IG4 Capital / Shine I - Round 1 position

Chair: Shine I FIP (IG4 Capital), joint controller of Braskem S.A.
Cutoff: 18 August 2026, evening (~20:50 CT). Public information only.
Labels used throughout: Verified = company filing, statute, or disclosed ownership. Reported = press (Pipeline, O Globo, Valor/Bloomberg, Estadão, Fitch commentary) that is not a 6-K. Inference = this chair’s negotiating judgment.

This is a control-and-call memo, not a recap memo. Pipeline Valor (18 August, 12:16 BRT) reported no willingness from IG4 to inject capital. That report is the working-case constraint for this round. A US$300 million Shine junior facility is not on the table as a clearing chip.

1. Objective and realistic alternative if talks fail

Objective. Preserve the product Shine bought: a call on Braskem’s cycle recovery plus joint control with Petrobras (Shine 50.1108% voting / 34.3234% total capital; Petrobras 47.03% voting / 36.15% total - verified, Braskem ownership structure and 20 April 2026 6-K). File a thin, plan-to-a-plan extrajudicial recovery (EJ) of invited financial creditors before the 60-day tutela expires on Monday 24 August 2026 (verified grant date 26 June; reported expiry). Keep Mexico as a separate estate that still delivers majority ownership at Braskem Idesa. Do not recapitalize the US$10.3 billion parent stack with fund cash.

Walkaway if talks fail: Brazilian RJ of Braskem S.A. That is the realistic alternative, not a U.S. Chapter 11 of the parent, not a converting DIP, and not a going-concern sale of the crackers. IG4 will not sign a deal that wipes Shine. IG4 will risk RJ.

Those are different statements:

IG4 is not bluffing a preference for RJ over a thin EJ. The thin EJ is strictly better: it protects the Idesa close, LCs, trade, and utilization. RJ is better than conversion. That ranking is the whole chair.

2. What Shine bought - and what it will not give away

Verified. Shine holds 226.335 million common shares and 47.294 million PNA shares. The stake came through a distressed-credit / share-exchange involving NSP/Novonor claims, not a cash recap of Braskem S.A. Joint control is contractual (shareholder agreement) and numerical (Shine + Petrobras). IG4 disclosed AUM around US$1 billion; that figure is not uncalled Braskem dry powder.

The July creditor menu asked Shine to swap that package for either (i) a converting instrument that leaves controllers below 10% or (ii) a five-year extension at existing coupon with all assets as collateral. Braskem rejected both (reported). This chair repeats the rejection:

• Conversion that drops controllers below 10% is rejected, including if it is dressed as a DIP, a rights offering that only creditors take up, or an EJ cram. A “smaller percentage of a solvent Braskem” is a warrant discussion, not a sub-10% control discussion.
• All-asset liens on the Brazilian crackers are rejected. Limited non-core / working-capital collateral is a concession, not an opening.
• A Petrobras put, keep-well, contingent equity, or parent guarantee is not Shine’s to give. Creditors who want that instrument must obtain it from Petrobras. Pipeline reported Petrobras has no willingness to inject or nationalize. Lei 13.303, TCU scrutiny, the 4 October 2026 election (verified, TSE calendar), and IFRS 10 consolidation of the US$10.3 billion stack if voting control is lost or common is issued are Petrobras hard stops, not IG4 talking points.

Warrants can be acceptable. A put cannot be delivered from this chair.

3. IG4 cash: working case is zero. Last-ditch is not the working case.

Reported (Pipeline, 18 August). Tuesday bondholder meeting: no willingness from Petrobras or IG4 to inject capital. Treat that as the public constraint this rerun exists to honor.

Verified, and not a substitute fund cheque. Braskem will contribute US$476 million to the Idesa prepack, of which about US$126 million was already made available and about US$350 million remains (18 August 6-K). That is company / Braskem Netherlands cash, not an IG4 fund cheque. Majority at Idesa is an option the company paid to keep. This chair will not let creditors re-characterize that US$476 million as “sponsor recap of the parent.” It does not delever the US$10.3 billion corporate stack. It does, however, exhaust the honest “sponsors already put up cash” narrative for this round.

