IG4 / Shine I - independent negotiation position

Public-information cutoff: 18 August 2026. This is Shine I FIP’s independent Round 1 position for Braskem S.A.

Labels follow house style. Fact means a verified public disclosure. Reported means a press account that is not a company filing. Inference means a negotiation judgment that counterparties may contest.

Shine I owns 226.335 million common shares and 47.294 million PNA shares, equal to 50.1108 percent of voting capital and 34.3234 percent of total capital (fact). Petrobras holds 47.03 percent voting and 36.15 percent total. The parties jointly control Braskem through consensus governance (fact). Shine nominates important management and transformation roles. Petrobras holds major operational and governance nominations (fact). The Shine stake was acquired through a distressed-credit and share-exchange structure involving NSP and Novonor claims, not a conventional full-cash bid at historical equity value (fact).

IG4 does not control creditor votes. IG4’s disclosed equity-fund AUM is around US$1 billion. That figure is not uncalled capital sitting ready for Braskem, and it will not be treated in this room as a US$1 billion cheque (fact / inference). The US$476 million Braskem Idesa contribution is Braskem S.A. and Braskem Netherlands cash. It is not a new IG4 fund cheque (fact).

1. Objective and realistic alternative

Shine’s objective is to preserve turnaround value and Petrobras/Shine joint control of a solvent parent while the petrochemical cycle recovers. A smaller percentage of a going-concern Braskem can be worth more than 34.3 percent of an insolvent one. The Shine thesis still fails if creditors take 70 to 90 percent of equity and management (inference).

Persistent PE-naphtha spreads argue for contingent warrants rather than immediate equitization. They do not argue for leaving the US$10.3 billion corporate stack untouched without sponsor cash at the parent.

The realistic alternative if this round fails is not a quiet return to the status quo. It is a generic plan-to-a-plan extrajudicial reorganization filed before the reported 24 August stay expiry, buying about 90 days of mediation with cash controls. If the stay expires with no EJ, no committed Petrobras cap, and no IG4 parent facility, the path is RJ after acceleration risk. It is not a standalone cure of the disclosed US$98 million of defaults (inference).

2. Mexico: separate estate, capped contribution, majority is the asset

On 18 August Braskem confirmed a consensual Braskem Idesa prepack. Idesa senior debt is to fall from about US$2.5 billion to about US$1.6 billion. Prepackaged Chapter 11 proceedings were commenced in the Southern District of Texas, with completion expected in 60 to 90 days (fact).

Braskem, as controlling shareholder, will contribute US$476 million, of which about US$126 million had already been made available. Upon completion Braskem will continue to hold a majority stake, with Grupo Idesa and affiliates as the largest minority. Operations continue without interruption. Unsecured creditors and trade vendors are to be paid in the ordinary course. Braskem described Idesa as a strategic asset (fact).

About US$350 million of the US$476 million remains to be funded. Treat that remainder as a live use of Braskem S.A. and Braskem Netherlands liquidity unless later filings show it is only a restatement of the existing term loan or the US$82 million working-capital facility (inference required by the public record). The term-loan lender is listed as a separate supporting party. The US$126 million already funded is not assumed to be that term loan (inference).

Q2 Mexico EBITDA of US$57 million at 43 percent utilization is not a run-rate. The remaining US$1.6 billion of Idesa senior debt is underwritten on operating the plant, not on annualizing that print (fact / inference). Cutting Idesa senior debt does not, by itself, reduce the US$10.3 billion corporate stack, the 6.74x corporate leverage, or the December standby. Consolidated IFRS gross debt may fall if the haircut is a real principal reduction. That is a consolidation-optics point, not a parent recapitalization (fact / inference).

Creditors will argue that IG4 already spent parent cash on Mexico and must therefore put more junior capital into the parent, approaching a dollar-for-dollar recapitalization of the US$10.3 billion stack. Shine’s reply is specific. Keeping majority of a delevered, in-place Mexican cracker is the turnaround thesis, not a leak to be abandoned.

