates We estimate higher refining costs in the long term (R$10.11/bbl in 2020E from R$7.57 in 2010) to reflect: i) recent pressure that, in our view, is not fully explained by the recent maintenance stoppages; and ii) the higher complexity of the new refineries starting up over the next several years. Gas & Power We value the G&P business at R$10bn, equivalent to 2% of our EV and an implied EV/EBIT 12E of 4x. This business unit is the sole buyer of the natural gas produced by the E&P arm. Therefore, we forecast the company s gas acquisition cost over the next few years to be at U$2.29/MM BTU; however, as long as the production from the pre-salt picks up, its average acquisition price should significantly increase, as we estimate that the new contract to be signed will be closer to international prices. We assume US$6.0 / MM BTU for the gas coming for these new fields. Meaningful deviations to our estimated transfer price from E&P to G&P have a neutral impact on our valuation as it is offset in consolidated terms. In terms of its sale price, we expect some readjustment from current levels (we estimate an average sale price of US$9.07 / MM BTU in 2011) to around US$6.6 / MM BTU over the next few years in order to reflect lower oil prices. We believe that the company could potentially pursue a higher sale price in order to pass through higher costs; however, it is not clear to us if the sale agreements in force will allow this change. Therefore, we assume stable prices. We also note that the company has a Take or Pay contract that obliges a minimum purchase of 24 MM cm / day from Bolivia. The agreement is valid through While we forecast that the Bolivian supply is key to guaranteeing domestic demand in the short term, we believe that it will generate an oversupply between 2016 and We estimate that this oversupply will be exported. However, due to the logistic costs involved in liquefying and exporting the gas, we are attributing no margin to this process, and we consider the export price at $6.21 / MM BTU. This arrangement leads to strong pressure on the company s operating results between 2016 and We think that the situation should be normalized in 2020 with the end of the Take or Pay contract. Our sales volumes assumptions are, however, more conservative than Petrobras. Therefore, if demand exceeds our forecast, we should not see any oversupply. NATURAL GAS SALES ( 000 M3) 100,000 90,000 80,000 70,000 60,000 50,000 40, Source: Petrobras and ESIBR estimates for 2011 to 2020 GAS COST/SALE PRICE ($/MM BTU) Avg Gas Cost Avg Gas Sale Price Source: Petrobras and ESIBR estimates for 2011 to 2020 Page 26 of 61

27 We project demand from thermoelectric between 2011E and 2014E to grow in line with the Brazilian National System Operator s estimate of c.16% pa. During this period, we expect demand ex thermos to grow by 1.5x GDP. After 2014E, we expect a growth rate of 1x GDP. G&P CASH FLOW R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 2,601 2,788 2,970 3,048 2,534 1, ,450 Taxes on EBIT ,010 1, ,173 NOPLAT 1,717 1,840 1,960 2,012 1,673 1, ,277 Depreciation & Amortization 1,658 1,753 1,852 1,947 2,032 2,095 2,143 2,177 2,197 Capex (4,440) (4,491) (4,544) (4,596) (3,894) (3,658) (3,142) (3,001) (2,308) in Working Capital 61 (50) (71) (78) (78) (99) (98) (96) (64) Free Cash Flow (1,005) (949) (802) (715) (267) (603) (636) (801) 2,103 Discounted Cash Flow (1,005) (852) (647) (519) (174) (353) (334) (378) 892 Source: Espirito Santo Investment Bank for estimates G&P EV BREAKDOWN (R$MN) (3,371) Terminal Value 13,659 G&P Value 10,289 Source: Espirito Santo Investment Bank for estimates Distribution We value the distribution business at R$21bn, which represents 4% of our EV and an implied EV / EBIT 12E of 12x. We estimate this business unit by deriving it from the downstream unit. We forecast its % of total sales from that unit at 33%.We expect a stable gross margin at 8.3% from 2012 onwards. As the company has lower capex needs and we estimate it to be FCF positive from 2012E, this gives it a higher implied valuation than for the other business areas. We believe that Ultrapar (UGPA3, NR) could be seen as a reasonable peer for this unit. According to Bloomberg consensus, UGPA3 is currently trading at 12.9x EV/EBIT 12E, which supports our valuation. DISTRIBUTION CASH FLOW R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 1,743 1,861 2,058 2,331 2,663 3,017 3,378 3,712 3,969 Taxes on EBIT ,026 1,148 1,262 1,350 NOPLAT 1,150 1,228 1,358 1,539 1,757 1,991 2,229 2,450 2,620 Depreciation & Amortization Capex (1,043) (1,055) (1,067) (1,079) (1,025) (1,094) (1,157) (940) (774) in Working Capital 296 (214) (293) (319) (322) (418) (417) (414) (276) Free Cash Flow ,012 1,115 1,327 1,799 2,287 Discounted Cash Flow Source: Espirito Santo Investment Bank for estimates Page 27 of 61

