Adani Green Energy: Building India's Renewable Colossus, Under Scrutiny

I. Introduction & Episode Roadmap

There is a stretch of land in the Rann of Kutch, in the Indian state of Gujarat, where almost nothing grows. Salt crusts the surface. The wind comes off the desert hard enough to make conversation difficult. Nobody farmed it, nobody built on it, and for most of independent India's history nobody wanted it. Today, a contiguous parcel there measuring roughly 538 square kilometres — about five times the area of Paris — is covered in bifacial solar modules, 5.2-megawatt wind turbines, and the largest single-location battery installation on the planet.¹

That site, Khavda, is the physical expression of a corporate thesis: that the binding constraint on India's energy transition is not technology or even capital, but land, transmission, and the willingness to build faster than anyone thought possible.

The company that made that bet is Adani Green Energy Limited. As of the quarter ended June 30, 2026, AGEL operated 20.1 gigawatts of renewable capacity, up 27% year over year, making it India's largest renewable generator by operating scale by a wide margin.¹ Management has committed to 50 GW by 2030.

The arithmetic of the last three years is what makes this story worth telling carefully. A company that was barely a decade old was hit in January 2023 by a short-seller report that helped erase more than $100 billion of combined market value across Adani Group listed entities within days.³ In November 2024, US prosecutors in Brooklyn indicted its founder, its executive director, and its then-CEO in connection with an alleged scheme to pay hundreds of millions of dollars in bribes to Indian officials.⁴ By mid-2026, India's securities regulator had found the central Hindenburg-era allegations "not established,"²⁵ the SEC had settled its civil case for $18 million without admissions,²⁶ and the Justice Department was asking a federal judge to dismiss the criminal indictment with prejudice, arguing it should never have been brought.³⁰

And through all of it, the plants kept getting built. Quarterly EBITDA from power supply reached ₹4,122 crore in the June 2026 quarter, at a 94% margin.¹

So the central tension is this. Is AGEL the best-executing infrastructure business in one of the world's fastest-growing power markets — an annuity machine wrapped in a growth story — or a heavily leveraged, related-party-dense platform whose governance discount is a permanent feature rather than a temporary overhang? The honest answer as of August 2026 is that both descriptions contain a lot of truth, and the evidence for each sits in different documents.

Here is where we're going: the origins and the founding strategic choice that still governs everything; the capital-markets machine and the SoftBank deal that bought scale; Khavda and how a contracted renewable independent power producer actually makes money; the storage pivot; the Hindenburg-to-DOJ arc, told as one continuous story rather than three headlines; management, ownership and capital allocation; the balance sheet; industry structure; the playbook; and a bull-versus-bear stress test with the handful of numbers actually worth tracking.

↑ Back to Top

II. Origins: From Trading House to Power Developer (2015–2018)

On June 18, 2018, a new ticker appeared on Indian exchanges. Adani Green Energy Limited debuted on the BSE at ₹29.40, locked in a 5% upper circuit on day one, and traded around ₹31 on the NSE.⁹ There was no IPO, no roadshow, no bankers' book. AGEL arrived by demerger — the renewable assets of Adani Enterprises spun out and handed directly to existing shareholders.

It is a telling entrance. AGEL did not begin life as a startup persuading strangers to fund a vision. It began as a carve-out of an existing conglomerate's balance sheet, and that origin explains a great deal about how it has behaved ever since.

The conglomerate's logic

The Adani Group traces to 1988, when Gautam Adani started a commodities trading business in Ahmedabad. Over three decades that trading operation became a ports, logistics, power generation, transmission, and gas distribution conglomerate — a set of businesses with one common characteristic. Every one of them is a physical asset with contracted or regulated cash flows, built at scale, financed with debt, and dependent on winning government processes.

Renewables fit that template with almost uncanny precision. India after the 2015 Paris Agreement had made large public commitments to non-fossil generation capacity while running a grid overwhelmingly dependent on coal. To close that gap, the government did not subsidise consumers directly. It built an auction machine. The Solar Energy Corporation of India, NTPC, and state distribution companies would tender for capacity; developers would bid a tariff; the lowest bidder would sign a power purchase agreement, typically for 25 years, at a fixed or lightly indexed price.

For a developer, that structure converts a weather-dependent commodity into something remarkable: a two-and-a-half-decade revenue stream with a sovereign-linked counterparty, signed before a single panel is installed.

The founding choice

AGEL was incorporated in 2016 and the group's renewable portfolio was consolidated into it ahead of the 2018 listing.⁴² The strategic decision made in those early years is the one that still defines the company in 2026: win auctions on price, contract essentially all output on long-term PPAs with government-backed offtakers, and finance construction against those contracted cash flows.

That is a deliberate refusal of the merchant-power model, where a generator sells into a spot market and captures — or loses — whatever the market pays. Merchant exposure offers upside when power is scarce. It also makes debt expensive, because lenders cannot underwrite a cash flow they cannot forecast.

AGEL chose the opposite. Give up the upside, lock the price, and use the contract as collateral. It is closer to how a toll road gets financed than how a technology company grows. And it means the company's returns are determined almost entirely at the moment of bidding — get the tariff and the cost of capital right, and twenty-five years of cash flow follow mechanically. Get them wrong, and there is no operating fix.

That single decision set up everything that came next, because a model financed against contracted cash flow scales only as fast as you can raise capital against it. The next chapter of AGEL's story is therefore not about engineering. It is about money.

↑ Back to Top

III. The Scaling Machine: Capital, Acquisitions & the SoftBank Bet (2018–2021)

In October 2019, three wholly-owned AGEL subsidiaries holding 570 MW of operating solar assets went to international investors with an unusual proposition: a 20-year, US-dollar-denominated green bond, amortising, secured against long-term PPAs with a weighted-average remaining life of roughly 24 years. It priced at a yield of 4.625% and carried investment-grade ratings of Baa3/BBB–/BBB–.¹¹

It raised $362.5 million — a modest sum against what would come later. Its significance was categorical rather than numerical. It was the first investment-grade US-dollar bond by an Indian renewable energy company, and the first 20-year green bond out of India.¹¹

What the bond actually proved

Indian infrastructure has historically been financed by Indian banks, at Indian rates, on Indian tenors — typically shorter than the asset life, which forces repeated refinancing. A 20-year amortising instrument sold to global fixed-income buyers matched the debt maturity to the contract maturity for the first time.

There is a deeper point about who was underwriting what. Those bondholders were not underwriting Gautam Adani. They were underwriting SECI's obligation to buy power for 24 more years, plus the physics of a solar plant. The financing structure quietly separated the credit of the asset from the credit of the sponsor — a distinction that would matter enormously in 2023 and 2024, when sponsor credibility came under attack while the underlying assets kept generating.