The US$476 million is company cash preserving that asset for the same creditors who would recover less in a parent RJ that also orphaned Mexico. Further Mexico leakage after the disclosed amount is off the table. Parent warrants and IG4 junior capital remain available at the Brazilian parent. Shine will not recapitalize the corporate stack dollar-for-dollar because of Idesa (inference).

The economic case for majority is not sentimental. Idesa Q2 EBITDA of US$57 million at 43 percent utilization implies substantial operating leverage if utilization recovers toward historical levels, which is why a delevered plant with a US$900 million senior-debt cut is an asset rather than a leak (inference). Abandoning majority to save the remaining US$350 million would crystallize a strategic loss in exchange for a few months of parent cash, and it would not retire the US$10.3 billion corporate stack. Shine will not trade that asset to appease a recapitalization demand that Mexico does not legally or economically support.

Required Mexico treatment in every package Shine will sign:

1. Idesa remains a separate estate. There is no disclosed parent guarantee of the remaining about US$1.6 billion of Idesa senior debt, and no Braskem S.A. Chapter 11 or RJ filing (fact).
2. No further parent cash, guarantee, keep-well, or cross-collateralization is permitted after the disclosed US$476 million (inference / authority limit).
3. The existing US$82 million working-capital loans due December 2026 remain secured by Idesa assets. The term loan of US$180 million committed and US$129 million disbursed remains an Idesa-estate instrument. Residual parent exposure after closing is equity plus amounts already disclosed, not a new parent assumption (fact / inference).
4. Consolidated Idesa debt reduction is not credited as parent deleveraging of the US$10.3 billion stack (inference).
5. Majority ownership is an asset worth preserving. It is not a cash leak to be stopped by handing the parent to creditors or by selling control of the plant for a short-term cash saving (inference).

3. Q3 operating cases, cash conversion, and remaining Idesa use

Corporate gross debt was US$10.3 billion and adjusted net debt was US$9.5 billion at 30 June. Adjusted corporate net leverage was 6.74x. These are Braskem S.A. corporate figures and already exclude Idesa project debt (fact). Q2 recurring EBITDA was US$1.043 billion, including US$869 million from Brazil and South America, US$147 million from the United States and Europe, and US$57 million from Mexico. H1 recurring EBITDA was about US$1.235 billion (fact).

Q2 working capital consumed US$547 million. Braskem attributed this to higher feedstock and product prices, higher inventory volumes, and reduced payment arrangements. The price and volume components are not a quarterly run rate if they stabilize. The financing component is persistent (fact / inference).

Cash and cash equivalents were R$3.931 billion at 30 June, with R$353 million inside Braskem Idesa. The June restructuring plan showed about US$795 million of unrestricted cash (fact). The US$1.0 billion standby facility matures on 31 December 2026. The June plan showed US$2.349 billion of H2 contractual debt service, including LC runoff and the standby maturity (fact).

In every case below, remaining Idesa cash is a use, not a source. The December standby cannot be paid from residual cash. It must be extended, refinanced, or included in the restructuring.

June-plan case, Q3 EBITDA US$586 million. The June plan projected Q3 contractual debt service of US$878 million, including US$572 million of LC runoff, and about US$337 million of unrestricted cash at 30 September under the status quo (fact). That September figure predates the 18 August Idesa announcement. Subtracting a remaining Idesa use of US$350 million leaves September cash at approximately nil to modestly negative if the cheque is Q3-weighted.

If closing stretches into the 60-to-90-day prepack window, some of the US$350 million may fall in October or November, which still makes it a first-order H2 claim six days before the reported stay expiry (inference). EBITDA does not convert one-for-one while working-capital financing stays tight. This case cannot carry LC runoff, Idesa closing, and a December standby without Petrobras trade credit and a creditor LC roll. Cure-and-continue is not a plan. It is a few weeks of optionality.