28 DISTRIBUTION EV BREAKDOWN (R$MN) ,985 Terminal Value 14,857 Distribution Value 20,842 Source: Espirito Santo Investment Bank for estimates International Operations PETR has activities in 27 different countries. We value its international division at R$12bn, which represents an EV / EBIT 12E of 6.8x. We largely base our production figures on PETR s business plan. We assume that total oil production will reach 240k bpd by 2020E from 151k in 2010 (CAGR 11E-20E of 6.4%). We estimate the oil realization price to remain stable in the long run at $64.23 / bbl and gas at $2.99 / MM BTU. We expect costs to drop as oil prices fall. INTERNATIONAL CASH FLOW R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT 1,713 1,559 1,531 1,571 1,718 1,949 2,259 2,659 3,179 Taxes on EBIT ,081 NOPLAT 1,131 1,029 1,011 1,037 1,134 1,286 1,491 1,755 2,098 Depreciation & Amortization 1,687 1,851 2,009 2,156 2,274 2,348 2,410 2,447 2,455 Capex (3,700) (3,742) (3,786) (3,830) (3,198) (3,140) (3,052) (2,531) (2,570) in Working Capital 111 (81) (115) (123) (125) (157) (147) (141) (91) Free Cash Flow (771) (944) (882) (760) ,530 1,893 Discounted Cash Flow (771) (848) (711) (551) Source: Espirito Santo Investment Bank for estimates INTERNATIONAL EV BREAKDOWN (R$MN) (734) Terminal Value 12,299 International Value 11,565 Source: Espirito Santo Investment Bank for estimates Corporate Overheads We estimate the negative contribution from corporate overheads to the company s EV at R$39bn, which implies an EV / EBIT 12E of 5.2x. We break down this calculation into two parts in order to: i) calculate the negative contribution of administrative expenses; and ii) account for the positive contribution of the benefits from interest on capital (IoC). As we have calculated the SoTP through NOPLAT and disregard any tax benefit from IoC, we need to carry out a separate DCF to evaluate this benefit. It is also important to highlight that we are not including estimates related to the company s biofuel division. In PETR s business plan, the company disclosed its intention to invest c.r$7bn in this business over the period. However, as this business has no proven track record and the company has not provided any financial guidance for the unit, we prefer not to attribute any equity value to this new business. However, as we recognize that the related capex should be carried out anyway and we are considering the disbursement together with the corporate overheads. Therefore, capex in the biofuel division has a negative contribution to our EV of R$4.8bn. Page 28 of 61