The record auction

In June 2020, SECI awarded AGEL what was then described as the world's largest solar development order: 8,000 MW of solar capacity, coupled with a commitment to establish 2 GW of domestic solar cell and module manufacturing.¹² AGEL completed power-offtake tie-ups for the entire 8,000 MW by late 2023.¹³

The manufacturing-linked structure is worth pausing on, because it reveals what the Indian state actually wanted. The government was not simply buying electricity; it was buying industrial capacity, using a guaranteed 25-year offtake as the inducement. Only a bidder able to commit simultaneously to a multi-gigawatt build and a manufacturing plant could compete. That is a structural filter favouring conglomerates over pure-play developers — and AGEL, sitting inside a group with construction, ports, and logistics arms, was built to pass through it.

The French validation

In early 2020, Total agreed to invest $510 million for 50% of a 2,148 MW operating solar portfolio at an enterprise value of about ₹17,385 crore, expanding the joint venture with a further ₹1,632 crore of solar assets shortly after.⁴⁴ The relationship deepened in January 2021, when Total acquired a 20% interest in AGEL itself.¹⁰ TotalEnergies has described the stake as a minority interest of 19.75%.³⁹

This was the first genuine outside-standard test of AGEL's numbers. A European supermajor with its own engineers, its own auditors, and its own board-level scrutiny examined the asset base and decided to write a very large cheque. That is not proof of anything about group-level governance — and TotalEnergies would later have to publicly address its Adani exposure after the 2023 short report. But it was a meaningful signal that the operating assets were real and the contracts were what they appeared to be.

Buying five gigawatts

On May 19, 2021, AGEL signed share purchase agreements to acquire 100% of SB Energy India from SoftBank Group (80%) and the Bharti Group (20%) for a fully completed enterprise value of approximately $3.5 billion.¹⁴ It closed in October 2021.¹⁵ At the time it was the largest renewable energy M&A transaction in India's history.

The portfolio contained 4,954 MW: roughly 84% solar, 9% wind-solar hybrid, and 7% wind — with only 1,700 MW actually operating, 2,554 MW under construction, and about 700 MW near construction.¹⁴

That composition is the key to judging the price. A headline of $3.5 billion for "5 GW" implies roughly $0.70 per watt, which would have looked inexpensive. But most of the megawatts were not producing anything. AGEL was buying a development pipeline, a land bank, an operating platform, and a set of signed PPAs — and then had to spend its own capital to finish the construction. The enterprise value included assumed project debt, so the equity cheque was smaller than the headline; AGEL described the transaction as lifting operational capacity by 46% on completion.¹⁵

What did AGEL actually get? Speed. Land and grid connectivity in India are acquired one landowner and one regulatory approval at a time, and the calendar cannot be compressed with money alone. Buying SB Energy bought years. It also bought signed tariffs from an earlier auction vintage, when pricing was more generous than what the market would offer later.

What it did not buy was any obvious operating edge. SB Energy's assets were scattered across states, which is the opposite of the concentrated-site model AGEL was about to pursue at Khavda. Judged honestly, this looks less like a bargain than like a rational purchase of time by a company whose entire strategy depended on getting to scale before competitors did — funded, notably, with a balance sheet that was already levered.

By 2022 the market had decided it liked the story a great deal, and AGEL's equity value had multiplied many times over from its unremarkable 2018 debut. That is the point at which the growth-story phase peaked. What came next was a project so large it changes how you have to think about the business.

↑ Back to Top

IV. Khavda and the Core Business: How a Renewable IPP Actually Wins

In December 2022, construction crews arrived at a place with no road, no water, no grid connection, and no permanent human population. Twelve months later, in December 2023, the first electrons flowed. By the June 2026 quarter, cumulative installed capacity at the site stood at 10.3 GW — 9,520 MW belonging to AGEL and 742 MW built for other Adani group companies — against a target of 30 GW by 2029.¹

To put that pace in perspective: AGEL added 5,051 MW of renewable capacity across FY26, which the company said was the highest annual addition by any developer globally outside China.²

The business model, in plain terms

Strip away the vocabulary and a contracted renewable IPP works like this.

You win an auction by promising to sell electricity at, say, ₹2.50 per unit for 25 years. You then borrow most of the money to build the plant, and the lender sizes that loan against the contracted revenue, not against your equity story. Once the plant is running, sunlight and wind are free. Your costs are interest, depreciation, and a modest amount of maintenance and land lease. Everything else is margin.

That is why AGEL's EBITDA margin on power supply reached 94% in the June 2026 quarter.¹ This number confuses people who compare it to industrial companies. It should be compared to a toll road or a cell tower: enormous upfront capital, then decades of collection with almost no variable cost. The profit was effectively determined on the day the PPA was signed.

The corollary — and it is a hard one — is that operating excellence has limited ability to rescue a bad bid. You can squeeze a percentage point of capacity utilisation through better cleaning and better forecasting. You cannot re-price a 25-year contract.

Why Khavda is different from a big solar farm

Khavda's advantage is not that it is large. It is that it is contiguous.

A developer building 10 GW across forty sites in eight states negotiates forty land assemblies, forty grid connections, forty sets of local approvals, and runs forty separate construction supply chains. AGEL builds on one parcel with one shared transmission corridor, one logistics base, one workforce camp, and one set of permits. Every incremental gigawatt at Khavda is cheaper and faster to install than the one before, because the hard infrastructure is already sunk.

This is a genuine scale economy in Hamilton Helmer's sense — declining unit cost with volume, in a way a competitor cannot match without building an equivalent site. And it rests on something close to a cornered resource: the site is a vast tract of non-arable land near the Pakistan border with exceptional solar irradiance and wind speeds, assembled early and at low cost. India does not have many more of these.

The honest caveat is that a head start is not the same as a moat. Nothing prevents a well-capitalised rival from assembling land in Rajasthan or Andhra Pradesh. What Khavda buys is a lead measured in years and a cost position that competitors must spend heavily to approach. That is valuable. It is not permanent.

Solar by day, wind by night

The hybrid model deserves a plain-English explanation because it drives the economics of everything AGEL is building now.

A solar plant produces only when the sun shines — a bell curve peaking at midday, zero after sunset. A wind farm at Khavda produces disproportionately at night and during the monsoon. Both need an expensive transmission line to carry power to demand centres. Built separately, each line sits idle much of the time.

Put both on the same site and the same line, and the line carries power far more hours per day. The transmission asset — one of the largest fixed costs in the chain — gets amortised across many more units of energy. That is why AGEL's recent additions have skewed toward hybrid: of the 4,327 MW added in the twelve months to June 2026, 3,051 MW was solar, 684 MW wind, and 592 MW solar-wind hybrid, with the wind and hybrid capacity concentrated at Khavda.¹

The commercial reason matters more than the engineering one. Indian buyers increasingly want firm, dispatchable, round-the-clock renewable power rather than intermittent output they must back up with coal. Firm power commands a higher tariff. Hybridisation is the cheapest first step toward it.