Working case for Shine fund cash: US$0. Reasons, stacked:

1. Pipeline is the 18 August public record. Inventing a US$200–300 million junior facility as the base case requires explaining why Pipeline is wrong. This chair will not.
2. Hide behind Petrobras on any recap. Shine will not be the sole cash sponsor of US$10.3 billion. If Petrobras will not inject - and Pipeline says it will not - an IG4-only cheque makes Shine the only visible equity check in an election year, while Petrobras remains the political residual. That is the worst optics for both controllers (inference, political overlay). Petrobras also has no reason to allow IG4 to be the only cash sponsor (common-brief constraint).
3. AUM ~US$1 billion is not a Braskem commitment line. Large contributions would require co-investors or a new raise. None is disclosed.
4. The parent does not have an equity hole that Shine can fill at the scale creditors want. Fitch’s menu if there is no new money is asset sales, shareholder injection, or liability renegotiation (reported). Injection at parent scale is a Petrobras-and-Shine problem. Shine will not solve it alone.

Last-ditch number - not the working case, not Round 1, not a gift. If, and only if, on 23–24 August the documented alternative is a converting instrument that drops controllers below 10%, Shine would consider a US$125–150 million junior / hybrid backstop, and only on these conditions:

• Petrobras has signed, capped naphtha or equivalent working-capital support of at least US$400 million. Shine matches politically; it does not lead. Never sole sponsor.
• Structure is junior, deeply subordinated, and preferably a backstop to a pro-rata rights offering - not a cheque written to the AHG.
• Warrants, not conversion. Joint voting control preserved.
• Political justification (inference): Law 13.303 related-party optics and the election-year “public-company money protecting a distressed sponsor option” narrative. A small, matching, junior amount exists only to neutralize that narrative so Petrobras can sign WC without being accused of bailing out IG4 for free. It is not loss-absorbing capital for US$10.3 billion (US$150 million is ~1.5% of gross debt). It is control-preservation insurance.

If those conditions are absent, the number is zero and the walkaway is RJ. Do not carry US$125–150 million into a term sheet as if it were Round 1 economics.

4. Required operating cases

No Q3 result is assumed. June-plan figures below are verified as plan, not as outcomes. Arithmetic labeled inference.

Plan identity (status quo, June). Unrestricted cash ~US$795 million at 30 June → Q3 EBITDA US$586 million → Q3 contractual debt service US$878 million (of which US$572 million LC runoff) → ~US$337 million unrestricted at 30 September. Implied other Q3 cash use: 795 + 586 − 878 − 337 = US$166 million (WC, capex, Alagoas, FX, any Idesa already in the plan). Q2 working-capital consumption was US$547 million (verified); price/volume components are level resets if prices and inventory stabilize; lost supplier finance is persistent. Do not assume incremental EBITDA converts 1:1. Use 60–80% cash conversion on EBITDA above the June plan (inference).

H2 contractual debt service in the June plan was US$2.349 billion including LCs and the standby. Residual Q4 service after the US$878 million Q3 line is ~US$1.471 billion, of which the US$1.0 billion RCF / standby due 31 December 2026 is the wall (verified maturity; Elliott reported as largest lender). After July cure periods, Braskem disclosed R$507 million / ~US$98 million of defaults under certain financial instruments (verified). No public acceleration as of this cutoff. Public bond terms generally: 30-day interest cure; 25% of a series can accelerate after an uncured event of default (verified as market-standard on these indentures / reported as applied). Fitch parent rating RD as of 17 August; a subsequent EJ or RJ would take ratings to D (reported Fitch commentary, 18 August).