29 We estimate corporate overheads to maintain their proportion of revenues in line with the recent average. CORPORATE OVERHEADS R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E EBIT (7,545) (8,027) (8,794) (9,813) (11,029) (12,310) (13,634) (14,877) (15,798) Taxes on EBIT (2,565) (2,729) (2,990) (3,337) (3,750) (4,185) (4,636) (5,058) (5,371) NOPLAT (4,980) (5,298) (5,804) (6,477) (7,279) (8,125) (8,999) (9,819) (10,427) Depreciation & Amortization ,039 1,097 1,104 1,114 1,120 1,120 Capex (2,187) (2,211) (2,237) (2,263) (1,169) (1,241) (1,251) (1,114) (1,120) in Working Capital Free Cash Flow (6,432) (6,666) (7,098) (7,701) (7,351) (8,262) (9,135) (9,813) (10,427) Discounted Cash Flow (6,432) (5,988) (5,728) (5,582) (4,787) (4,833) (4,800) (4,632) (4,421) Source: Espirito Santo Investment Bank for estimates CORPORATE OVERHEADS EV BREAKDOWN (R$MN) (47,203) Terminal Value (67,739) Overhead (114,943) Source: Espirito Santo Investment Bank for estimates Interest on Capital (IoC) is a tax efficient method for remunerating shareholders. The amount of IoC paid is deductible from the company s taxable profit. However, there is a legal limit to IoC, beyond which the company is obliged to pay shareholders with dividends (non deductible). This limit is determined by: i) the shareholders equity remunerated by the Brazilian long-term interest rate (TJLP) currently at 6%; ii) Below 50% of profit before taxes. We estimate this benefit through a DCF of 34% (tax rate) of the amount estimated to be paid. FISCAL BENEFIT FROM INTEREST ON CAPITAL R$mn 2012E 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E Interest on Capital 10,436 10,436 10,436 10,436 12,236 14,730 16,901 19,084 21,914 Fiscal 34% tax rate 3,548 3,548 3,548 3,548 4,160 5,008 5,746 6,489 7,451 Fiscal PV 3,548 3,187 2,863 2,572 2,709 2,930 3,019 3,063 3,159 Source: Espirito Santo Investment Bank for estimates FISCAL BENEFIT PV BREAKDOWN (R$MN) ,051 Terminal Value 48,403 Fiscal Benefit from IoC 75,454 Source: Espirito Santo Investment Bank for estimates NET CORPORATE OVERHEADS (R$MN) Overhead (114,943) Fiscal Benefit 75,454 Net Corporate Overhead (39,489) Source: Espirito Santo Investment Bank for estimates Page 29 of 61

30 Section 4: Key Investor Themes Equipment & Local Content: A risk but not insurmountable Petrobras has stated that it intends to add 54 new production systems by 2020 (19 between 2011 and 2015 and 35 between 2016 and 2020). This will require not only substantial investments but also major efforts from suppliers to meet this demand. Moreover, all concession contracts signed in Brazil require that a minimum level of goods and services from Brazil-based providers are used in the exploration, development and production phases (the local content requirement, or LC). In the absence of an effective national industrial policy in the sector, the purpose of the LC is to develop the local industry. Therefore, we see little chance of significant relaxation of this obligation. We acknowledge that the level of investment needed and the LC requirement represent material execution risks; however, we believe that the goals are achievable, due to: i) the company s efforts to strengthen the supply chain and promote investments; ii) the government s commitment to guaranteeing the minimum necessary investment conditions; and iii) a coalition of international and national oil equipment and services providers attracted by the large expenditure planned for the coming years. PETROBRAS NEW RESOURCES Critical Delivery Plan Resources Dec-09 Dec-10 By 2013 By 2015 By 2020 Drilling Rigs (above 2,000m) Supply & Special Vessel Production Platforms SS & FPSO Others (Jacket & TLWP) Source: Petrobras As can be seen in the table above, the company s new plan requires the addition of a significant amount of equipment by 2020 to make its targets viable. However, it is important to note that this plan is relatively mature, with 75% of the projects approved by the end of 2011 (whereas approved projects represented 54.7% of the total in the plan). We believe that this should provide more security for Petrobras and its suppliers as they pursue the necessary investments. We note that, according to Petrobras, all equipment for projects expected to start up by 2013 has already been contracted. Page 30 of 61

31 PETR s Business Plan Approved Projects US$bn Approved up After 2014 Total Source: Petrobras Petrobras is working to enable its suppliers planning and financing We believe Petrobras understands that, in addition to the big global providers, smaller domestic names will play a vital role in making its expansion plans happen. Petrobras has stated that it has been in close contact with its supplier base to help them with their long-term growth planning. In terms of concrete measures, the company has launched several operating initiatives, including: i) a workforce qualification program in partnership with the government through the Brazilian Industry Mobilization Program PROMINP. The course uses Senai, a payroll-funded network of training bases, and a national network of technical colleges. So far, 78k people have received free training to prepare them to work in the sector and the program aims to invest R$554mn in training more than 212k professionals by 2014; ii) management solutions for small & mid-sized companies; iii) identification and cooperation agreements to address technological bottlenecks; iv) simplification of the registration process at PETR s suppliers base (which enabled the addition of 1k new suppliers in 2010). Petrobras has also been attentive to the financing needs of its suppliers. The company has taken the lead in structuring a funding program together with Brazilian banks to make financing lines available more cheaply and with less demanding conditions (i.e., the Progredir program). Progredir, which aims to assist all of PETR s suppliers (whether direct or indirect), is based on granting credit to the supply chain, backed by pending receivables in each of the contracts signed between chain participants. The credit will not involve Petrobras resources, but they will be the anchor of the program, since the receivables generated by Petrobras will provide greater support and assurance for the granting of non-performed credit to its entire supply chain. Some of the largest banks in Brazil (Banco do Brasil, Bradesco, Caixa Econômica Federal, Itaú, HSBC and Santander) are part of the program, competing to offer the lowest rates to suppliers. Between its official launch in June/11 and early Oct/11, the program carried out 115 operations reaching around R$604mn in financing. The average cost reduction to the loan takers is around 20%, but it has exceeded 40% in some cases, according to Petrobras. Most of the operations have been carried out in the states of São Paulo and Rio de Janeiro to fund the working capital of small and medium-sized companies. Where is the money coming from? PETR s expansion plans will require large investments not only from itself but also from the whole chain. Analysis carried out by the Brazilian National Oil Industry Organization (ONIP) and Page 31 of 61