Who else is in the ring

AGEL's lead on operating scale is currently substantial, but the competitive set is strengthening from an unhelpful direction.

NTPC Green Energy crossed 10 GW of renewable capacity in FY26 and reported consolidated profit after tax of ₹521 crore.⁴⁰ It is the renewables arm of India's largest state-owned generator, and its structural advantage is not execution — it is the cost of money. A public-sector undertaking with sovereign association borrows more cheaply than any private developer. In a business where the winning bid is essentially a function of capital cost, that is the single most dangerous form of competition AGEL faces.

JSW Energy reported total installed capacity of 13,295 MW with renewables at roughly 57% of the mix, and has publicly targeted a much larger green portfolio by 2030.⁴⁰ Tata Power operates over 7,000 MW of renewable capacity and is vertically integrated into module and cell manufacturing.⁴⁰ ReNew brings deep wind expertise and a US-listed capital base.

The pattern is clear. AGEL wins on speed and scale. Competitors are converging on cost of capital, manufacturing integration, and — in the PSU case — on a balance sheet that does not need to earn a private-equity return.

How AGEL wins, and how it loses

It wins through three mechanisms that are all observable rather than rhetorical. First, land and transmission optionality at Khavda that shortens time-to-power. Second, in-house engineering and operations, including a centralised network operations centre that monitors the fleet. Third, demonstrated access to long-tenor global debt — including a $1.06 billion refinancing completed in March 2025, notable because it was the first significant dollar raise after the US indictment and showed that the credit markets would still transact.³⁴

It loses when auction tariffs compress below what its cost of capital can support. This is not hypothetical. As PSU bidders scale, the marginal auction is increasingly priced off government-linked funding costs. AGEL's embedded fleet — contracts signed in earlier, richer vintages — remains highly profitable. The question investors should hold onto is whether new capacity earns the same spread.

There is also a live operational drag that management disclosed candidly. On the Q1 FY27 call, the company said curtailment — grid operators instructing plants to reduce output because transmission cannot absorb it — was reducing EBITDA by 5% to 7%, with relief expected as roughly 7 GW of evacuation infrastructure comes online by end-2026 and the remainder within one to two quarters after.¹⁹ That is a real, quantified haircut on reported earnings, and it is the direct consequence of building generation faster than India can build wires.

Which raises the obvious question: if the grid cannot always take the power when it is produced, what if you could store it?

↑ Back to Top

V. Beyond Generation: Storage, Hybrid Firming & the Next Leg

On November 10, 2025, the Adani Group announced its entry into battery energy storage with a project sized at 1,126 MW / 3,530 MWh at Khavda, with commissioning targeted for March 2026 and a stated ambition of 15 GWh by March 2027 and 50 GWh over five years.²² Gautam Adani framed it as infrastructure rather than experiment: "Energy storage is the cornerstone of a renewable-powered future."²²

Chairman quotes are not evidence. The build schedule is more informative. AGEL commissioned an initial 1,376 MWh of BESS at Khavda during FY26,² then added 1,972 MWh in the June 2026 quarter alone, taking total installed battery storage to 3,551 MWh.¹ Management has guided to more than 10,000 MWh of additional BESS capacity in FY27 and 50 GWh by 2030.¹

For context on why this is happening now: batteries have gone from a research topic to a procurement line item in roughly five years, on the back of global cell manufacturing scale. India's grid operators and distribution companies have simultaneously shifted from tendering raw renewable capacity to tendering firm capacity — power that can be delivered at a specified hour.

The economics management described

On the Q1 FY27 call, management gave unusually concrete numbers for a business this new: batteries store power at around ₹2.5 per unit and discharge into evening peak demand at ₹4 to ₹5 per unit, with expected EBITDA of roughly ₹25 lakh to ₹30 lakh per MWh of installed capacity.¹⁸¹⁹

That spread is the whole thesis of storage in one line. Solar makes electricity worthless at noon and scarce at 8pm; a battery is a machine for moving energy across that gap. If those unit economics hold at 10 GWh of scale, storage becomes a materially higher-return use of the Khavda footprint than selling raw solar.

The "if" is doing real work. This is a young asset class in India with limited operating history, and management itself declined to specify quarterly commissioning phasing on the call, citing "deep commissioning and stabilization" periods.¹⁹ Analysts on the same call pressed on battery fire risk; management attributed incidents in the industry to inverter and power-conversion-system failures rather than cell defects.¹⁹ That is a reasonable answer, but it is an answer, not a track record.

Pumped hydro: the hundred-year battery

In February 2025, AGEL won 1,250 MW of energy storage capacity from Uttar Pradesh Power Corporation for the Panaura Pumped Storage Project in Sonbhadra district, under a PPA with a minimum 40-year term, with completion expected within six years.²³²⁴ The company has additional pumped storage projects under development at Chitravathi and Gandikota in Andhra Pradesh and Tarali in Maharashtra, and has stated a target of over 5 GW of pumped hydro by 2030.²³

Pumped hydro is the oldest and least glamorous storage technology: use cheap daytime solar to pump water uphill, release it through turbines at night. As management put it, the plants "use electrons generated from solar in the day to pump water and meet peak time energy demand at night."²³ The assets last for generations, which is why the contract runs forty years rather than twenty-five. They also take the better part of a decade to build and involve civil works, water rights, and environmental clearance — categories of risk quite different from bolting modules onto steel.

How to size this honestly

Storage is the most credible new source of value AGEL has added since Khavda itself, and it fits the existing asset base with unusual precision — same land, same transmission, higher-value product.

But it is not yet a proven profit contributor. Storage revenue is not broken out at material scale in reported segment data, the pumped hydro projects are years from commissioning, and the entire BESS fleet has less than a year of operating history. Treat contracted and commissioned GWh as a forward-looking indicator to monitor, not as earnings.

There is also a structural detail from the Q1 FY27 call that belongs here rather than in a footnote, because it changes what storage actually is for AGEL. All 3.5 GWh of installed battery capacity has been contracted to Adani Energy Solutions Limited — a fellow Adani Group listed company — on 15-year fixed-price agreements, with solar and wind merchant capacity similarly moved onto 25-year contracts with AESL.¹⁸¹⁹ Management explained the reclassification as de-risking: converting volatile merchant exposure into contracted revenue, at prices management said were benchmarked to market and approved by the board following SECI PPA guidelines.¹⁹

That is a defensible commercial rationale. It is also a very large related-party arrangement that converts an arm's-length market price into an intra-group negotiated price. Which brings us to the part of this story that no amount of operating excellence has been able to settle.

