Transmissora Aliança de Energia Elétrica Deep Dive: Contracted RAP Growth, a 2030 Cliff
An Availability Franchise, Not an Energy Business
Transmissora Aliança de Energia Elétrica, traded as TAEE3, TAEE4 and the unit TAEE11, is a pure-play owner-operator of regulated transmission concessions in Brazil. It does not sell megawatt-hours. Revenue is Annual Permitted Revenue, or RAP, set by ANEEL and paid for keeping lines and substations available, not for the energy that flows through them. The economic customer is the National Interconnected System: generators, distributors, free consumers and exporters pay transmission-use charges that are pooled and settled through the system operator. Volume risk is structurally absent, and credit losses have historically been negligible because non-payment is a system problem, not a bilateral counterparty problem. That architecture is why regulatory EBITDA margins sit in the mid-80s — 83.7% for full-year 2025 on regulatory net revenue of R$2.51 billion, and 85.4% in the first half of 2026 — while cash costs are dominated by a thin operations base rather than fuel, offtake or commodity inputs.
Indexation and contract vintage determine how that RAP compounds. Category III concessions, the post-2006 book, reprice annually by IPCA; Category II assets, the 1999–2006 cohort, reprice by IGP-M and have already absorbed the 50% RAP reset that followed the earlier tariff reviews. In the cycle that began on 1 July 2026, IPCA assets were adjusted by 4.72% and IGP-M assets by 1.95%. The growth book therefore inflates faster than the legacy book, and the mix shift over the next several years is quietly improving revenue quality even before any new project is considered. Concession length on the newer assets runs into the late 2040s and early 2050s. The investment case is the quality and duration of that RAP stack, not a view on Brazilian power demand.
Who Pays, Who Competes, Who Supplies
There is no concentrated industrial customer to lose. End demand for the wires sits with the load centers of the Southeast and the generation surplus of the North and Northeast, but TAESA’s cash flow does not depend on any generator’s dispatch or any distributor’s volume. The relevant counterparties are ANEEL, which sets RAP and the rules for reinforcements, and ONS, which certifies availability and applies the variable portion — the penalty for unavailability. In 2025 that penalty was 0.53% of transmission revenue; in the first half of 2026 it was 0.73%, against line availability of 99.91%. Operational underperformance is a margin leak measured in tens of basis points, not a thesis risk, provided that standard holds as the network gets longer and older.
The competitive set is other concessionaires, not technology challengers. ISA Energia Brasil is the larger listed platform, with an operational network above 20,000 km and a RAP base that still exceeds TAESA’s. Alupar is the growth-oriented listed alternative, consistently willing to take greenfield lots. State Grid and, in the 2026 auctions, Axia have been the aggressive bidders on new capacity, with pension and infrastructure capital — Perfin via Infra 2 alongside Alupar — also clearing lots. Energisa remains both a peer and a recent seller: the five concessions TAESA bought were 27% of Energisa’s transmission RAP, leaving Energisa with the majority of its own wires book. Equipment and EPC capacity — converters, transformers, line contractors — is a project-timing constraint, not a structural supplier dependency; once energized, the asset’s economics are regulatory, not procurement-driven. TAEE4, the preferred line, has essentially no controller ownership: Cemig and ISA hold the voting block, and preferred shares are almost entirely free float. Economic exposure is liquid; control is not.
What the Franchise Actually Owns
Control sits with Cemig, at 36.97% of ordinary shares, and ISA Investimentos, at 26.03%, bound by a shareholders’ agreement covering 63% of voting capital. Together they appoint the majority of a 13-member board; four seats are independent. The economic capital is different: the market holds 63.44% of total shares, concentrated in units and preferred stock. That split matters for TAEE4 holders. Dividend priority is contractual; strategy, related-party dealings and the appetite for leverage are set by a state-controlled utility and a Colombian transmission group. The operating partnerships mirror the control group. AIE, owned 50/50 with ISA Energia, holds Aimorés, Paraguaçu and Ivaí. TBE, alongside Alupar, holds a cluster of shared concessions. The corporate structure has been simplified toward the holding company — 14 concession contracts sit there — to improve tax and financing efficiency, which is the right use of a holding that already consolidates a large share of RAP.