Remaining Idesa overlay. Treat US$350 million as live parent / Netherlands cash use unless later filings prove it is only a restatement of the existing term loan (US$180 million committed / US$129 million disbursed) or the US$82 million Idesa-secured WC loans. First-day reporting (BankruptcyData, 18 August - court-document reporting, not a 6-K): Braskem-backed DIP up to US$279 million new money plus ~US$130 million roll-up, 10% PIK, converting into Braskem’s equity allocation; plus US$71 million effective-date cash. 279 + 71 = 350. Interim DIP target within two business days of the Idesa filing; confirmation / final DIP within 40 days; effective date within 55 days. Inference: the remaining cheque can leave the parent in late August, inside the tutela window, not after a leisurely Q4 close.

Case A - Q3 EBITDA US$586 million (June plan)

Chair view. Case A is an LC-roll + standstill case, not a cure-and-continue case. Spending US$98 million to look current, then writing US$350 million to Mexico, then facing US$1.0 billion in December, is how a controller donates the call to the RCF.

Case B - Q3 EBITDA US$750 million (sensitivity)

Incremental EBITDA vs plan: US$164 million. At 70% conversion: ~US$115 million extra cash. Sep unrestricted ~US$452 million before incremental Idesa; ~US$102 million after a full US$350 million Idesa draw. LC roll still required to have a December discussion. Standby still cannot be paid. Slightly more room to accept a US$250 million naphtha cap rather than the ceiling.

Case C - Q3 EBITDA US$1.0 billion (upside sensitivity)

Incremental EBITDA vs plan: US$414 million. At 70% conversion: ~US$290 million extra cash. Sep unrestricted ~US$627 million before Idesa; ~US$277 million after. Optics for a thin EJ improve. The December US$1.0 billion RCF is still a restructuring event, not a refinance at RD. Do not annualize Q2’s US$1.043 billion or Mexico’s US$57 million (verified instruction: Mexico utilization 43%; do not annualize).

Standby treatment, all cases. Paid: no. Extended: only inside an EJ/RJ or a signed AHG standstill. Restructured: the live path. Elliott as reported largest RCF lender is why a notes-out EJ does not neutralize Elliott after 24 August - the RCF still matures in four months.

5. Mexico - required treatment

Idesa was already an unrestricted subsidiary on older parent indentures. The prepack is not a J.Crew drop-down of Camaçari.

Pipeline’s Idesa paragraph in the same 18 August piece is not this chair’s analysis of parent exposure.

6. Unsigned naphtha terms after Idesa - do they satisfy burden-sharing?

Reported. Petrobras is weighing capped naphtha payment-term relief (purchases now often cash / short-dated, conceptually toward six months). Not a loan. Not signed. Cap, price, and duration open. Parties far apart. Goal described as one-third support for a generic 90-day EJ, not a finished plan. Mathematical ceiling of moving about nine-day payables to 180 days is roughly US$950 million; do not assume the ceiling. Test US$250 million / US$500 million / US$950 million. Supplier credit is liquidity, not loss-absorbing capital.

This chair’s answer: unsigned naphtha does not by itself satisfy creditor burden-sharing. Combined with Idesa, it can satisfy controller burden-sharing for a thin EJ, at a high enough cap, because the US$476 million is already real cash leaving the group.

Shine will not replace unsigned Petrobras terms with a Shine cheque. If Petrobras cannot sign before Monday, the filing is a thin EJ on local-bank plus peeled-notes support, or it is RJ. It is not an IG4 recap.

7. Elliott / Contrarian and the one-third filing

Statute (verified). Lei 11.101 art. 163 §7: more than one-third of each affected class (espécie) to protocolar an EJ; more than 50% of each class to homologate. Pipeline’s “two-thirds to file” is wrong. If NY notes are one class, more than about US$2.35 billion of the listed ~US$7.02 billion parent securities is needed to file that class. Listed outstandings (verified, 30 June, US$ million): ’28 1,250; ’30 1,500; ’31 850; ’33 1,000; ’34 850; ’41 587; ’50 750; hybrid ’81 231. Remainder of the US$10.3 billion stack is the drawn RCF plus debentures, CRAs and bank / ECA lines.