↑ Back to Top

VI. Crisis, Scrutiny & Vindication: Hindenburg to the DOJ Dismissal (2023–2026)

On January 24, 2023, a small New York short-selling research firm published a report accusing the Adani Group of what it called decades-long "brazen stock manipulation and accounting fraud," alleging improper use of offshore entities and undisclosed related-party dealings. It said the work followed a two-year investigation.

What happened next was one of the fastest destructions of paper wealth in corporate history. By February 2, 2023, losses across Gautam Adani's main businesses had reached roughly $107 billion since publication.³ A ₹20,000 crore follow-on public offer at Adani Enterprises — which had been fully subscribed — was withdrawn. AGEL's dollar bonds traded down sharply.

The group's response was immediate and combative. In a media statement issued the day after publication, it characterised the report as a calculated attack timed to damage the FPO.³³ Alongside the rhetoric came something more consequential to creditors: a campaign of debt prepayment to release pledged promoter shares, aimed at removing the mechanism by which a falling share price could trigger forced selling.

That distinction — between rebutting the argument and removing the financial vulnerability the argument exploited — is the most instructive thing about the group's crisis management. The pledge unwinding addressed a real fragility.

Partial vindication, first instalment

In January 2024, India's Supreme Court declined to transfer the investigation to a special investigation team or the CBI, accepting the securities regulator's position and directing SEBI to complete its remaining inquiries. GQG Partners, a US-based asset manager led by a well-known emerging-markets investor, had by then taken and increased positions across Adani entities — the most-cited external signal that institutional capital was willing to underwrite the group again.

None of that was a finding of innocence. It was a finding that, at that stage, fraud had not been established.

November 20, 2024

Then Brooklyn. On November 20, 2024, the US Attorney's Office for the Eastern District of New York unsealed a five-count indictment charging Gautam Adani, his nephew and AGEL executive director Sagar Adani, then-AGEL chief executive Vneet Jaain, and five others in connection with an alleged scheme to pay hundreds of millions of dollars in bribes to Indian government officials to secure solar power supply contracts.⁴ Prosecutors alleged the contracts were projected to generate more than $2 billion in after-tax profit over roughly twenty years, and that the defendants concealed the scheme from US investors while raising capital in American markets.²⁹

The SEC filed parallel civil complaints the same day, alleging that in September 2021 — while the alleged scheme was ongoing — the defendants falsely touted AGEL's compliance with anti-bribery principles in connection with a $750 million bond offering.²⁸²⁶

Note who was charged. This was not a group-level allegation that happened to touch AGEL. The named individuals included AGEL's executive director and its chief executive, and the alleged conduct concerned AGEL's core commercial activity: winning solar PPAs. The company publicly denied that its directors had been charged with FCPA violations specifically and said it would appoint independent law firms to review the indictment.⁴⁵⁴⁶ A planned dollar bond raise was shelved. The stock and bonds sold off again.

The operational consequence was measurable and then, notably, short-lived: by March 2025, AGEL had completed a $1.06 billion refinancing — its first substantial dollar raise since the indictment — which told the market that international lenders would still price the assets even while the sponsor was under criminal charge.³⁴

September 2025: SEBI closes the loop

On September 18, 2025, SEBI issued final orders finding that the transactions flagged in the 2023 short report did not violate securities law at the time. "Having considered the matter holistically, I find that the allegations... are not established," wrote whole-time member Kamlesh C. Varshney.²⁵⁷ The regulator found no evidence that the conglomerate had routed funds into its listed companies through related parties.²⁵

Balance requires two qualifications. First, these orders addressed specific allegations; the Supreme Court's earlier direction had covered a docket of twenty-four investigations, of which twenty-two had been completed, with the remaining ones to be finished expeditiously.²⁵ Second, and more substantively, commentators noted that the orders turned significantly on what Indian related-party disclosure rules required at the time of the transactions — a legal-standard question rather than a factual finding that nothing of concern occurred.

2026: the case the government abandoned

In May 2026, two things happened in quick succession.

On May 14, Gautam Adani and Sagar Adani filed proposed consent judgments with the SEC, agreeing to civil monetary penalties of $6 million and $12 million respectively — $18 million in total — without admitting or denying the allegations.²⁶²⁷ For men of their means, the amounts are close to symbolic. The structure is what matters: a settlement without admission resolves exposure without producing a factual record.

Around the same time, the Justice Department moved to dismiss the criminal indictment with prejudice.⁵ The stated grounds were jurisdictional weakness, the absence of demonstrable US investor losses, parallel Indian proceedings, and poor prospects at trial.⁵ Defence counsel Robert J. Giuffra Jr. of Sullivan & Cromwell described a compressed and intensive campaign — hundreds of pages of filings assembled over roughly ten weeks in early 2026.⁵

Judge Nicholas Garaufis of the Eastern District of New York did not simply sign it. On June 26, 2026, he declined to dismiss immediately, writing that "the Government's terse, bland and conclusory statement affords the court neither a sufficient basis to reach any conclusion," and ordering prosecutors to file, by July 13, each reason for seeking dismissal with sufficient factual support.⁶³²

The DOJ complied in early July with a brief arguing the indictment was legally flawed and should never have been filed, and urging permanent dismissal with prejudice.³⁰ Principal Associate Attorney General Trent McCotter told the court the government had decided to abandon the charges before any investment offer was raised.

That last point was necessary because of the elephant in the courtroom. The Adani Group had announced a proposed $10 billion US investment on November 13, 2024 — one week before the indictment was unsealed.³¹ Judge Garaufis ordered Gautam Adani to answer, in writing and under oath, whether he was aware of "anything promised, offered, sought, received, agreed to, or accepted, by anyone in connection with the dismissal of the indictment."[^32] Adani filed a sworn affidavit on July 15, 2026, denying any such arrangement and stating that the proposed US investment played no role in the DOJ's decision.³¹

As of August 10, 2026, no public record confirms that Judge Garaufis has granted the motion. The dismissal has been sought, defended, and interrogated. It has not, on the available evidence, been finally ordered. Several outlets reported the case as closed in July 2026;⁵ the court docket record available publicly indicates the motion remained pending judicial approval after the affidavit was filed.³¹ Investors should treat this as an open matter, not a settled one.

Reading the arc

Neither of the two available narratives survives contact with the documents.

The "total vindication" story fails because the criminal case was abandoned on prosecutorial discretion and jurisdictional grounds, not resolved on the facts; because a judge found the government's initial explanation inadequate and demanded a sworn denial of any quid pro quo; and because the SEC matter was settled with payment and no admission. Nothing in that sequence establishes that the alleged conduct did not occur.

The "cover-up" story fails too. SEBI conducted a multi-year investigation and issued reasoned orders finding the flagged transactions did not breach the rules in force. The DOJ's stated jurisdictional concerns — conduct predominantly in India, involving Indian officials, with contested US investor harm — are the kind of arguments that defeat foreign-bribery prosecutions on the merits with some regularity.