Scale, on a proportional basis after the August 2026 close, is 49 concessions across 18 states and the Federal District, on the order of 16,600 km of lines and roughly 160 substations, with transformation capacity lifted about 33% by the acquired assets to around 18,000 MVA. Proportional RAP for the 2026–2027 cycle is R$4.9 billion, of which about R$4.3 billion was already operational at the end of June before the acquisition fully contributed. Geographic spread cuts single-corridor weather and permitting risk, but it does not diversify regulatory risk: every real of RAP is an ANEEL number. The EBITDA margin above 83% for a decade is the market’s evidence that the model works when availability is held near 100% and opex grows below inflation, which management again delivered in the first half of 2026.
Growth That Is Already Contracted
The 2026–2027 cycle is a step-change, not a forecast. ANEEL Dispatch 2,268 of 22 June 2026 set the new RAP, and the company’s own bridge shows a 22.3% increase, or about R$886 million, versus the prior operational base. Roughly R$341 million of that came from energizations over the prior year and into July — partial entries at Tangará, Saíra and Ananaí, full contribution from Pitiguari, and reinforcements at TSN, São Pedro and ATE III — plus the inflation adjustments. The Energisa portfolio, closed on 12 August 2026, adds a further R$305 million of RAP in the same cycle, 1,305 km and 12 substations, at an equity price of R$1.638 billion after CDI adjustment on a R$1.545 billion December 2025 base, and an enterprise value of R$2.293 billion. Average remaining life on those five concessions is about 22 years. Residual construction, mainly the last phases of Saíra and the reinforcement book, is another R$110 million of RAP still to be enabled. As of today, 12 October 2026, none of this has been visible in a full reported quarter: the new cycle started 1 July, the acquisition closed mid-August, and third-quarter results are due 11 November.
Delivery timing has been a genuine source of value, not a slide-deck claim. Tangará, Ananaí and Saíra were energized as much as 25 months ahead of ANEEL deadlines, pulling RAP forward. On the Saíra concession — Lot 5 of the December 2022 auction, the former Enel Cien Garabi interconnection with Argentina — TAESA completed what it describes as Brazil’s first HVDC converter retrofit, Garabi II, about 20 months early, enabling an additional R$24.2 million of RAP and taking the concession to 97.5% of its R$200.8 million cycle RAP. The technical point is narrow but investable: the company can execute converter work on an existing 2.2 GW cross-border link, which is relevant to a grid that will keep needing HVDC reinforcement between the Northeast resource and the Southeast load. It is not, by itself, a new earnings line. Regulatory net revenue in the first half rose 9.5% and regulatory EBITDA 10.5% to R$1.139 billion. Regulatory net income did not follow: R$399.4 million in the half, down 16.4%. The gap is financial expense in a high-rate book that just absorbed R$1.7 billion of debentures for the acquisition, within R$2.5 billion of debenture issuance in 2026. EBITDA growth and earnings growth are not the same object here.
The Duration Problem That Dominates the Equity
The offset to that contracted growth is already dated. Category II concessions begin to expire in 2030 — TSN, Novatrans, ETEO and ECTE among them — and sell-side work puts roughly 40% of RAP at maturity by 2032. These are the assets whose capital component was largely reset in the prior reviews; what remains is still large in absolute terms, with TSN and Novatrans alone historically in the vicinity of R$500 million of RAP each. Under current rules, renewal is at the grantor’s discretion. The concessionaire must declare interest at least 36 months ahead, and ANEEL must respond by the 18th month before expiry. Renewal would reopen the economic terms, including RAP, to reflect that the original investment has been remunerated. Rebidding would do the same from the other direction. Either path removes the investment-remuneration slice of RAP.