Reported. Named holders: Elliott (international notes and reported largest RCF lender); Contrarian (hard group with Elliott on the Petrobras-cash ask, Bloomberg 10 August); Strategic Value Partners (notes and some RCF, reported as not a formal Elliott pairing); AllianceBernstein, Capital Group, PGIM (Bloomberg compilation). Local banks (Itaú, Safra, KfW, Bladex, DZ, SMBC and others) described as supportive of an EJ. Estadão: without some bondholders the company cannot reach one-third. O Globo (17:36 BRT, 18 August): company-side sources say Braskem is preparing to file an EJ next week, first filing a 90-day process plan, no asset sales in the first filing; the URL slug claims creditor approval, the body still recites the one-third bar. Do not treat O Globo as locked adhesion.

Will Elliott / Contrarian support a one-third filing without a Petrobras put? No. (Inference, from the Tuesday ask and the July menu.) They demanded a Petrobras commitment to invest if metrics are missed, as the price of supporting an EJ and a 90-day negotiation (reported, Pipeline). This chair cannot deliver that put. Petrobras will not. Therefore Elliott/Contrarian are outside the working-case filing coalition.

How a filing still happens.

1. Local banks supply the bank / ECA / debenture / CRA classes. Pipeline describes them as supportive. Those classes are not the notes class.
2. Peel the notes class with Capital Group, AllianceBernstein, PGIM, and any SVP-or-similar that will not die in a ditch for a Petrobras put. Target: >US$2.35 billion of the ~US$7.02 billion. This chair does not claim a public blocking percentage for Elliott and does not assume the steering committee is 51% of the notes (common-brief constraint).
3. File a thin 90-day plan-to-a-plan EJ: standstill, cash controls, information rights, Idesa left alone, naphtha unsigned or capped, no completed five-year economics, no Petrobras put, no Shine cheque. That is what “conditional support for a temporary plan-to-a-plan EJ with protections” in June actually was - not adhesion to Braskem’s five-year / three-year-grace / no-haircut offer (reported).
4. Homologation (>50% per class) is a different fight. A notes-out filing does not stay Elliott after 24 August, especially with the RCF. If homologation fails, the path is RJ. Shine prefers that to a converting deal.

Opening message to the long-onlys: you are not being asked to finance IG4’s option. You are being asked to preserve a Brazilian going concern through the election, with warrants as the upside instrument, while Elliott’s put is an instrument Petrobras will not sign and Shine cannot forge.

8. Opening, first concession, narrowest acceptable

Opening (Round 1 working case)

• Thin EJ next week. 90-day process plan. No asset-sale roadshow in the first filing (consistent with O Globo company-side color - reported).
• Five-year maturity extension, no principal haircut. Company still offering five-year extension, three-year grace, no haircut (reported, Pipeline). Shine supports that as the opening ask, not as already-agreed economics.
• PIK through December 2028 on the notes; cash pay at existing coupon thereafter. Opening is generous on cash coupon relative to what a completed economic EJ will require.
• Shine fund cash: US$0. Idesa US$476 million company cash is the deployed contribution.
• Petrobras: capped naphtha WC, target US$500 million, accept unsigned at filing if that is what Lei 13.303 timing requires, with a covenant to continue good-faith documentation. Not a put, not common equity, not a keep-well.
• Creditors: roll LCs (the US$572 million Q3 line item). Ask ~US$200 million incremental LC / WC capacity from the bank group. Not assumed.
• Warrants: none in the opening draft of the 90-day plan (it is a process filing). In parallel economic talks, opening warrant discussion is 10%, non-voting or limited-voting, on the fully diluted parent.
• Collateral: no blanket liens on core Brazilian assets. No pledge of crackers.
• Covenants: cash controls, information rights, no dividends, Alagoas ring-fence (see grid).
• Mexico: separate estate; remaining US$350 million allowed to fund; no further parent cash.
• Conversion / Petrobras put: not offered.