The defensible reading is narrower and less satisfying: allegations of grave conduct were made by two US agencies and a short seller; none has been proven in a contested proceeding; one regulator has cleared the specific transactions it examined under the standards then applicable; and the criminal case is being withdrawn by the prosecutor rather than defeated at trial.

What a skeptic would look at today

A short seller opening a file on AGEL in 2026 would probably not re-litigate 2023. They would start with what is disclosed now.

They would start with the AESL contracts — an entire storage fleet and reclassified merchant capacity sold to an affiliate at prices set inside the group, with the pricing benchmark asserted by management rather than validated by an external market.¹⁹ They would map intercompany loans, guarantees, and equipment sourcing across group entities, including module supply from affiliated manufacturing. They would count independent directors relative to promoter influence. And they would ask the single sharpest question available: whether disclosure practice around US investor risk actually changed after November 2024, or whether the company simply stopped raising money in the venue where the standard applied.

None of that is an accusation. It is the diligence agenda the last three years created. Which makes it worth examining who is actually running the company.

↑ Back to Top

VII. Current Management, Ownership & Capital Allocation

Listen to an AGEL earnings call and you hear four voices: Vijil Jain from investor relations, chief financial officer Saurabh Shah, Rajat Seksaria who runs the battery storage business, and chief executive Ashish Khanna.¹⁹ Only one of them carries the Adani surname on the board — and he is not on the call as the principal presenter.

The professional layer

Vneet S. Jaain is Managing Director, re-appointed with effect from July 10, 2025, and is also chief executive of Adani New Industries.³⁵ He is a career power-sector executive rather than a family member, and he ran AGEL through the entire scaling phase from 2018 onward.

He is also, and this cannot be treated as biography, one of the individuals named in the November 2024 US indictment.⁴ The charges against him related to securities fraud conspiracy and wire fraud conspiracy rather than FCPA violations. He remained in an executive role at the company throughout. Whatever one concludes about the merits, the board's decision to retain named defendants in senior positions while charges were pending is itself a governance data point that shareholders were entitled to weigh.

Ashish Khanna was appointed chief executive officer effective April 1, 2025, succeeding Amit Singh, having previously led the group's international energy business.³⁶ The functional effect was to separate day-to-day operating leadership from the managing director role during the period the indictment was live.

On the evidence of delivery, this team has done what it said it would. AGEL committed to adding 5 GW in FY26 and reported 5,051 MW — 3,409 MW solar, 686 MW wind, 956 MW wind-solar hybrid.² It has guided to another 5 GW plus more than 10 GWh of storage in FY27 and said in July 2026 it was on track.¹ Capacity guidance is the one promise this management has consistently kept, and in an industry where projects routinely slip by years, that is not a small thing.

The family layer

Sagar Adani, Gautam Adani's nephew, is Executive Director and the operating face of the family's next generation. He is the company's most quoted executive in results releases — "This milestone strengthens our progress towards our target of 50 GW by 2030" on the FY26 capacity announcement,² and detailed contribution statistics on the FY25 results.⁸

He is also the co-defendant who agreed to pay the larger of the two SEC penalties: $12 million against Gautam Adani's $6 million.²⁶²⁷ The SEC's complaint identified him as AGEL's executive director and alleged he was directly involved in the conduct at issue.²⁸ He remains in his role. For investors, the relevant question is not moral but structural: the individual most closely associated with the company's regulatory exposure is also the individual with the longest expected tenure.

Ownership

The promoter group held 1,028,396,636 shares, or 62.43% of capital, and filed a declaration under Regulation 31(4) of SEBI's takeover regulations confirming no new encumbrances were created during the year ended March 31, 2026.³⁷ Within that, Adani Trading Services LLP holds 28.80% and the S.B. Adani Family Trust 19.96%.³⁷

The absence of new pledges is a meaningful contrast with the pre-2023 posture, when pledged promoter shares were the transmission mechanism between a falling stock price and a funding crisis. It is worth stating the limit of the signal: this is a declaration about a completed financial year, not a commitment about future ones, and encumbrance levels can change quickly. It should be re-checked against each shareholding pattern filing.

The TotalEnergies question

TotalEnergies acquired its stake in January 2021 and has described holding 19.75%.³⁹ By September 30, 2025 the reported holding was 18.99%, and in November 2025 reports indicated preparation to sell up to a further 6% in a transaction estimated near ₹10,000–10,200 crore, with the residual stake valued at multiples of the original investment.³⁸⁴³

How should a long-term investor read that? Not as a verdict. A supermajor trimming a position after a very large mark-to-market gain, while retaining the large majority of it and maintaining the operating relationship, is portfolio management. If TotalEnergies were exiting entirely, that would be a different signal. The nuance worth holding is that AGEL's most sophisticated outside strategic shareholder has become a net seller at current valuations — while GQG, its most visible outside financial shareholder, moved the other way after 2023. Reasonable professionals disagree about the price.

Capital allocation

AGEL pays no meaningful dividend. Every rupee of operating cash flow, plus a great deal of borrowed money, goes back into greenfield capacity and now storage. FY27 capex guidance is ₹42,000 crore, against Q1 FY27 capex of ₹8,800 crore, up 41% year over year.¹⁸¹⁹

That is internally consistent with the model — a contracted-infrastructure compounder should reinvest while returns on new capital exceed cost of capital. It also means shareholders have no cash return and are wholly dependent on the build continuing to work.

Two credibility observations from the most recent call. First, management framed the merchant-to-AESL reclassification as de-risking, and did so in response to a direct analyst question rather than proactively — analysts had to ask why the classification changed.¹⁹ Second, and more notable: management provided no net debt to EBITDA guidance on the Q1 FY27 call, offering instead a run-rate EBITDA path from roughly ₹17,000 crore currently to about ₹21,000 crore in FY27.¹⁹ Deleveraging trajectories circulating in the market — a path from roughly 7.4x in FY25 toward the mid-5x range by FY28 — come from sell-side modelling, not from a company commitment reiterated on the record.²¹ The company's own FY26 business and credit compendium remains the primary document against which any leverage claim should be checked.⁴¹

That distinction matters enormously, and it takes us to the balance sheet.

↑ Back to Top

VIII. The Financial Architecture: PPAs, Margins & Balance Sheet Risk

Here is the single most important thing to understand about AGEL's financial statements: the income statement and the balance sheet tell two completely different stories, and both are true.