Management has stopped hedging that point in public. In early September 2026, chief executive Rinaldo Pecchio stated that RAP on these contracts will fall whether the government chooses renewal or rebidding, and that there is no scenario in which revenue stays at today’s level. The company expects an ordinance setting the treatment of expiring concessions by the end of 2026 — a catalyst inside the current quarter — and has not ruled out litigation if compensation for sunk investment is not settled. Moody’s Local has estimated the non-amortizable asset base on the 2030 cohort at about R$3 billion before inflation, which is the cash that would, in principle, come back if the concessions are not renewed on terms that keep that capital inside the RAP. The form of payment — lump sum or installment — is itself unresolved, and an installment schedule would turn a balance-sheet event into a multi-year cash trickle. The Energisa assets lengthen average portfolio life by about a year. That is a mitigation, not a solution. Investors underwriting TAEE4 through the early 2030s are underwriting a regulatory outcome that management itself describes as a revenue decline, with the quantum and the cash timing still open.
Why Greenfield Auctions Are the Wrong Use of Capital
The expansion cycle in the wires is real and will continue. Northeast wind and solar still need evacuation, São Paulo’s load — including data-center demand competing for scarce transmission margin under the 2026 access season — needs connection, and ANEEL’s Auction 4/2026, set for 30 October, offers eight lots, 1,866 km and R$8.8 billion of investment against a maximum RAP of about R$1.63 billion. The problem is the clearing price. The second phase of Auction 1/2026, in July, averaged a 53.21% discount to the RAP ceiling; the full auction averaged 51.58%. Axia took three lots at discounts of 51.8% to 59.0%. Alupar and Perfin took the largest lot at a 52% discount. A bid at half the regulatory ceiling is a bid at roughly half the return ANEEL embedded in the cap, before construction, permitting and cost overrun. That is not a market TAESA should be trying to win on volume.
Management has acted on that arithmetic. After signing the Energisa agreement in May, Pecchio described integration as plan A, B and C for the rest of 2026, and the company was selective rather than aggressive in Auction 1/2026. The multiple paid for operating assets — Energisa’s framing was 8.3 times EV/RAP and 10.1 times EV/EBITDA on the 2025–2026 cycle and trailing EBITDA — is the rational alternative to a 50% auction discount. Reinforcements and improvements, where competition is an authorization process rather than an open auction, are the other. The company has a book of these, with incremental authorized investment above R$190 million, and ISA Energia’s own R$7.2 billion authorized reinforcement pipeline is a reminder of how large this channel is for incumbents who already sit on the corridors. The growth that compounds shareholder value at TAESA is early energization of what it already won, plus operating assets bought from sellers who want leverage down, not new 30-year concessions won at a 50% haircut.
Management: Execution Has Been Better Than the Earnings Print
Pecchio has been chief executive since August 2024, after serving as CFO from May 2023 and as interim CEO from February 2024. His prior operating history is ISA CTEEP, AES Brasil and Elektro — transmission and distribution, not generation merchant risk. The bench around him is newer than the franchise: Catia Pereira as CFO since February 2025, from Sabesp and industrial issuers; Luis Alves on the technical side since 2021, from Neoenergia; Jell Andrade on implementation since late 2024, from Sterlite, GE Grid and Alstom; Mauricio Dall’Agnese on new business since May 2025. Continuity sits in the asset base and in the ISA operating culture, not in a long-tenured C-suite. The 2024–2026 record is nevertheless specific. Regulatory EBITDA in 2025 was R$2.10 billion, up 12.5%, the highest in the company’s history, on opex down 10.7%, with a 100% payout of regulatory net income of R$1.12 billion. Projects have been delivered early. The Energisa transaction moved from signing in May to ANEEL prior consent in July to close on 12 August, with shareholder approval in between. That is a three-month regulatory path on a R$2.3 billion enterprise-value deal, which is fast for this sector.
