Round 1 - Braskem S.A. Management & Board

Date 2026-08-15. V = VERIFIED (source). I = INFERENCE. FX US$/R$ 5.22 at Jun-30 (derived: R$36,744M bonds ÷ US$7,030M face, V q2itr n.15(c) vs Lazard p.030).

Post-run correction, 16 August 2026. Q2's US$547M working-capital outflow is not treated as a quarterly run rate. The H2 operational release in the June plan is US$518M, not US$548M. A targeted role review confirmed that this changes the cash-conversion framing but not the need for maturity extension and replacement trade finance. The refreshed mediator estimate uses 15% upfront warrants, stepping to 22.5% on cash-conversion, liquidity or PIK misses.

OCR corrections to Lazard p.030: Bond 2041 is US$579M not 877 (R$3,021M ÷ 5.22; 7.125% × 579 = US$41M interest, matches deck). Bond 2081 is US$250M not 207 (12.004% × 250 = US$30M, matches). Corrected stack ties: bonds 1,189+1,510+872+1,016+853+579+761+250 = 7,030 = R$36,744M ÷ 5.22 ✓.

1. Situation assessment - liquidity, runway, minimum need

Debt stack (Apr-30-26, US$m, V p.030 as corrected): Bonds 7,030 · Debentures/CRAs ~656 · RCF/stand-by 1,004 (31-Dec-26, SOFR+2.98%) · Bilateral 413 · ECA/MLA 381 → 9,484 (deck total 9,497). + leases ~759 (V n.13(b), R$3,964M ex-Idesa) = US$10.24B ≈ corporate gross debt US$10.3B (V q2release 4.9.1). Net US$9.5B, 6.74x (V). Contractual cash interest US$654-665M/yr (V p.030).

2H26 contractual debt service = US$2,349M (derived & tied to V p.005: Jul 549, Aug 152, Sep 177, Oct 234, Nov 118, Dec 1,119; 3Q26 sum 878 = deck's 878 ✓, of which LCs 572). Jan-28 bar 1,326 (bond 1,189 + interest).

Persistent-spread case (A). Plan monthly 2026 EBITDA (V p.022) sums to 2,166 ✓. Actual 1H = 192 (1Q, V q2pres) + 1,043 (2Q, V) = 1,235 vs plan 1,229 - the plan converts spreads to EBITDA accurately; it is only a price-taker. Plan 2H26 = 937. Persistence delta uses Braskem's own effective spreads (V p.020): domestic PE 2Q26 1,738 → plan 3Q 1,088 / 4Q 977; PP 622 → 390 / 305. Domestic resin volume 763kt/qtr (V q2pres), split PE 412 / PP 252 / PVC 99 (I).

• 3Q: 412×650 = 268; 252×232 = 58; exports +38; chemicals +40; US/EU +25 = 429; −20% inventory-cost-lag haircut = +345
• 4Q: 412×761 = 314; 252×317 = 80; +45/+40/+25 = 504; −20% = +403
• 2H26 EBITDA ≈ 937 + 748 = 1,685; FY26 ≈ US$2.92B (range 2.62–3.19B).

Q2 cash conversion should not be treated as a run rate. 2Q26 evidence: EBITDA 1,043 → WC (547) → capex (112) → op cash 385 → recurring cash generated 210 = 20% conversion (V q2pres p.12). Braskem attributed the WC outflow to higher feedstock and product prices, higher inventory volumes and reduced payment arrangements. The price and inventory components are one-time level changes if naphtha and volumes stabilize. Persistence may reduce the plan's assumed 2H26 WC release of +518 (V p.024), but forgoing a release is not the same as repeating the Q2 outflow. Model flat operational WC as a central case and the full release as upside. → 2H26 incremental CFADS = 748 − 480 − 60 = +208 (27.8% conversion). Plan 2H26 CFADS = (29) → persistence +US$179M vs US$2,349M of debt service (7.6% coverage).