The income statement

Power supply revenue in the June 2026 quarter was ₹4,280 crore, up 29% year over year, generating EBITDA of ₹4,122 crore at a 94% margin, with cash profit of ₹2,225 crore, up 28%.¹ Energy sales reached 13,657 million units, up 30%.¹ A year earlier, the comparable quarter delivered ₹3,312 crore of revenue and ₹3,108 crore of EBITDA at 92.8%.¹⁶

FY25 was the year AGEL first crossed $1 billion of annual EBITDA — a threshold that put it in a different conversation among global renewable platforms.⁸ Across the nine months to December 2024, EBITDA margin ran at 91.7%, with solar capacity utilisation of 32.4% in the March 2025 quarter and PPA generation at 107% of annual commitment.⁸

The first half of FY26 showed the same pattern at larger scale: revenue of ₹6,088 crore and EBITDA of ₹5,651 crore at a 91.8% margin, on energy sales up 39% year over year.¹⁷

Read the PPA-generation figure carefully, because it is the one that validates the model. Generating 107% of contracted commitment means the plants are delivering more than the PPAs require — the asset base is performing above the assumptions underwritten in the contracts.

The balance sheet

Now the other side. Net debt reportedly rose to approximately ₹91,252 crore by March 2026 from about ₹64,462 crore a year earlier — an increase of some ₹26,790 crore in a single year, against reported equity of roughly ₹19,965 crore.²¹ Reported debt-to-EBITDA measured on trailing figures was cited at 10.92x as of Q4 FY26, with EBIT-to-interest coverage averaging around 1.11x and long-term debt having risen 66% in the prior year.²⁰

Different leverage figures circulate because different people measure different things: net versus gross debt, run-rate versus trailing EBITDA, and whether under-construction assets contributing no earnings yet are included in the denominator. Management's ₹17,000 crore run-rate EBITDA against roughly ₹91,000 crore of net debt implies something near 5x; trailing reported figures imply roughly double that.¹⁹²¹

Both calculations are defensible, and the gap between them is the investment question. If you believe the under-construction pipeline converts to earnings on schedule, current leverage is a timing artefact. If commissioning slips, or curtailment persists, or tariffs on new capacity disappoint, the trailing number is the honest one.

An interest coverage ratio near 1.1x deserves to be said plainly: operating profit is covering interest with very little cushion.²⁰ That is characteristic of project-financed infrastructure in a heavy build phase, where each new asset adds interest immediately and earnings with a lag. It is also exactly the condition under which an unexpected shock — a rate cycle, a payment delay, a construction overrun — becomes a financing problem rather than an earnings problem.

Why leverage is a feature, not an accident

A 25-year contracted cash flow is precisely what lenders like to lend against. Using less debt would depress return on equity without materially improving the safety of an asset whose revenue is already contracted. The model is designed to be levered.

The genuine risk is therefore not solvency in a static sense. It is refinancing. Project debt must be rolled or amortised across decades and multiple rate cycles. AGEL's mitigation — visible in the original 2019 structure and reinforced since — is long-tenor amortising debt matched to PPA life, which reduces the number of times it must return to the market. The March 2025 dollar refinancing showed access survived even the indictment.³⁴ That is real evidence, not assertion.

The counterparty question inside the "de-risked" contract

There is a caveat buried inside the phrase "sovereign-backed offtake." Indian state distribution companies have a long, well-documented history of paying generators late. A PPA with a state discom guarantees the price and the obligation; it does not guarantee the timing of cash. Receivable days are a genuine risk in this sector, and central intermediaries such as SECI mitigate but do not eliminate it.

Cost of capital versus peers

Set AGEL against its two most relevant comparators and the picture is sharp. NTPC Green borrows against sovereign association and therefore needs a lower tariff to hit its return threshold — a permanent structural edge in auctions. ReNew, listed in the US, accesses a different investor base with different disclosure requirements and a different tolerance for governance complexity.

AGEL sits in between: genuine access to global capital markets, demonstrated repeatedly, but at a spread that reflects both group-level governance history and the leverage of the model. Whether that access is a competitive advantage or simply a higher risk appetite depends on where the next auction clears. Which is really a question about industry structure.

↑ Back to Top

IX. Industry Structure: Porter's Five Forces & Durable Advantage

Picture an Indian renewable auction. A dozen bidders, an online reverse-auction clock, and a single variable that decides the winner: the tariff in rupees per unit. No brand, no product differentiation, no customer relationship. Just a number.

That image tells you most of what you need to know about the industry's structure. Under Porter's framework, this is a business with weak differentiation and powerful buyers — which makes the question of where advantage can possibly live unusually interesting.

Supplier power: moderate, and complicated. Solar modules are a global commodity subject to violent price cycles and shifting trade policy; India's domestic content requirements have raised prices while building local capacity. AGEL is partly hedged by the group's affiliated solar manufacturing, which secures supply and captures manufacturing margin inside the family. That is efficient. It is also another related-party channel where transfer pricing is set internally, and it should be read alongside the AESL arrangements rather than separately.

Buyer power: high, and rising. SECI, NTPC, and state discoms design the tenders, set the terms, and can time issuance to market conditions. They face many hungry bidders and are structurally motivated to push tariffs down. Critically, buyer power operates asymmetrically across time: it compresses the economics of new capacity while leaving the installed base's signed contracts untouched. AGEL's existing fleet is insulated. Its growth is not.

Threat of entry: high on paper, moderate in practice. There is no intellectual property here. Anyone with capital can buy panels. What is genuinely scarce is contiguous land with irradiance and grid access, plus the execution capability to install gigawatts per year. The realistic threat is not new entrants but incumbents with cheaper capital — NTPC Green, SJVN, and other public-sector platforms with a lower hurdle rate.

Substitutes: policy, not technology. Coal remains India's baseload backbone. The substitution risk is not a rival clean technology displacing solar; it is a slowdown in renewable procurement, a change in auction design, or a shift in political priorities. This is a demand-side policy risk dressed as a substitution risk.

Rivalry: intense at the margin. Because bids are price-only, rivalry expresses itself entirely through tariff compression rather than any dimension where AGEL's operational strengths count.

Mapping to Helmer's 7 Powers

Against that structure, three of Helmer's powers plausibly apply to AGEL, and several clearly do not.

Scale economies — the strongest claim. Khavda's shared transmission, logistics, and workforce base mean AGEL's marginal cost per gigawatt at that site is genuinely below what a fragmented builder achieves. This is verifiable in the build rate: 5,051 MW in FY26.²

Cornered resource — the second-strongest. The Khavda land assembly, with grid corridors under construction, is not readily reproducible. India has other suitable geographies; it does not have many with this combination of scale, resource quality, and an already-committed evacuation corridor.

Process power — real but temporary. Building at five gigawatts a year is a learned organisational capability. Learned capabilities can be learned by others, particularly by well-capitalised rivals recruiting from the same labour pool.

What AGEL does not have: branding (electrons are electrons), network economies, counter-positioning (competitors are copying the model, not being blocked from it), and switching costs — though there is a mild incumbency effect in tender qualification, where demonstrated delivery history matters to procurers.