Sensitivity case, Q3 EBITDA US$750 million. Relative to the June plan this adds US$164 million of EBITDA. Estimated September unrestricted cash after scheduled Q3 requirements is about US$250 million to US$500 million before any new Petrobras facility (fact of the brief’s construction). Treating remaining Idesa US$350 million as an additional parent use compresses that range to about −US$100 million to +US$150 million, depending on working-capital conversion and how much of the US$350 million actually funds before 30 September (inference).

Better earnings do not pay the standby. Petrobras trade credit is still required. The ask can sit at the US$250 million to US$500 million test points rather than the mathematical US$950 million ceiling.

Upside case, Q3 EBITDA US$1.0 billion. Persistent PE-naphtha spreads would add about US$414 million of EBITDA versus the June-plan print. Mapping that increment onto the US$750 million cash band implies roughly US$500 million to US$750 million of September unrestricted cash before Idesa, and about US$150 million to US$400 million after a US$350 million remaining contribution (inference). Cash conversion still has to be proved. Q2’s US$547 million working-capital use is the warning.

Even this case cannot retire the US$1.0 billion standby from cash. It does strengthen contingent warrants over immediate equitization. It can support a smaller Petrobras cap if LCs roll and the standby is extended. It does not let Shine skip a numbered junior facility at the parent.

Across all three cases the H2 problem is the same shape. Remaining Idesa cash, LC runoff, and the December standby are uses. Recurring EBITDA is a source only to the extent it converts after working capital. Petrobras trade credit changes timing. It does not absorb losses. IG4 junior capital and creditor warrants are the loss-absorbing and upside-sharing tools at the parent. Mexico’s US$476 million does not substitute for those tools.

4. Petrobras trade credit after Idesa

Valor Economico reported on 17 August, citing unnamed sources, that Petrobras put commercial support on the table. The concept could extend terms on naphtha purchases currently paid in cash for as long as six months. It would not be a loan or new financial debt. Format, duration, pricing, and the exposure cap remained under discussion. Petrobras, IG4, and Braskem did not comment. No 18 August filing confirmed that the concept is signed, capped, or accepted by creditors (reported).

Valor also reported an objective of one-third creditor support and an EJ before 24 August, possibly a generic plan-to-a-plan EJ providing another 90 days, with positions still far apart (reported).

Braskem purchased R$5.098 billion of raw materials, finished goods, services, and utilities from Petrobras and subsidiaries in H1 2026. Annualized flow is about R$10.2 billion, or about US$2.0 billion. The June parent-company payable to Petrobras was R$257 million, roughly nine days of the H1 purchase run rate (fact). The mathematical ceiling from moving approximately nine-day terms to 180 days is roughly US$950 million of incremental trade liquidity. The report says exposure will be capped, so do not assume the ceiling. Test US$250 million, US$500 million, and US$950 million (reported / inference).

Petrobras trade credit is still required in every Q3 case after the remaining Idesa cheque. It is liquidity. It is not loss-absorbing capital. If commercially priced, its subsidy value is small even when its cash-timing value is large (inference).

The remaining US$350 million parent cheque to Mexico makes Petrobras less willing to extend uncapped or equity-like support. Election-year optics and Law 13,303 scrutiny punish a story in which public-company money follows a large parent transfer to Mexico without matching IG4 junior capital into the Brazilian parent (inference). The same cheque strengthens the creditor demand for shareholder burden-sharing. Shine accepts that demand at the parent. Shine does not accept it by re-characterizing Idesa as IG4’s contribution. A Petrobras naphtha facility, at any of the three test caps, does not by itself satisfy burden-sharing.

Opening Petrobras ask: a capped, commercially documented naphtha or working-capital facility of US$400 million, about six months, arm’s-length pricing, no parent guarantee of Idesa, and no open-ended keep-well. Shine will test US$250 million if Q3 prints toward US$1.0 billion and LCs roll, and US$500 million if the June-plan case materializes. Shine will not treat US$950 million as a planning number. Duration should bridge through year-end standby negotiations, then reset on disclosed terms. Independent Petrobras approvals and a defined exposure cap are features Shine wants, not obstacles, because they make the facility politically durable.