First concession (still not a Shine cheque)

• Warrants 10–15%, with a step-up of up to an additional 5% (to 15–20%) if cash-conversion, EBITDA, liquidity or PIK milestones are missed. Persistent PE–naphtha spreads are why contingent warrants dominate immediate equitization (inference).
• Limited non-core and WC collateral (receivables, inventory, non-core holdings). Not Camaçari, not Triunfo, not the cracker complex.
• Accept that the first filing is thinner than June’s economic ask: grace period shortened from three years toward a 12–18 month PIK window if that peels long-onlys.
• Accept signed naphtha at US$250 million rather than US$500 million if that is all Petrobras can document under Lei 13.303 before Monday.
• Still US$0 Shine fund cash.

Narrowest acceptable package (economic EJ, completed term sheet)

This is the tightest Shine will sign without walking to RJ. It is not a converting deal and not a Petrobras put.

• Five-year extension, no principal haircut. Mix of cash and PIK such that the package is not obviously negative-NPV at the current curve; Shine will not pre-concede a number that makes the AHG whole at par plus default interest. Positive-NPV is their word from June (reported); the instrument is duration plus warrants, not par cash.
• PIK floor: through at least mid-2027. Cash coupon thereafter at a rate that does not cash-starve Case A.
• LCs rolled for the tenor of the EJ / the 90 days at minimum; bank group asked to keep trade lines open.
• Petrobras: capped WC / naphtha, not a put. Shine will not fill a Petrobras-shaped hole.
• Shine fund cash: US$0 working case. Last-ditch US$125–150 million junior/hybrid only under the §3 conditions, and only if the alternative in the room is sub-10% conversion. Not the working case.
• Creditor new money: LC roll is the ask; incremental US$200 million is optional, not a Shine walkaway.
• Warrants: 15%, stepping to 20–22.5% on missed milestones. Old shareholders retain ~77.5–85% fully diluted. Shine economic stake at 15% warrants: 34.3234% × 0.85 ≈ 29.2%. If warrants vote, Shine voting ≈ 50.1108% × 0.85 ≈ 42.6% of fully diluted; Petrobras ≈ 47.03% × 0.85 ≈ 40.0%; combined ≈ 82.6%. Joint control intact. That is acceptable. Sub-10% is not.
• Collateral: limited non-core / WC only.
• Covenants: cash sweep while leverage is above an agreed level; information rights; no upstreaming; Alagoas protected.
• Mexico: as in the table. No second cheque.
• Rights offering: acceptable on a pro-rata basis if permanent capital is later required, with economically fair mechanics for BRKM3/5/6 and BAK where legally practical. Non-subscribers are diluted; that is not the same as a trough-price issuance solely to controllers (red line).
• Minority: preserve PNA economic preferences; parallel-class or equivalent rights.

Anything below this line on control - converting DIP, controllers <10%, all-asset cracker liens, Shine as sole cash sponsor - is a walk to RJ.

9. Term grid

10. Political overlay (labeled)

Verified. Election first round 4 October 2026, possible second round 25 October. Lula’s PT confirmed a reelection bid. Petrobras is state-controlled but subject to Lei 13.303. REIQ and Law 15.294/2025 (PRESIQ) exist as industrial-policy instruments. Provisional PE antidumping measures have been used; treat as scenario variables, not permanent protection. Alagoas remains live.

Inference - election-year double bind, applied to this chair. The two bad federal narratives are (i) public-company capital protecting IG4’s distressed option and (ii) a disruptive RJ that is blamed on foreign creditors and lost Brazilian influence. The path that sits between them is a commercially documented, capped, shared package: Petrobras WC, creditor duration, Shine not extracting dividends, warrants for creditors, Alagoas protected. An IG4-only cash recap recreates narrative (i) with Shine as the sucker and does not bind Petrobras. A Petrobras-only recap recreates narrative (i) with Shine as the free rider. Hence zero Shine fund cash and no Petrobras put is not only Pipeline-consistent; it is the political-equilibrium package. Last-ditch matching junior cash exists solely to kill the free-rider version of narrative (i) if conversion is the alternative.