The honest summary is that AGEL's advantage is a cost-and-time lead built on a genuinely scarce physical asset. That is a real edge. It is not an edge that widens automatically, and it does nothing to protect against a competitor whose cost of capital is structurally lower. Recognising that is the beginning of the general lesson.

↑ Back to Top

X. Playbook: Business & Investing Lessons

The most transferable idea in this story is the one that is easiest to skip past: a contract can change what kind of business you are in.

Electricity is the purest commodity imaginable. An electron from a solar panel is indistinguishable from an electron from a coal plant, and a merchant generator selling into a spot market has no pricing power whatsoever. Wrap that same electron in a 25-year fixed-price agreement with a creditworthy counterparty and you have manufactured a bond-like instrument out of a commodity. The margin stops being a function of competition and becomes a function of contract structure and financing cost.

Every capital-intensive business should ask the corresponding question: what is the longest-duration, highest-certainty contract my customer will realistically sign, and how does signing it change what my lenders will do?

Second: in emerging-market infrastructure, land and permits are the scarce input, not capital. Global capital is abundant and mobile; contiguous land with grid access near a demand centre is neither. Khavda's value is not that it is impressive. It is that it was assembled early, cheaply, and at a scale nobody else attempted. Any investor evaluating an infrastructure builder should look hard at what it controls that cannot be bought quickly.

Third: conglomerate backing cuts both ways, and the two edges have different time signatures. The benefits show up fast — capital access, land relationships, in-house EPC, affiliated manufacturing, and a group balance sheet that can absorb early losses. The costs show up slowly and all at once. Reputational contagion from group-level controversy hit AGEL's cost of capital even though the underlying assets kept performing. Related-party density that is operationally convenient becomes an analytical liability the moment anyone starts asking questions about transfer prices.

Fourth: speed is a real strategy, but only when it is paid for. AGEL's five-gigawatt-a-year build rate is a genuine competitive weapon in a market where the auction pipeline rewards demonstrated delivery. It is also the reason net debt rose by roughly ₹27,000 crore in a single year.²¹ Speed and financing discipline pull in opposite directions, and the 2023–2024 stress test showed exactly how a leveraged growth story behaves when external confidence evaporates.

Fifth, and least appreciated: crisis management and deleveraging are a separate skill from industrial execution. AGEL's operating team never stopped delivering through the crisis years. But the pledge unwinding, the refinancing that reopened dollar markets after the indictment, the regulatory defence, and the eventual settlements were a different discipline entirely, executed by a different set of people. Investors evaluating any founder-led, high-leverage growth company should ask whether the organisation has both capabilities, because it will eventually need both.

The final lesson is the one this story is still writing: whether repairing a governance reputation is possible at all, or whether it is simply priced in permanently.

↑ Back to Top

XI. Bear vs. Bull Case & Risk Radar

The bull case

The bull case rests on three things that are documented rather than promised.

First, position. Twenty gigawatts operating, in the world's fastest-growing major power market, with a 27% year-over-year growth rate and the largest single renewable site on earth under one owner.¹ India's electricity demand is not a forecast; it is a demographic and industrial fact.

Second, economics. Ninety-four percent EBITDA margins on contracted power, with plants generating above contracted commitment.¹⁸ Once built, these assets throw off cash for decades with minimal variable cost. Every gigawatt commissioned adds a near-permanent annuity.

Third, optionality. Storage converts the Khavda footprint from a producer of cheap midday electricity into a supplier of expensive evening electricity — using the same land, the same transmission, and a stated capital programme running to 50 GWh.¹ If the ₹2.5-in, ₹4-to-5-out spread holds at scale, storage improves both revenue mix and margin without requiring new land.¹⁹

Add the removal of two years of legal overhang — SEBI's orders, the SEC settlement, the pending criminal dismissal — and the bull argues the discount applied to this equity since 2023 was compensation for a risk that has now substantially resolved.

The bear case

The bear case is not that any of the above is false. It is that the equity has no margin for error.

Leverage first. Net debt of roughly ₹91,252 crore against reported equity near ₹19,965 crore, interest coverage around 1.1x, and ₹42,000 crore of further capex committed for FY27 alone.²¹²⁰¹⁹ Management gave a run-rate EBITDA path on the most recent call but no leverage target.¹⁹ The deleveraging trajectory investors cite is a sell-side model, not a company commitment — and there is a meaningful difference between the two.

Governance second. Charges being withdrawn by a prosecutor on jurisdictional grounds is not the same as a factual exoneration, the criminal motion had not been granted as of early August 2026,³¹⁶ and the individuals named remain in senior roles. Meanwhile the related-party surface area has increased, not decreased: the entire battery fleet and reclassified merchant capacity is now sold to a sister listed company at internally benchmarked prices.¹⁹

Competition third. PSU bidders with lower capital costs can win auctions AGEL cannot profitably win. The embedded fleet is safe; the growth is contestable.

Execution and grid fourth. Curtailment is already costing 5–7% of EBITDA, and the fix depends on transmission infrastructure AGEL does not control.¹⁹ Sustaining five gigawatts a year of additions at a remote desert site, while simultaneously commissioning a novel battery fleet and beginning multi-year pumped hydro civil works, is a genuinely difficult operating programme.

And finally: no dividend, ever, so far. The entire return depends on the reinvestment engine continuing to work at the current pace.

The activist lens

If an activist took a position tomorrow, the pitch would not be about solar panels. It would be about structure: an intra-group buyer for a majority of new capacity, affiliated equipment sourcing, promoter control at 62.43%,³⁷ board independence relative to that control, and a capital programme scaling faster than internal cash generation.

The rebuttals are real — the AESL contracts were board-approved, priced against SECI PPA guidelines, and genuinely reduce merchant volatility;¹⁹ SEBI examined related-party routing and found no violation;²⁵ the promoter group added no new pledges in FY26.³⁷ But the pattern that a skeptic would highlight is that each individual arrangement is defensible while the aggregate makes the enterprise harder for an outsider to verify independently. That opacity is a real cost of capital, whatever the underlying facts.

Risk radar, by mechanism

Refinancing and rate risk. The mechanism is not default — it is that project debt must be serviced and rolled across decades, and each 100 basis points of higher refinancing cost reduces distributable cash on assets whose revenue is contractually fixed and cannot be re-priced upward.

Regulatory and political risk. Two distinct channels: Indian auction and tender rules that determine future returns, and residual international scrutiny of the group that determines the cost and availability of foreign capital. These are independent risks that can fire simultaneously.

Execution risk. Concentrated at Khavda. A single site delivering the majority of growth means a single set of construction, logistics, and grid dependencies.