32 Booz & Company estimated that the domestic supply chain will receive investments of US$324bn (and US$76bn of opex) by We believe that PETR, the government and the industry are adopting the necessary measures to meet this financing demand: PETR should be responsible for the lion s share of this amount, as its planned E&P investments should reach US$127.5bn by Progredir will play a vital role in widening credit access. The program benefits suppliers up to the 4 th degree in PETR s supply chain, generating a credit cascade effect. Foreign investments will also likely meet some of the demand. As we detail in Appendix 1 (All roads lead to... the pre-salt), there have recently been several announcements of new investments in Brazil by foreign companies with an eye to meeting PETR s strong demand. Following the strategy adopted by some Asian countries in the past, the Brazilian government is using state-owned entities to develop comprehensive credit mechanisms to support the entire supply chain: Marine Merchant Fund (FMM) has recently authorized additional funds of R$9bn to the industry. These resources are passed to companies in the industry through some stateowned banks, with costs usually below the market level and linked to Brazilian longterm interest rates (currently at 6%), or Libor; State-owned banks have shown appetite for increasing their exposure to the sector above the funds from FMM. Caixa Economica Federal (Caixa), e.g., has stated that it aims to become the main lender to the sector and that funding from FMM does not represent a cap. The Brazilian development bank, BNDES, recently launched a R$4bn new credit line tailor-made for the Oil & Gas supply chain. This new line is built around the concept of anchor companies (revenues above R$90mn), around which credit facilities are made available to sub-contractors and smaller suppliers. Lending costs in some cases are as low as 4.5% pa. BNDES also supports the sector through equity operations; Regional governments also make direct loans to the industry. For example, the state of Rio de Janeiro announced that it will make R$3bn in financing available to smaller supply chain companies through the Invest Rio development agency. Furthermore, alternative funding vehicles have also been developed to help fill any gaps. Over recent quarters, several new private equity funds and Receivables Investment Funds (FIDCs) have been created to invest in the sector. Production Units The company is facilitating the supply of production units through standardized projects. By this strategy, it has managed to not only reduce costs through better economies of scale but also to meaningfully reduce delivery time. We believe that the example of the sister semi-submersible platforms P-51 and P-56 illustrate the success of this strategy. Both platforms were built at the Brasfels shipyard in the state of Rio de Janeiro. While P-51 took 56 months for delivery, P-56 took 39 months (excluding 3 months of a strike). In addition to the significant time reduction, the yard managed to deliver a higher local content level of 73% (above the previous 65%). The case of P-57 is also noteworthy. The platform was delivered 2 months ahead of schedule, reaching a delivery period of 33 months. Even with construction works being carried out in both Singapore and Brazil, the local content level reached 65%. Following the standardization and simplification strategy, the P-57 project will set the standard for upcoming projects (P-58, P-62 and P-63, all of which are scheduled for delivery in 2013), according to Brasil Energia. We recognize that, even after this improvement, Brazilian yards are behind international standards and have a long way to go before they will be anywhere close to the productivity of Asian yards. However, we view the improvement as notable if we take into consideration the 20- year blackout of the Brazilian shipyard industry. After this long period of lack of investments, the Page 32 of 61