Trade finance is the real constraint even if operational working capital stabilizes. LC-backed trade payables R$7.8B→R$5.9B in 6M (V n.14); LC book 1,274 (Jun) → 502 (Dec) → 0 in 2027 (V p.026/027); reverse factoring gone (V n.1); supplier advances R$544M→R$1,625M (V n.1); R$929M + R$703M of LCs drawn by banks and reclassified as loans (V n.1/14). WC-financial drain 2H26 = (860) regardless of spreads. Client factoring: US$769M revolving, ~US$6.4B/yr originations (V p.027).

Runway, Track A (stay holds / EJ filed, financial debt service suspended). Opening 15-Aug unrestricted ≈ US$800M (I: plan EoP-Aug 444 + suspended Jul/Aug cash debt service 295 + ~60 persistence). Sep (39) → 761 · Oct +137 → 898 · Nov +55 → 953 · Dec +84 → ~1,037 at 31-Dec-26. Runway to ~mid-2027, when the LC book hits zero and 2027 CFADS-before-interest is ~(360). Track B (stay lapses ~Aug 25, no filing): all US$9.5B current (V n.1: non-current reclassified to current from Jul-26); factoring stops (769 over 60-90d); LC issuance stops (200-400); Petrobras terms compress (85-170) → cash negative in 45-75 days. Not survivable.

Minimum NEED vs WANT.

2. Fiduciary framing

Negative consolidated equity R$13,087M (parent R$12,586M), negative net WC R$8,701M, material-uncertainty going-concern emphasis (V q2itr auditor + n.1). Equity has no accounting value; creditors are the residual claimants.

• LSA art. 154: duty runs to the company. §1: a director elected by a group has identical duties and may not favour those who elected him - dispositive for IG4- and Petrobras-nominated directors. Art. 156: conflicted directors must disclose and abstain (Petrobras is simultaneously 36.15% holder, R$5,098M/H1 feedstock supplier (V n.8) and prospective new-money provider). Art. 155 loyalty; art. 117 abuse of control (incl. §1(b), steering a reorganization for undue advantage); art. 116 sole ¶ duties to minorities/employees/community; art. 158 personal liability.
• Lei 11.101 arts. 129/130 (clawback in the termo legal), 168/172-173: no selective payments to financial creditors, no non-ordinary transfers to sponsors or to Braskem Idesa.
• Conclusion the board must state on the record: it cannot reject a superior creditor proposal to preserve Shine's 50.11% voting or Petrobras's 47.03%. Mechanism: an independent committee of non-sponsor directors with its own counsel/banker to score every proposal against the RJ counterfactual, with sponsor nominees abstaining under art. 156. What the board can defensibly reject is a proposal worse for the company than RJ (§5).
• Hard constraint (V n.24.1): authorized capital 1,152,937,970 vs 797,207,834 outstanding = 355.7M shares of headroom, of which only 84.0M common. A 650M-share raise requires an EGM; both sponsors must agree (joint control, consensus on relevant activities, V n.1). Fallback executable by board resolution alone: up to 271.6M class-A preferred ≈ US$416M.

3. RJ vs EJ vs out-of-court

Out-of-court exchange: needs ~90%+ across 8 series against an organised AHG. Dismiss.

EJ (arts. 161-167). File with ≥1/3 adhesion → 90-day stay (art. 163 §7); homologation needs >50% of each affected class. Trade, employees, tax and Alagoas stay outside and unimpaired; no administrador judicial; management retained; Ch.15 recognition already seeded (V n.1, granted 30-Jun). Impossible without AHG support - but the AHG has already said in writing it "is willing to support entry into any 'plan to a plan' EJ proceedings, subject to appropriate parameters and creditor protections" (V p.014).