5. Cure-and-continue after the remaining Idesa cheque

The Brazilian court granted 60 days of protection on 26 June to support mediation. The stay is reported to expire on 24 August 2026. Chapter 15 provisional protection was also obtained in New York. The stay covers invited financial creditors in the Wind Chamber mediation. It does not stay trade, suppliers, or customers (fact / reported).

Braskem suspended payments covered by that injunction. After cure periods expired in July, it disclosed R$507 million, or US$98 million, of defaults under certain financial instruments. It did not publicly disclose acceleration as of the cutoff (fact). Public bond terms generally provide a 30-day interest cure period. Holders of at least 25 percent of an affected series can generally accelerate after an uncured event of default. A majority can generally rescind acceleration before judgment if overdue amounts, expenses, and other defaults are cured. Cross-default terms vary by instrument (fact).

Paying US$98 million does not fund remaining Idesa cash, US$572 million of Q3 LC runoff, or a US$1.0 billion December standby. After the Idesa announcement, cure-and-continue is not credible as a path to 31 December (inference). It is only a tactical option to reduce near-term acceleration risk while an EJ is filed.

Using scarce parent cash to cure coupons while sending US$350 million to Mexico, without a signed parent package, is the picture Shine must not present. If 25 percent of a major series accelerates after 24 August, Shine’s fallback is a defensive EJ or RJ that preserves joint-control optionality, not a scramble to write a larger fund cheque under fire (inference).

6. Opening package, first concession, narrowest acceptable

Opening. Five-year maturity extension on the corporate stack, with no principal haircut. PIK through December 2028, then cash coupon at existing contractual rates. IG4 provides a US$200 million junior shareholder or hybrid facility into the parent. That facility is new Shine risk capital and is not the Idesa contribution.

Petrobras provides at least US$400 million of capped, commercially priced trade or working-capital support. Creditors roll LCs and provide about US$200 million of incremental LC or working-capital capacity. Creditors extend or refinance the December standby into the five-year package. Creditor warrants start at 10 percent, plus 5 percent if cash-conversion, EBITDA, liquidity, or PIK milestones are missed.

No blanket liens over core Brazilian assets. Mexico as specified in section 2. Alagoas safety and remediation cash is ring-fenced, with reporting preserved and no dividends while leverage is high. File a plan-to-a-plan EJ before 24 August if a full term sheet is not signed.

First concession. IG4 increases to US$250 million. Warrants move to 15 to 20 percent. Limited non-core or working-capital collateral is acceptable. The Petrobras cap can be set at US$500 million if independently approved and commercially documented. Shine will accept cash controls, information rights, and a temporary EJ with those protections. The PIK window can be shortened if the US$1.0 billion EBITDA case is printing and converting to cash.

Narrowest acceptable, which is the authority limit for this room. IG4 junior capital of US$250 million to US$350 million, structured as a junior convertible or a rights-offering backstop, and only with comparable Petrobras support of at least US$300 million of commercially priced working-capital or trade credit. Shine will not put more than about US$350 million of fund cash into Braskem without that Petrobras match.

Creditor warrants of 15 percent, stepping to 22.5 percent if milestones fail. A pro-rata rights offering is acceptable if BRKM3, BRKM5, BRKM6, and BAK holders receive economically fair participation where legally practical. PNA economic preferences are preserved. Non-subscribers are diluted, which is different from diluting holders who fund pro rata. BAK ADS holders need a depositary or registration mechanism; if direct subscription is impossible, rights should be sold for their benefit where feasible.