33 situation has materially changed and now the sector is a significant destination for national and international investments. Currently, according to the Brazilian Naval Construction Association (Sinaval), there are 43 shipyards in Brazil, with 7 already carrying out some sort of expansion and 11 new facilities currently under construction. Shipyards in Brazil Source: Brazilian Naval Construction Association (Sinaval) We also note that Petrobras generally assumes a minimum 40-month period for the construction of producing units longer than necessary for an international player but a reasonable target, in our view, for the Brazilian industry as it comes up the learning curve. In order to run its order book more efficiently, Petrobras manages its LC on an individual basis as each concession assumes different commitments, enabling it to mix its demand from local and international suppliers depending on each project calendar and available capacity. Page 33 of 61

34 International Charter vs. Petrobras EPC contracts International Charter Contracts Unit Capacity Signed Targeted Contractual Actual Scheduled vs. Main (bpd) Start delivery period Start prod. Start Contractor Marlim Sul 100,000 Jan-03 Apr months Jun months SBM Capixaba 100,000 Nov-04 Apr months May months SBM Cd. Rio de Janeiro 100,000 Jul-05 Apr months Jan-07-3 months Modec Cd. Vitoria 100,000 Aug-05 Jan months Nov months Saipem Cd. Niteroi 100,000 Jan-07 Oct months Feb months Modec Cd. Andra dos Reis 100,000 Aug-08 Nov months Oct-10-1 months Modec P ,000 Feb-08 Jan months Dec-10-2 months SBM* Petrobras-owned contracts Unit Capacity Signed Targeted Contractual Actual Scheduled vs. Main (bpd) Start delivery period Start prod. Start Contractor P ,000 Jun-00 Apr months Aug months KBR P ,000 Jun-02 Jul months Feb months KBR P ,000 Feb-02 Oct months Apr months Jurong / Maua P ,000 Dec-03 May months Nov months Keppel Fels / Technip P ,000 Apr-04 Jul months Dec months Jurong / Maua P ,000 Apr-05 Jul months Nov months Keppel Fels / SBM / Quip P ,000 Apr-04 Apr months Dec months Keppel Fels / Technip P ,000 Oct-07 Mar months Aug months Keppel Fels / Technip Source: Upstream * Hybrid contract ownership shifts to Petrobras Brazilian Petroleum Agency (ANP) & Domestic Content: A Threat ANP administers LC policy, verifying the fulfilment of contractual requirements and penalizing companies when contract terms are not completely fulfilled (fines are proportional to the nonrealized local investment). Over the last few years, ANP has built up significant technical expertise and we expect it to maintain an open-minded and business-friendly regulatory approach. Moreover, the development of the regulatory framework and Petrobras increasing experience of LC should reduce the likelihood of future problems, in our view. After applying no fines in its prior audits (2008/10), ANP has recently led to investors concerns being raised after citing several oil companies for non-compliance with minimum LC requirements from the 5th and 6th rounds. We believe that this seemingly more aggressive approach is a consequence of: A new technical approach: ANP representatives have confirmed to us that over the last few years the agency has developed more efficient ways of measuring the LC level. As reported in Upstream, PETR s CEO Mr. Gabrielli stated that Petrobras had fallen only about 5% short of the target during exploration activities but he described the measure as unfair due to the changes in measurement criteria during the period; Unrealistic commitments from the oil companies during the 5th and 6th rounds. These rounds increased the weight of LC in the final offer (from 15% to 40%) and also included an exponential factor to LC, encouraging competitors to maximize the content offer instead of increasing the fee (due to the exponential effect). As we can see in the chart below, this change in methodology significantly increased the LC offer in these bids. Page 34 of 61