The misses are financial, not operational. Proportional net debt to regulatory EBITDA was 4.1 times at the end of 2025, after a record capex year of R$1.78 billion. The 2026 debenture program, including R$1.7 billion earmarked for the acquisition, adds duration-matched funding — average life on the book has been managed in the five-year area — but it also explains why regulatory net income fell in the first half while EBITDA rose. Distributions have not been cut to defend a leverage target: the board has continued to pay out 100% of quarterly regulatory net income, with R$1.036 billion declared year-to-date through the August decision, including R$206.8 million from the second quarter payable on 26 November. Full-year 2025 remuneration was R$3.26 per unit. The policy is consistent. It is also pro-cyclical with respect to rates: if Selic stays elevated and the acquisition debt seasons at current coupons, payout capacity in reais grows more slowly than RAP. Dual control is the governance residual. Cemig’s own political cycle flows into five board seats. Related-party operations with ISA Energia through AIE are visible and, so far, ordinary. Neither has produced a capital-allocation error of the kind the auction market is currently offering. Sitting out 50% discounts is the decision that should define this management team’s reputation more than the early energizations.
Industry Dynamics: The Grid Must Expand, the Return May Not
Demand for new transmission is not the constraint. EPE and ONS planning still points at multi-year capex to connect renewable surplus, to relieve curtailment out of the Northeast, and to create margin in São Paulo, where qualified demand for the 2029 connection window already exceeds available capacity by about two to one. Auction 4/2026 at the end of this month is only the next installment; a Brazil–Bolivia interconnection is already being discussed for 2027 and depends on treaty approval, which is a reminder that the political layer sits on top of the engineering layer. For an incumbent with corridors, substations and a functioning operations center, reinforcements on existing concessions are the highest-return expression of that demand. They do not require winning a room full of bidders. Climate exposure cuts the other way: a 16,000 km linear network is a physical asset in a country with rising extreme-weather frequency, and the variable portion is the mechanism that turns an outage into lost RAP. Availability at 99.9% has absorbed that so far. It is not guaranteed as the average age of the Category II book rises into its final years.
There is no credible technological entrant threatening the availability model. The disruptive force in this industry is the cost of capital and the willingness of State Grid, Axia, Alupar and infrastructure funds to accept RAP discounts above 50%. Storage, discussed by management as an adjacency alongside climate-resilience investment, is not a revenue line in the current RAP and should not be underwritten as one. HVDC retrofit capability is a real operating skill, demonstrated at Garabi, and it positions TAESA for reinforcement work on existing DC links rather than for a new product category. The regulatory threat is larger than the competitive threat. An ordinance before year-end will start to price the 2030 cohort. A rebidding design that socializes indemnification slowly, or a renewal that strips investment remuneration without a clean cash settlement, would convert a high-quality cash compounder into a shrinking RAP with a levered balance sheet. A clean indemnification near the R$3 billion uninflated figure, redeployed into operating assets at Energisa-like multiples rather than into auction discounts, would do the opposite. That fork, not the next quarter’s EBITDA, is what the TAEE4 cash flows after 2029 depend on.
The Scorecard
TAESA is a high-quality availability franchise: system-settled RAP, mid-80s EBITDA margins, availability above 99.9%, and a 2026–2027 cycle that is already contracted at R$4.9 billion, up 22% before any further construction, with the Energisa assets closed and only partially in the reported numbers as of mid-October 2026. Management has earned credibility on delivery speed and on refusing to bid greenfield returns down to half the regulatory ceiling. The preferred share, TAEE4, is an economic claim on that cash flow with dividend priority and no vote, under a Cemig–ISA control block that has so far allocated capital more carefully than the auction market.
The equity is nonetheless a duration problem with a growth offset, not a growth story with a duration footnote. About 40% of RAP matures by 2032, the first expiries arrive in 2030, and the chief executive has said revenue on those contracts falls under either renewal or rebidding. Indemnification of roughly R$3 billion before inflation is the bridge, and its form is undecided, with an ordinance expected before the end of 2026 and litigation left open. Incremental debt has already decoupled earnings from EBITDA. Relative to ISA Energia’s longer reinforced base and Alupar’s construction torque, TAESA’s advantage is near-term contracted RAP and operational discipline; its disadvantage is the size of the legacy book and a leverage ratio that entered this acquisition cycle at 4.1 times. The next decisive data point is regulatory, not operational.