Target: old shareholders retain about 70 to 85 percent fully diluted. Shine can accept 20 to 30 percent dilution if enterprise value is repaired and joint control remains. Economic control floor: Shine at about 35 percent of voting capital, and the Petrobras/Shine group above 60 percent. No automatic DIP conversion after a technical default. No blanket liens on core assets.

Creditor equity above 35 percent is acceptable only against roughly US$2.5 billion to US$3 billion or more of principal cancellation. Controllers below 50 percent without that scale of cancellation is outside authority. No further parent cash to Mexico after US$476 million. Majority of Idesa is retained. The US$10.3 billion corporate stack is not recapitalized dollar-for-dollar because Idesa was funded. Unequal treatment or trough-price issuance solely to controllers is rejected.

Red lines, stated as complete limits rather than slogans. IG4 and Shine cash above roughly US$350 million without comparable Petrobras support is outside authority. Controllers below 50 percent without several billion dollars of debt cancellation is outside authority. Shine below roughly 35 percent voting, or loss of agreed governance, is outside authority. Automatic DIP conversion after a technical default is rejected. Blanket liens on all core assets are rejected. Creditor equity above 35 percent without substantial principal cancellation is rejected. Further parent cash, guarantees, or cross-collateral to Idesa after US$476 million is rejected.

Preferred security sequence. First, a sponsor and creditor working-capital bridge, including the Petrobras cap. Second, a signed maturity and PIK agreement that folds in the standby. Third, the junior bridge converts into a pro-rata rights offering only if permanent capital is required. Fourth, creditor warrants provide contingent upside rather than an immediate takeover.

That sequence exists because Shine’s strongest argument against a 70 to 90 percent creditor takeover is repaired enterprise value plus visible junior capital, not a claim that IG4’s option should be protected for its own sake. Creditors can fairly say that a distressed-credit acquisition is not a reason for below-market creditor returns. Shine answers with cash at the parent, warrants, and a capped Mexico estate, not with rhetoric.

7. Instrument-by-instrument treatment

Maturity: five-year extension of the corporate stack. The 31 December 2026 standby is included in that extension. It is not left as a cash pay-at-maturity in any of the three Q3 cases. Interest: PIK through December 2028 at opening, cash thereafter. The first concession can shorten PIK if cash conversion is proven. There is no opening principal haircut.

LC and trade facilities: creditors roll existing LCs and add about US$200 million of capacity. The June Q3 plan’s US$572 million LC runoff cannot be ignored. Petrobras: capped trade credit as in section 4. It is not loss-absorbing. It is not an Idesa backstop. IG4: US$200 million opening, US$250 million first concession, US$350 million authority cap, parent-level junior capital only. Large contributions above that band would require co-investors, creditor-bank investors, or newly raised fund capital, which this room should not assume exists today.

Creditor new money: LC capacity and, if needed, a modest working-capital participation pari passu with rolled facilities. It is not a control-seeking DIP. Warrants and equity: contingent creditor upside, not an immediate takeover. A rights offering is only a permanent-capital backstop. Collateral: limited non-core or working-capital security at the first concession. No blanket Brazilian-asset lien at any authority level Shine now holds.

Covenants: cash controls, information rights, no dividends, an Alagoas ring-fence, and a hard stop on further Mexico parent leakage. Mexico: separate estate; US$476 million is the parent cash cap; majority preserved; the Idesa term loan and the US$82 million working-capital loan stay on Idesa assets; the consolidated Idesa haircut is not parent deleveraging. Alagoas: budgeted safety and remediation protected in every consensual case. No upstreaming of value before those obligations. Any attempt to pay sponsors or bondholders ahead of remediation would destroy political support for joint control.

Q2 filings continued to describe creditor proposals as indicative and nonbinding, including possible capitalization and collateral. No signed term sheet, sufficient adhesion, or committed sponsor funding was publicly disclosed for Braskem S.A. by the cutoff (fact). The June creditor framework rejected Braskem’s initial five-year extension and coupon reduction. Creditors required positive-NPV treatment, shareholder burden-sharing, cash controls, information rights, and direct Petrobras participation. They expressed conditional support for a temporary plan-to-a-plan EJ with protections (fact). Shine’s package is built to meet those headings with parent junior capital and warrants, not with a second Mexico cheque and not with a Petrobras-only rescue.