35 Local Content Commitments in Concession Agreements 100% 90% High possibility of penalties 80% Local Content Offerings (%) 70% 60% 50% 40% 30% 20% 10% 0% Onerous Assignment Bid Round Main Pre-salt fields in Santos Basin Source: ANP; * The 8 th Bid Round was canceled Exploration Development After a recent meeting with ANP, we believe that the agency is following very closely the development of LC in Brazil. We believe that ANP is unlikely to relax LC requirements; however, we feel that that it could grant occasional waivers if strictly necessary to make projects viable if the domestic industry is proven unable to meet demand. Moreover, ANP has said that, for the 11 th Bid Round & 1st Pre-salt Bid Round, it is considering offering simplified procedures for measuring and evaluating LC. This change, in our opinion, is a result of a more open dialogue with the industry to make the fulfilment of this requirement more achievable. As an example of ANP s open dialogue, Brasil Energia reported in Sep/2011 that ANP was studying the possibility of creating a system of credits that could be used for operators and vendors to increase the LC in oil exploration and production projects. The measure could be put into practice as soon as the 12 th Bid Round, according to the report. This change would be very important not only to guarantee higher levels of LC but also to transform the country into a specialist for some goods and consolidate local R&D centers. This idea was previously presented by Mr. Francisco Itzaina, president of Rolls-Royce s Brazilian unit, targeting to receive credits when Brazilian production is exported. Regulatory Framework: Stable but Noisy The pre-salt oil production will generate billions in tax revenues for municipal, state and federal governments, which could turn the subject into a highly sensitive political matter. Theoretically, the laws already approved (we discuss the recent regulatory changes in Appendix 2: Regulatory Background ) would be sufficient to create a regulatory framework for the new regime and allow the auction for the pre-salt areas not under concession yet. However, during the proceedings for the approval of the new legislation, Congress changed the government take distribution methodology, benefiting the oil non-producing regions to the disadvantage of producing regions. However, this change was subsequently vetoed by the President. Congress now has to vote on whether or not to drop the presidential veto. Discussions around this issue have been quite intense since the beginning of 2011 and we expect to see a decision by the end of the year. Generally, the debate has focused on the government take sharing between various levels of government and not on a taxation change for the oil companies. Therefore, we do not expect the discussion to have a significant impact on Petrobras. We highlight the following key points from current discussions: Page 35 of 61

36 Representatives from oil-producing states are proposing an increase in the total tax burden in order to increase the tax revenues distributed to non-producing cities and states without decreasing their share. This goes against our view of a solid regulatory framework in Brazil, as it would be seen as a contract breach; therefore, we attribute very low probability to this scenario. We base our view on the fact that the concession contracts, as per the draft version available at the time of the 10 th Bid Round, through its Annex V, specifically define the government special participation rates based on decree #2,705 from Aug/03/1998. In contrast with some other countries (e.g. the UK), we believe that these contracts do not anticipate any changes to accommodate new legislation. If such a discussion were taken to the courts, we believe it would be the worst case scenario for all the parties involved, as it would likely take several quarters to be solved. We also highlight that, during discussions, the Federal Government has shown its commitment to keep contracts in force unchanged. Although we believe that the taxation environment for current concessions looks secure, there is poor visibility on what will be applied to new contracts. We believe that we are likely to see an increase in current royalties and/or special government take (SPT). The President has proposed an increase in the royalties to these areas from the current 10% to 15%, but this could still be further altered by Congress. However, we highlight that we are not taking new producing rights into consideration in our valuation; therefore, we do not see this as a risk to our fair value. As soon as this debate is resolved, the government should start working on the proceedings for the 1st Bid Round for the pre-salt. We do not expect this auction to occur before Production Sharing Agreement (PSA) & Petrobras: Risks vs. Benefits The new regulatory framework (details in Appendix 2: Regulatory Background ) secured a 30% minimum stake for Petrobras in all the remaining pre-salt areas. Therefore, Petrobras will have facilitated access to one of the most important oil provinces in recent oil history, and this should be reflected in the company s production curve over the coming years. However, we are not convinced that this will be as important in terms of value creation. We recognize that many things can change as we are still far away from the 1 st bid round under the new regime. Therefore, due to a lack of visibility, we prefer not to attribute any value from rights under the new regime yet. Points to watch for include: The compulsory 30% stake: Although it is generally seen as a benefit, the minimum 30% stake could also be seen as an obligation. We understand that the new legislation does not offer an alternative to the company, which means that Petrobras will have to take this stake regardless whether or not it is in its best interest. We believe that with such a busy investment plan over the coming years, the company may not be looking for more projects, which could put even more pressure on its capex plan and increase execution risks. We also note that, as per current indications, Petrobras is only a pricetaker. Therefore, it will be obliged to join the winning consortium with a minimum 30% stake, replicating the conditions offered by the winning bid. This raises a material risk as other companies, with very different cost of capital assumptions, could reach very different valuations for the projects; Taxation: While we do not expect the government to change the total tax burden on current concessions, we believe that higher taxes from the new strategic areas would be an alternative to meeting the request of non-producing states for a higher % of oil taxes. Therefore, in addition to having only a certain % of the profit oil, the company is also likely to face a less benign taxation environment. However, we emphasize that such changes should affect only new concessions (or contracts under the PSA regime), which are not yet included in our valuation. At this point, we see more downside risks than upside. We would follow the development of this process very closely. However, we assign a neutral equity value contribution to the new PSA projects due to: Page 36 of 61