Inference - control is politically relevant, not legally protected. A sudden foreign-led creditor takeover has labor, regulatory and reputational execution cost. That cost is why RJ residual equity is more valuable than a U.S. APR reading suggests. It is not a veto. Do not price a government backstop.

Inference - industrial policy. REIQ/PRESIQ and trade defense support going-concern value and therefore support extension-plus-warrants over fire-sale equitization. They do not protect today’s 50.1108% / 34.3234%. Do not capitalize temporary policy at full value.

Scenario assumption. Base case remains debt-plus-warrants around a pre-election framework, with implementation that can slip past 4 October. That is a thin EJ, not a completed economic EJ.

11. Outcome probabilities

Must sum to 100%. These are this chair’s probabilities as of the 18 August evening cutoff, not targets.

Modal path: thin EJ attempt. True risk: RJ. The US$300 million Shine facility is not in any of these working probabilities.

12. The single fact that would most change this position

Whether more than one-third of the NY notes class will protocolar a thin EJ without Elliott/Contrarian and without a Petrobras put.

If Capital / AllianceBernstein / PGIM (and any non-paired notes) plus the local-bank classes can deliver that filing, Shine stays at US$0 fund cash, Idesa funds, and the call-plus-joint-control package is intact through the election. If they cannot, the working case is RJ (accepted) or the last-ditch matching US$125–150 million (not the working case, and only behind a signed Petrobras WC). Everything else in this memo - naphtha cap, Q3 EBITDA print, even the remaining Idesa US$350 million timing - is second-order to that filing-coalition fact.

13. Red-line recap (so there is no later ambiguity)

1. Working-case Shine fund cash = 0. Pipeline 18 August is the constraint. US$300 million is not a base-case gift.
2. Idesa US$476 million is company cash. Majority at Idesa is an option purchased with that cash. No second cheque. No parent guarantee of residual Idesa debt. No credit as parent deleveraging.
3. Call plus joint control is the product. 50.1108% voting / 34.3234% total. Conversion below 10% rejected. All-asset cracker liens rejected.
4. Not willing to be wiped. Willing to risk RJ because residual equity survives Brazilian process. Distinguish those.
5. Will not be the sole cash sponsor of US$10.3 billion. Hide behind Petrobras on any recap. Last-ditch junior cash only as matching, political, control-preservation insurance, and labeled as not the working case.
6. Warrants may be acceptable. A Petrobras put is not Shine’s to give.

Sources (public)

• Braskem ownership: https://www.braskem-ri.com.br/en/the-company/ownership-structure/
• Shine acquisition 6-K: https://www.sec.gov/Archives/edgar/data/1071438/000129281426002421/bak20260420_6k.htm
• Idesa 6-K, 18 August: https://www.sec.gov/Archives/edgar/data/1071438/000129281426004279/bak20260818_6k.htm
• Idesa announcement: https://www.prnewswire.com/news-releases/braskem-idesa-sapi-takes-strategic-action-through-consensual-restructuring-to-reduce-its-debt-by-more-than-us920-mm-302853634.html
• Idesa docket: https://restructuring.ra.kroll.com/BraskemIdesa/
• Pipeline Valor, 18 August: https://pipelinevalor.globo.com/negocios/noticia/o-impacto-do-chapter-11-da-braskem-idesa-para-a-holding.ghtml
• O Globo, 18 August: https://oglobo.globo.com/economia/negocios/noticia/2026/08/18/braskem-consegue-aval-de-credores-e-prepara-pedido-de-recuperacao-extrajudicial.ghtml
• Lei 11.101 art. 163; Lei 13.303; TSE 2026 calendar; PRESIQ Law 15.294/2025; IG4 site https://ig4capital.com/