RJ operating damage (quantified, I unless flagged): LC replacement/cash collateral 400-700 · factoring loss 500-769 · supplier-terms compression 200-400 = US$1.1-1.9B of cash need in the first 90-180 days vs US$0.8B on hand → RJ requires a US$1.0-1.5B DIP on day one. Feedstock: Petrobras payable R$257M on R$5,098M/H1 ≈ 18 days - 30 days of prepayment = US$170M. Utilization 70% (V) could fall 10-20pp → US$300-700M of lost EBITDA over 12-18 months. REIQ/PRESIQ benefits at risk: US$305M (2026) + ~US$225M/yr 2027-31 Insumos, plus 114/47/45/29/16 Investimentos (V p.019) - art. 57 tax-regularity exposure. Customers already substituting (V q2pres: PE/PVC volume down on higher imports). Ratings already C/D (V n.1). Duration 2-4 years.

Conclusion: EJ dominates for the company across most of the range. Immediate, unconditional move (highest value / near-zero cost): accept the AHG's June 19 protection list in full - non-ordinary-course and related-party restrictions binding on debtors and non-debtors, capped fee reimbursement, deep-dive diligence access, Petrobras at the table within 20 days - and file the EJ on or before Aug 24 with a parallel motion to extend the tutela. This is separate from the economic negotiation.

3B. Brazil political economy - a first-order input, not colour

The election window. First round 4 Oct 2026, runoff 25 Oct - ~7 weeks out. An RJ filing by Brazil's largest petrochemical producer inside that window is a national political event, not a corporate one. Consequences the board must price:

• Asymmetry of timing. Filing before 4 Oct maximises political cost and minimises political help: no minister, no BNDES board and no Petrobras executive will authorise discretionary support for a company that has just filed, in the middle of a campaign. Filing (or being forced to file) after 25 Oct is materially cheaper politically but a transition period runs to 1 Jan, freezing decision-making through exactly the Dec-26 RCF maturity. The realistic window for state-linked support is now through late September, or not until Q1-2027. The Dec-26 wall sits precisely in the dead zone. This argues for closing the transaction by end-September, not December.
• Conversely, it is leverage. No incumbent government wants Camaçari, ABC Paulista and Triunfo/Rio Grande do Sul headlines in October. That raises the probability of tariff renewal, PRESIQ generosity and BNDES accommodation - but only for a company that is negotiating, not one that has filed.

Petrobras's constraints. As a sociedade de economia mista (Lei 13.303/16), any Braskem capital injection is a related-party transaction requiring an independent valuation, a statutory-committee opinion and board approval, is disclosable under CVM rules, and is subject to TCU review and minority-shareholder challenge (Petrobras has its own ADR/minority base and a history of litigation over Braskem exposure). Practical read for the board: Petrobras cannot commit new equity quickly, cannot commit it at a price that looks like a subsidy, and cannot be the first mover. Expect 8-14 weeks from term sheet to a Petrobras board resolution - which does not clear before the election. Therefore the board should not build the plan around Petrobras cash on the critical path. What Petrobras can deliver fast, at low political cost and with no independent valuation, is commercial: extending naphtha/utility payment terms from ~18 to 60 days releases ~US$225-280M of permanent working capital (R$10,196M annualised ÷ 365 × 42 days) and waiving insolvency/change-of-control termination rights in the supply contracts. This is the single highest-value, lowest-cost ask in the negotiation and it should be Ask #1, ahead of cash.

Industrial policy already in the file (all V, q2itr n.1). REIQ raised 0.73% → 5.8% by LC 228/Mar-2026, capped R$2bn sector-wide, expiring 31-Dec-2026; PRESIQ (Law 15,294/25) effective 1-Jan-2027 to 31-Dec-2031; definitive five-year antidumping duties on US/Canada PE (GECEX 876/2026, Apr-2026); and the 20% import tariff on PE/PP/PVC expiring 16 Oct 2026 (GECEX 800/2025). Two discrete events to flag in the model: the REIQ→PRESIQ handoff at 31-Dec-26, and the 16-Oct-26 tariff cliff, which falls between the two election rounds. Renewal is plausible precisely because of that timing, but it is not in the base case and its loss is a direct 4Q26/2027 margin hit. The board should seek renewal before any filing - a company in RJ is a far weaker supplicant at GECEX.