37 ANP as a gatekeeper: We expect ANP to remain committed to preventing value erosion at Petrobras, as this could have a negative effect on the entire Brazilian oil industry. Therefore, ANP has stated that such risks will be considered when preparing the bidding conditions of the 1 st Bid Round under the new regime in order to align key assumptions and prevent interference from unqualified companies. We would not be surprised to see occasional waivers to Petrobras 30% mandatory stake if required; Significant capex needs expected only at the end of the decade: Even if the auction takes place in 2012 (which we see as unlikely), capex over the subsequent years will not be material as it would only refer to the minimum exploration program. Material capex needs for the development of new areas should not occur before 2018, when we expect the company to have become a positive cash flow generator and the Brazilian oil equipment industry to be much better prepared. Life Outside Pre-Salt: Upside Risks We recognize that a key pillar of our investment thesis is the pre-salt reserves and our expectation for faster ramp-up and better productivity than currently priced in by the equity market. However, it is important to keep in mind that the pre-salt acreage represents only 2% of the total area of Brazilian sedimentary basins. Additionally, according to ANP director Mrs. Magda Chambriard, as quoted in Upstream online, only 6% of the territory is well studied and 16% is partially studied from an upstream perspective. Therefore, the oil agency will encourage further analysis of the remaining territory, even directly acquiring data in frontier areas where the exploration risk is still unattractive for private capital and investors. Brazilian Sedimentary Basins Acreage ('000 km2) Granted to concession Currently under concession # Main Basins Offshore Onshore Total Area ('000 km2) % Total # blocks Area ('000 km2) % Total 29 2,500 5,000 7, % % Source: ANP Pre-Salt Acreage Area km2 % Pre Salt % Brazilian Acreage Offshore Total W/ PETR's interest 35,750 24% 1.4% 0.5% W/O PETR's interest 6,000 4% 0.2% 0.1% Pre-Salt Under Concession 41,750 28% 1.7% 0.6% Area not Under Concession 107,228 72% 4.3% 1.4% Total Pre-Salt Area 148, % 6.0% 2.0% Source: ANP and Petrobras We believe Petrobras is cognisant of all potential opportunities. However, due to lack of visibility, we prefer not to assign great value to them. Nevertheless, we acknowledge that this approach may prove to be too conservative in the future. We discuss below these relevant upside risks. Page 37 of 61

38 Brazilian Sedimentary Basins Source: ANP Brazil's 11th Bid Round: Interesting trend of geographical diversification Since the implementation of the new regulatory environment, ANP has held 10 series of auctions to grant concession rights for oil exploration and production. The consolidation of this new regulatory framework has been vital for guaranteeing the development of the Brazilian oil industry through a higher level of competition and increased participation of international players. However, as a consequence of the recent regulatory changes, ANP did not hold auctions in either 2009 or 2010, creating a shortage of new exploration opportunities. Thus the 11th oil auction has been much awaited by the sector. In 1H11, the National Energy Policy Council (CNPE) approved the new auction and only the final approval by President Dilma is now pending. Based on the information currently available about the 11th Bid Round, we note: It foresees no changes to the current concession model; No blocks will be offered close to the pre-salt area. The geographical focus will be the Equatorial Margin in an attempt to promote exploration activity in the region and deepen the geological knowledge of the area; Concession of 174 blocks (87 onshore, 87 offshore). After the long blackout on new exploration opportunities, we expect intense competition in this auction. However, we believe that Petrobras will be able to leverage its superior knowledge of the Brazilian sedimentary basins to maintain its dominant position. As there will not be any offer of the main pre-salt blocks, this new auction should not require the payment of high concession fees or pressure short-term capex. Very little is known about the Brazilian Equatorial Margin. There are only 30 exploration blocks under concession and only 5 drilled wells in deep waters. However, expectations are very high. According to the Pangaea theory, this region was once near the African coast (close to Ghana and the Ivory Coast), where there have been a series of oil finds. Therefore, these regions may have a similar geological structure, which could suggest the presence of substantial oil reservoirs in the Brazilian Equatorial Margin as well. Page 38 of 61