Counterparty risk. Discom payment delays convert profit into receivables. Watch working capital, not just EBITDA.

Curtailment and grid risk. Already quantified by management at 5–7% of EBITDA and structurally worsening if generation continues to outrun transmission.¹⁹

The three numbers that matter

Everything above can be monitored through three indicators.

One: megawatts commissioned against guided targets. This is the cleanest test of whether the execution machine still works. Management has committed to roughly 5 GW in FY27 after delivering 5,051 MW in FY26.²¹ A miss here is the earliest signal that something in the model has broken.

Two: net debt to EBITDA, measured consistently. Pick one definition — net debt over trailing reported EBITDA is the conservative choice — and track its direction quarterly. Whether the ratio actually falls as commissioned capacity converts to earnings is the entire deleveraging thesis, and it currently rests on analyst models rather than company guidance.

Three: gigawatt-hours of storage contracted and commissioned. This is the tell for whether the firm-power pivot is real economics or a capex line. Against 3,551 MWh installed and a stated 10,000+ MWh FY27 target,¹ the delivery rate — and eventually the disclosed margin on it — will settle whether storage is a second act or an expensive detour.

↑ Back to Top

XII. Epilogue & "If We Were CEOs"

As of the June 2026 quarter, AGEL operated 20.1 GW of renewable capacity, with 10.3 GW of that at Khavda against a 30 GW target for 2029, 3,551 MWh of battery storage installed, quarterly EBITDA of ₹4,122 crore, and a capital programme of ₹42,000 crore committed for FY27.¹¹⁹ The 50 GW target for 2030 implies roughly 30 GW of additional capacity in under four years — about six gigawatts a year, above even FY26's record pace.¹²

Reaching it requires not only continued execution but continued access to enormous amounts of capital at rates that leave a spread over compressed auction tariffs. That is the whole story in one sentence.

If we were running this company, three things would be different.

We would trade some speed for balance sheet. The 50 GW target is a public commitment made when auction economics were richer. Hitting it at any cost, funded with debt at interest coverage near 1.1x, optimises for a headline rather than for return on equity. A slower build with lower leverage would likely produce a better outcome for shareholders and a lower cost of capital that compounds into every future bid.

We would over-correct on board independence. With promoter holding at 62.43% and a fellow group company as the counterparty for the entire storage fleet, a related-party committee composed entirely of genuinely independent directors — with published, externally benchmarked pricing for intra-group contracts — would cost almost nothing and would remove the single most persistent objection to owning this equity.

We would disclose ahead of the requirement, not up to it. The most damaging feature of 2024 was not the indictment itself but that AGEL's disclosure posture toward US investors became the subject of a federal complaint. Voluntarily adopting the more demanding standard — on related-party terms, on curtailment, on leverage definitions — would be the cheapest reputation repair available.

The larger question this story raises is not really about one company. India needs an extraordinary volume of renewable capacity built in a short window, and AGEL has demonstrated it can build faster than anyone. But concentration of that build inside a single family-influenced conglomerate creates a systemic dependency: the country's transition timeline becomes partly hostage to one group's cost of capital and reputation. A more diversified developer base would be slower and probably more expensive. It would also be more robust. Which trade-off India has actually chosen is a question that will be answered by the auction results of the next five years.

↑ Back to Top

XIII. Outro & Links

Four ideas are worth carrying away from this story.

Contract structure is strategy. AGEL turned the world's purest commodity into a bond-like annuity by choosing 25-year sovereign-linked PPAs over merchant exposure. The 94% EBITDA margin is not operational brilliance; it is the arithmetic of a fixed-price contract meeting a zero-variable-cost asset. Understand the contract and you understand the business.

Land is the moat, and moats made of land have a shelf life. Khavda is genuinely hard to replicate and gives AGEL a cost and time advantage measured in years. It does not stop a rival with cheaper money from winning the next auction.

Governance overhang has a price even when the allegations do not stick. Between January 2023 and mid-2026, AGEL's operating performance improved almost monotonically while its equity story was hostage to events in a Brooklyn courtroom and a Mumbai regulator's office. Neither proceeding established wrongdoing. Both cost real money in cost of capital, shelved bond raises, and management attention. For investors, the lesson is that governance risk is not binary — it is a discount rate.

Growth infrastructure balance sheets must be sized against the definition you choose. The gap between roughly 5x leverage on run-rate EBITDA and roughly 11x on trailing EBITDA is not an accounting quibble. It is the difference between a business in a temporary build phase and a business that has borrowed against earnings it has not yet demonstrated.

For anyone going deeper, the primary documents matter more than the commentary: AGEL's quarterly results releases and earnings presentations, the FY25 annual report, the FY26 business and credit compendium, the Q1 FY27 earnings call — where the merchant-to-AESL reclassification, curtailment quantification, and battery unit economics were all disclosed in Q&A rather than prepared remarks — SEBI's September 2025 orders, the SEC's litigation releases, and the Eastern District of New York docket in the criminal matter, which as of August 10, 2026 remains the single most consequential open item in this story.

Adani Green Energy

Table of Contents

Adani Green Energy: Building India's Renewable Colossus, Under Scrutiny

I. Introduction & Episode Roadmap

II. Origins: From Trading House to Power Developer (2015–2018)

The conglomerate's logic

The founding choice

III. The Scaling Machine: Capital, Acquisitions & the SoftBank Bet (2018–2021)

What the bond actually proved

The record auction

The French validation

Buying five gigawatts

IV. Khavda and the Core Business: How a Renewable IPP Actually Wins

The business model, in plain terms

Why Khavda is different from a big solar farm

Solar by day, wind by night

Who else is in the ring

How AGEL wins, and how it loses

V. Beyond Generation: Storage, Hybrid Firming & the Next Leg

The economics management described

Pumped hydro: the hundred-year battery

How to size this honestly

VI. Crisis, Scrutiny & Vindication: Hindenburg to the DOJ Dismissal (2023–2026)

Partial vindication, first instalment

November 20, 2024

September 2025: SEBI closes the loop

2026: the case the government abandoned

Reading the arc

What a skeptic would look at today

VII. Current Management, Ownership & Capital Allocation

The professional layer

The family layer

Ownership

The TotalEnergies question

Capital allocation

VIII. The Financial Architecture: PPAs, Margins & Balance Sheet Risk

The income statement

The balance sheet

Why leverage is a feature, not an accident

The counterparty question inside the "de-risked" contract

Cost of capital versus peers

IX. Industry Structure: Porter's Five Forces & Durable Advantage

Mapping to Helmer's 7 Powers

X. Playbook: Business & Investing Lessons

XI. Bear vs. Bull Case & Risk Radar

The bull case

The bear case

The activist lens

Risk radar, by mechanism

The three numbers that matter

XII. Epilogue & "If We Were CEOs"

XIII. Outro & Links

References