Employment and regional politics. Camaçari (Bahia), ABC Paulista (São Paulo) and Triunfo (RS) are concentrated, unionised, politically salient sites feeding a large SME transformation chain. RS carries additional post-flood sensitivity. This is the source of judicial protectiveness: Brazilian courts have consistently privileged preservation of large employers under the princípio da preservação da empresa (Lei 11.101 art. 47) - Oi, Samarco, Americanas. Read honestly, this cuts both ways: it means an RJ court would likely grant generous stays, extend the 180-day period and tolerate a long process - which reduces the AHG's ability to threaten a fast liquidation, and is genuine negotiating leverage for the company. It also means the AHG's collateral will be slow and contested to enforce.

Alagoas remains an unresolved political liability: total provision R$18.2bn, R$14.6bn disbursed, ~R$3.2bn net remaining, 32% current (V q2pres); R$10,827M of "possible loss" litigation (V n.23). Any structure that appears to subordinate Alagoas remediation to bondholder recoveries is politically unsurvivable and, in an EJ, legally impossible (not a financial claim). Keep it wholly outside and pre-funded.

BNDES is the swing financier for Transforma Rio (~US$900M strategic capex 2026-30, V p.019). The AHG expressly objected to using internally generated funds rather than BNDES for it (V p.013). Here the board and the AHG agree, and the board should say so: strategic capex will be BNDES/ECA/project-financed or deferred. But note the political dependency - BNDES lending to a filed RJ debtor is far harder than to a company executing a consensual EJ.

4. Round 1 opening package - "Recapitalization & Reinstatement Plan"

Structure: EJ, two classes - (I) Financial Unsecured; (II) LC/Trade Finance. Trade, labour, tax, Alagoas unimpaired. Ch.15 recognition. Target close 30-Sep-26 (political window), hard deadline 31-Dec-26.

New money - US$1.70B of capacity, US$1.40B cash:

Treatment by bucket (yrs 1-2 = 24 months from close):

Fees: consent 100bp on ~US$7.8B = US$78M PIK · new-money 20 · LC 21.5 · AHG advisors capped US$45M + US$15M completion (V p.014 demand accepted, capped) · company advisors ~60. Cash ≈ US$105M, total economic ≈ US$240M (deck assumed US$100M, V p.028).

Collateral. First lien (new money + LC, pari): WC pool - eligible Brazil inventory, non-factored receivables, collection accounts, equity of Braskem Trading & Shipping B.V.; 1.35x borrowing base, US$2.0B cap. Second lien (all reinstated claims ex-2081): silent 2nd on the pool + 1st on a non-core basket (Braskem Green, Wise, Oxygea, 8 wind SPEs, Cetrel, RPR, Bioglycols; ~US$300-500M, I). Red lines: no liens on the Brazilian crackers, Alagoas assets, Braskem Idesa equity or REIQ/PRESIQ credits. New guarantees from Braskem Netherlands, Netherlands Finance, America Finance and Trading & Shipping - closes the structural gap where the RCF is guaranteed by Braskem S.A. "only" (V p.032). Real, cheap consideration.

Warrants. 15.0% FD at close, penny-strike, 7yr, to Class I pro rata. Step-ups tested 2027-29: +2.5% each if LTM EBITDA <US$2.0B at 31-Dec-27 · liquidity <US$750M any quarter-end · PIK used beyond 31-Dec-28 · LC utilisation >90% two quarters. Max 25.0% FD. Convertible into common (deliberate: the board concedes theoretical control transfer on failure rather than defend sponsor control). No equitization at close.

Dilution. Base 797.2M → post-raise 1,449.7M. 15% warrants = 255.8M (FD 1,705.5M); 25% = 644.3M (FD 1,932.9M).

Non-subscribing minorities lose 53% of their ownership share; subscribing minorities lose only the warrant dilution. Board obligations: full 30-day preemptive period, ADR holders enabled via the depositary (Reg S/144A tranche), sobras round, non-punitive backstop fee, price set by the independent committee.