39 Equatorial Margin vs. African Coast Source: ANP Also, according to the preliminary multiannual plan , the government intends to carry out two Bid Rounds under the new PSA regime and 7 rounds for mature basins and marginal fields. New Frontiers under Concession Over the last few years, Petrobras has accumulated a rich exploration portfolio even outside the main Campos, Santos and Espirito Santo basins. Due to poor visibility and lack of material results, this potential has not been a focus for investors. We take a similar view and attributed no value to this potential due to limited visibility. However, we emphasise this as a potential upside risk. Just as the Santos basin was not in the spotlight last decade before the pre-salt discovery, other accretive oil provinces could arise from PETR s exploration portfolio, in our view. In its F, PETR disclosed that it had 138 exploration agreements covering 198 blocks, corresponding to a gross exploratory acreage of 130k km2. We detail below the number of exploration blocks per basin according to ANP data; however, we point out a slight discrepancy between ANP and Petrobras figures as we could identify only 192 exploratory blocks. However, we believe that this difference is not important and possibly attributable to a timing difference between the company and the agency. We highlight that the company has 21 exploratory blocks in the mature basin of Campos, which, if successful, would benefit from the infrastructure already in place. As ANP has indicated a commitment to developing further geological knowledge of the Brazilian sedimentary basins, we include in the tables below the current status (this data was made available by the agency in early 2011) of some analysis that is part of its multiannual plan of geological and geophysics studies. When there is not enough appeal to private investors to carry out analysis of unexplored areas, ANP assumes this role. The basic sequence of ANP s works is: 1) Aerial Survey; 2) Geochemistry Data; 3) Seismic Studies; 4) Processing and Reprocessing of Data; and 5) Drilling. Page 39 of 61

40 SEDIMENTARY BASINS: RESERVES & EXPLORATION BLOCKS Total Reserves (3P) Sedimentary Exploratory Acreage (km2) Onshore Offshore Total Aerial Survey Geochemistry Data Seismic Studies Basin Blocks (#s) Onshore Offshore Oil (MM bbl) Gas (MM m3) Oil (MM bbl) Gas (MM m3) Oil (MM bbl) Gas (MM m3) Concluded Ongoing Concluded Ongoing Concluded Ongoing Almada Amazonas x x Barreirinhas Camamu Campos Ceará x Espírito Santo Foz do Amazonas Jequitinhonha Pará - Maranhão Parana** x x Parecis - Alto Xingu x x Parnaíba x x x Pelotas x Pernambuco - Paraíba Potiguar Recôncavo* x Rio do Peixe x Santos São Francisco x x Sergipe / Alagoas Solimões x Tucano Total Source: Companhia de Pesquisa de Recursos Minerais (CPRM), ANP * According to ANP, there is one offshore field in the Reconcavo basin (Dom Joao Mar, extension of the onshore Dom Joao). There is no opening of its acreage. ** It refers to the acreage in the Brazilian territory ANP STUDIES IN SEDIMENTARY BASINS WITH NO CONCESSIONS Sedimentary Aerial Survey Geochemistry Data Seismic Studies Basin Concluded Ongoing Concluded Ongoing Concluded Ongoing Acre x x x São Luis x x x Araripe x Source: Companhia de Pesquisa de Recursos Minerais (CPRM ), ANP We have identified some sort of exploration activity (even if very limited) in basins that sum around 5.2mn km2 (thus it has absolutely no ongoing explorations in basins that represent around 2.3mn km2). Therefore, we share the view of ANP director, Mrs. Chambriard, that the geological knowledge of the Brazilian sedimentary basins is still very poor, reducing visibility of PETR s exploration potential. We believe that if ANP works were at more advanced stages, they would likely raise investor confidence. However, considering the currently low level of exploration activity (340k km2 vs. total sedimentary area of 7,500k km2), we believe there is room for intense newsflow on this front over the coming years. Page 40 of 61

41 PROJECTS CONCLUDED PROJECTS ONGOING Source: ANP Source: ANP PROJECTS PROPOSED Source: ANP As an example of this potential, we note the recent oil discovery in ultra deep waters in block SEAL-M-426 in the Segipe-Alagoas basin. Although PETR has not released volume estimates, it has stated that the information obtained is sufficient to confirm the discovery of a new oil province in the Sergipe-Alagoas basin. The sample indicated excellent oil quality, with API as high as 43 (far better than the current Brazilian average of c.23 API). Petrobras has 15 exploration blocks in the area. Page 41 of 61