8. Political overlay

Brazil’s 2026 general-election first round is 4 October, with a possible second round on 25 October. Party conventions ran from 20 July through 5 August, so this negotiation sits inside the campaign (fact). President Lula’s Workers’ Party confirmed his reelection bid. In May, Lula said Petrobras must consider Brazil’s priorities, while also saying the government discusses priorities but does not command the company (fact).

Petrobras is state-controlled, but Law 13,303 subjects it to governance, transparency, risk-control, and related-party standards. Nonmarket public-policy obligations require defined conditions and compensation. A political preference is not enough to justify an uneconomic Braskem rescue (fact).

Federal industrial policy is supportive of the chemical chain through REIQ and Law 15,294/2025 creating PRESIQ, and through provisional polyethylene antidumping measures. Treat those as going-concern supports and scenario variables, not as a permanent protection of today’s ownership percentages (fact / inference). Alagoas remains live. In June 2026, Braskem and former executives became defendants in a federal proceeding related to the Maceió mining disaster, and Braskem continues to disclose administrative proceedings and socio-environmental obligations (fact).

Shine’s useful role is the private turnaround partner that lets the government avoid both a Petrobras-only nationalization story and a disruptive RJ (inference). That cover exists only if IG4 contributes real junior capital at the parent and accepts warrants. Public-company money protecting a distressed sponsor option is the worst political outcome for Shine. The Idesa parent cheque makes that outcome easier for counterparties to allege unless this round puts a numbered IG4 facility on the parent (inference).

A foreign-led creditor takeover would face labor, political, regulatory, and reputational friction. That raises the execution cost of a 70 to 90 percent creditor-equity proposal. It is not a legal veto and should not be priced as a government backstop (inference). The politically easiest path remains a commercially documented, capped, and shared solution: Petrobras working-capital support tied to profitable operations, creditor maturity relief, IG4 burden-sharing at the parent, and contingent creditor upside. An explicit Treasury bailout, an open-ended Petrobras guarantee, or a Petrobras-only equity rescue is less likely and is not something Shine should request (inference).

9. Outcome probabilities

These are negotiation judgments, not filings, and they sum to 100 percent.

• Consensual or plan-to-a-plan EJ with capped Petrobras trade credit, IG4 parent junior capital, LC roll, standby extension, and creditor warrants, after the Idesa prepack: 40 percent.
• Creditor-favorable consensual restructuring with tighter collateral, warrants at the 22.5 percent step-up or a limited rights offering, still with joint control: 28 percent.
• Cure-and-continue without a signed package, paying disclosed overdue interest and hoping to reach December: 7 percent.
• RJ, or equivalent hard enforcement, if the stay expires, acceleration begins, or Petrobras and IG4 cannot commit parent liquidity: 25 percent.

The Idesa prepack cuts both ways. It shows that Braskem can close a consensual deal with noteholders, which supports the 40 percent EJ path. It also consumes parent cash immediately before stay expiry, which supports the 25 percent RJ tail and is why cure-and-continue is only 7 percent (inference).

10. Single fact that would most change this position

Public confirmation, in a 6-K or the Idesa disclosure statement, that the remaining approximately US$350 million is not incremental unrestricted parent or Netherlands cash but only a restatement or roll of the existing term loan or the US$82 million Idesa-secured working-capital facility. That one fact would restore the June-plan September cash band, weaken the creditor demand that Shine pay again for Mexico, and let IG4 hold the opening US$200 million cheque.

The opposite confirmation, a dated funding schedule that drains September unrestricted cash below US$100 million before any Petrobras facility, would force Shine to the US$250 million to US$350 million authority band immediately and make a 24 August EJ non-optional.