Contributions. Petrobras: US$361.5M + up to 151.5M backstop = US$513M, plus the 60-day payment terms (US$225-280M of permanent WC) and a 5yr supply agreement with insolvency-termination waivers. Shine/IG4: US$343.2M + 143.8M = US$487M - and note Shine acquired control by delivering NSP debentures, not cash (V 6-K 5-Jun-26), so the board will not defend a Shine contribution below US$300M cash; Shine must also release the ex-NSP bank restrictions on its shares. Creditors: 400 new money + 300 new LC + ~800 roll + US$285M/yr cash-interest relief. Company self-help: 2026 capex 465 → 380 (V US$465M plan), SG&A −US$50M, strategic capex BNDES/ECA-funded or deferred, no dividends.

Governance. Board of 11: Petrobras 3, Shine 3, creditors 3, independents 2. Creditor consent over non-ordinary transactions, related-party >US$25M, incremental debt/liens, Idesa investments, amendments to the Petrobras supply agreements (mirrors V p.014). Weekly 13-week cash flow; quarterly borrowing base. CFO/Treasurer appointments need creditor-director consent while PIK is on.

Mexico / leakage. Braskem Idesa fully ring-fenced (US$900M 2029 + US$1.2B 2032; R$11,650M, V n.16), separate Ch.11. No new equity, loans or guarantees beyond the existing US$82M WC loan due Dec-26 (secured on BI assets, V n.1) and the pre-existing TQPM Equity Support Agreement - disclose and cap at 50% of R$1,789M ≈ US$171M, rising to ~US$343M on collateral perfection (V n.16(v)). No dividends/JCP until net leverage <3.0x. Nothing to Novonor/NSP.

Alagoas. Outside the compromise; US$150M dedicated reserve account at close plus a first claim on operating cash for scheduled payments, senior to the sweep.

Covenants. Min unrestricted liquidity US$900M monthly (warrant step-up <750, EoD <600) · 50% excess-cash sweep above US$1.20B · 75% asset-sale sweep >US$50M/yr · capex cap US$500M maintenance · 3.5x incurrence + US$250M basket · trade finance carved out.

Cash interest (US$m/yr).

Post-deal liquidity at close (31-Dec-26): opening ~970 + equity 1,000 + notes 400 − fees 105 = US$2.27B cash + US$300M undrawn LC (normalization case ≈ US$2.0B). 2027 pro forma (normalization, EBITDA 1,494): +99 WC-op (V p.027), WC-financial ~0 (the LC facility replaces the (602) runoff), capex/leasing/Alagoas/taxes (1,140), strategic (150) → CFADS ~303; interest (430); amort (40) → (167) → cash ~US$2.10B. 2028 ≈ (130) → US$1.97B.

Pro forma leverage. Gross ex-lease at close = 9,484 + 78 + 400 = US$9,962M; net = 7,692 (cash 2,270). End-2028 after PIK: gross 10,862, net 8,892.

5. Rejection terms and the RJ-dominance test

Reject: (R1) first liens on the Brazilian crackers or an all-assets pledge - it destroys the unencumbered base that supports LC/factoring, i.e. RJ damage without RJ benefits. (R2) mandatory equitization >35% without ≥US$2.0B of real principal cancellation - the test is dollars cancelled per point of equity, not the percentage. (R3) any pre-2029 cash amortization of pre-petition principal beyond the ECA schedule. (R4) cash interest >US$450M/yr in yrs 1-2 or >US$650M/yr from 2029. (R5) closing liquidity <US$900M or LC facility <US$900M - monthly gross feedstock/utility outflow is ~US$700-900M on US$13.0B COGS (V p.023). (R6) consolidation of Braskem Idesa's ~US$2.1B, or new parent support to it. (R7) any impairment of Alagoas, labour or the Petrobras supply relationship. (R8) forced sale of Braskem America/Europe (2Q26 EBITDA US$147M) to fund the deal. (R9) uncapped creditor advisor fees or a >5% backstop. (R10) any proposal conditioned on removing a specific sponsor - that is a control fight, not a restructuring.

RJ dominates for the COMPANY when: (i) post-deal 2029 cash service exceeds 2029E CFADS-before-interest in the normalization case; (ii) the collateral demanded removes >~US$700M of LC/factoring capacity - at which point RJ's art. 69-A super-priority DIP is a cheaper source of the same money; (iii) new money <US$400M while liens are demanded - creditors taking the lien without funding it; (iv) closing unrestricted liquidity <US$1.2B, given the 16-Oct-26 tariff cliff and the REIQ→PRESIQ handoff.

Test (i) already fails under my own opening. 2029E: EBITDA 1,710 (V p.023) + WC (71) − ~1,150 ≈ CFADS 490 vs cash interest ~860 → a US$370M deficit. The board must record this: the year-3 coupon step-up is not survivable at plan EBITDA, and the board should expect to trade the step-up away for additional warrants - pay in equity, not cash. My opening is an opening, not a reservation price; the true reservation price is materially worse for shareholders.

6. Strongest argument against us, and the answer

The argument. Braskem's own plan says 2027E US$1,494M and 2028E US$1,524M (V p.023). Against US$9.5B that is 6.4x gross before PIK, ~7.2x after. Contractual cash interest US$654M = 44% of 2027E EBITDA, before US$600M capex, US$200M leases and US$250M Alagoas. Extension plus PIK converts a 2026-28 liquidity problem into a larger 2031-33 solvency problem. Braskem itself calls 2Q26 "a tactical value capture opportunity in a volatile environment, rather than a structural change in the cycle," with global PE surplus +2.6Mt (2026) and +3.7Mt (2027) at 77%/76% operating rates (V q2pres). Equity is negative R$13,087M. IG4 paid in NSP debentures, not cash; Petrobras has written no cheque. The rational creditor demand is conversion of US$2.5-3.5B into majority equity, not warrants.

Answer - three parts, the first conceded.

1. Concede. At US$1.5B EBITDA the structure is unsustainable and equity has no accounting value. Creditors are entitled to positive-NPV compensation and real upside; sponsors must contribute loss-absorbing cash. Hence, unlike June: US$1.0B of permanent equity, a coupon increase not a 200bp cut, first liens for new money, up to 25% creditor equity convertible into voting stock, and three board seats. The June proposal was not a clearing proposal and the AHG was right to reject it.
2. Factually. (a) US$1,494M is not a fact - it is a May-2026 GMI forecast on a 2027 PE-naphtha spread of US$377/t (V p.007). The Q2 deck is expressly "updated as of June 30, 2026" (V q2pres disclaimer) and its 3Q26e of US$316/t was contradicted by July Platts and mid-August observation. The same process forecast 2Q26 at US$700/t and got 773, forecast US$1,052M of quarterly EBITDA and got US$1,043M. Persistence puts 2026 at US$2.9B (2.62-3.19B). A structural downcycle and a 12-18 month supply-shock window are not mutually exclusive; converting principal now prices the equity off a forecast already wrong twice. (b) More important: conversion is a balance-sheet answer to a liquidity question. LC-backed payables R$7.8B→R$5.9B; the LC book goes to zero in 2027; reverse factoring is gone; supplier advances tripled; R$1,632M of LCs already drawn and converted to loans. An equitized Braskem with US$6B of debt and no LC capacity cannot buy naphtha and is worth less than a levered Braskem with US$1.1B of committed trade finance.
3. Procedurally. If the AHG delivers more new money, more trade-finance capacity and lower total cash service than the sponsors, the board will recommend it even if Shine and Petrobras fall below control - and will say so on the record, through an independent committee with its own advisors. What the board will not accept is a proposal that takes the equity and leaves the trade finance unsolved. The test is the company's, not the shareholders'.