BlackRock’s group pays with equity and adds no acquisition debt, but that cheque is a third of AES’s enterprise value. What funds the plan, and the bond clause that could still reshape AES’s debt, sit in its own filings.
AES closed at $17.28 on 27 February 2026. Two days later its board agreed to sell at $15.00 a share to Bayo Ogunlesi’s Global Infrastructure Partners (GIP), which BlackRock owns, EQT, CalPERS and the Qatar Investment Authority.
They will fund 100% of the purchase price with equity. That $10.68bn buys a company with an enterprise value of about $33.4bn; the rest is existing net debt. “No new debt” buys a third of the company.
I laid the proxy’s two forecasts side by side and set the board’s need against the dividend in AES’s quarterly filing. Board and bidders priced AES on different ones, each unused by the other.
The paid half settles what pays for the board’s growth plan: a dividend the owners stop paying, about the size of the board’s stated need. It also prices a put: two AES bond series can demand repayment at 101 if Moody’s and S&P both cut AES to junk. I put 75% on no such cut; the deciding number is how much of the pair is handed back after the 2 December close.
All equity, a third of AES
The merger proxy of 15 May puts the cash needed to close at “approximately $10.69 billion” and says Horizon Parent, L.P. (the proxy’s “Parent”), the buyers’ Delaware acquisition partnership, expects “this amount to be funded through equity financing”. On the release’s count of 712 million shares, the cheque is a shade less: the $10.68bn above.
The other $22.7bn of the enterprise value is AES’s proportional net debt at 31 December 2025: its share of the debt of the businesses it owns, net of cash, the basis the release itself uses. On that basis the equity is 32% of the value.
Three vehicles signed equity commitments on 1 March: GIP’s fifth flagship fund, Tower Bridge Infrastructure Partners, which GIP manages for CalPERS, and EQT’s Infrastructure VI fund, each for its “pro rata portion”. The proxy never prints those portions. A protest filed at the Federal Energy Regulatory Commission (FERC) on 20 July gives them as control. Its authors were Public Citizen’s Tyson Slocum, Jim Baker of the Private Equity Stakeholder Project and Kerwin Olson of the Citizens Action Coalition of Indiana.
“BlackRock affiliates would control 56.625% of AES (which includes BlackRock’s management of CalPERS’ stake)”, they wrote, with EQT at 33.375% and Qatar at 10%. They asked FERC to “compel BlackRock to legally and fully separate its non-controlling asset management business from its active management of utilities”. They also quoted former FERC chair Mark Christie: “a public utility has public service obligations; it is not just another company seeking to maximize returns to its shareholders.” FERC approved the deal on 17 November anyway, and closing is expected on 2 December.
PJM’s market monitor, the independent monitor of the grid operator for much of the eastern US, gives the cheques from the buyers’ FERC application. GIP pays 36.625% of the purchase price and EQT 33.375%. CalPERS’s 20% and Qatar’s 10% make up the rest. So GIP writes the largest cheque without paying a majority, and BlackRock’s affiliates control a majority through the CalPERS stake GIP manages.
That equity took nine months to assemble. GIP bid alone in June 2025 at $15.00, with sponsor equity and “no incremental leverage on the Company”. On 3 November GIP said its main partner, a party the proxy calls Party D, “had unexpectedly withdrawn from the transaction”. On 24 November Ogunlesi called Andrés Gluski, AES’s chief executive, with $14.25, citing “changes in certain tax assumptions and renewables operating performance.”
With EQT beside GIP, the price was $14.75 on 19 December, and the final $15.00 came on 5 February 2026, though GIP’s first bid had assumed no dividends after 2025. That price is 35.5% above $11.07, AES’s close on 8 July 2025, the last before a press report put the sale talks in public. It is 13.2% below the $17.28 close of 27 February. Gluski said the deal “maximizes value for existing stockholders”; on announcement day the shares closed 17.8% lower, at $14.21.
The buyers’ own words were about the build: Ogunlesi saw “a need for significant investments in new capacity”, EQT’s Masoud Homayoun “expanding power demand worldwide”, CalPERS’s Sarah Corr a “landmark investment”, and the Qatar Investment Authority’s Mohammed Saif Al-Sowaidi a way of making “energy transition a reality”. None of the four lines says what pays for the build.
Jay Morse, then AES’s chair, put the need plainly in the deal release: “AES has a significant need for capital to support growth beyond 2027.” Absent this deal, he said, the company “would likely require a plan that includes reduction or elimination of the dividend and/or substantial new equity issuances.” The cheque goes to the shareholders who sold, and the need Morse named stays with AES.
The backstop is for bondholders
The board could not borrow for that need. Its reasons to sell name “high leverage relative to peers” and warn that more debt “would potentially result in the loss of the Company’s investment grade credit ratings.” AES’s September quarter filing shows $33.3bn of consolidated debt at 30 September, every subsidiary’s counted in full, only $6.36bn of it, about 19%, recourse debt AES’s holding company must repay. The rest is repaid only from project and subsidiary assets.
“No new debt” may cover only closing day, with borrowing moving above AES afterwards. Ohio’s utility commission approved the merger application for AES Ohio on 17 September 2026 with no duty to stay investment grade. The Ohio Consumers’ Counsel wants a rehearing to add one, and ring fencing, the legal walls that keep a utility’s cash from its owners.
AES’s deal deck, relaying the buyers, says “no back leverage”, meaning no borrowing at a holding company above AES secured on its equity, and that AES “will cease paying fixed quarterly dividends at closing”. A deck binds no one. Any debt above AES still needs cash paid up to Horizon Parent, and AES’s own reports will show that.
Goldman Sachs and Citigroup committed on 1 March to about $4.72bn of term loan, and the proxy says this debt financing “is not intended to fund the Merger Consideration and is intended only to refinance” AES debt whose terms would demand repayment on a change of control.
Two bond consents later cut the term loan to about $3.0bn. It is expected to be drawn only if holders of named notes at AES, at DPL, its Ohio utility holding company, and at IPALCO, its Indiana one, “accept an applicable ‘change of control’ offer” after closing.
That offer is narrow. Under AES’s consent solicitation of 5 March, the merger “may result in a ‘Change of Control Triggering Event’” for a series only “if the rating on such series of Notes is lowered below an investment grade rating by both of the Rating Agencies” as a result of the merger. For the 2030 and 2031 notes those agencies are Moody’s and S&P.
The 2031 indenture counts a cut unless the agency says it “was not the result, in whole or in part” of the deal, and an upgrade back to the starting rating inside the window cancels it. The cut must land between the 2 March announcement and 60 days after closing, later while either agency has a downgrade review open.
AES offered holders of four series, $3.4bn in all, $1.00 per $1,000 to waive the clause. The 2032 notes consented on 18 March, and the 5.450% notes due 2028 by 31 March for a larger fee.
Two said no: the 3.950% notes due 2030 and the 2.450% notes due 2031, coupons well under what AES paid on new 2029 notes in June. The proxy lists them among five series still able to force an offer; the other three sit at subsidiaries: DPL’s $400m of 4.35% notes due 2029 and IPALCO’s $475m due 2030 and $400m due 2034.
The subsidiaries’ indentures need a cut by two of Fitch, Moody’s and S&P, below investment grade for IPALCO and by any notch for DPL, so I keep them outside my odds, which on the AES pair are a judgement with no counted base rate.
For no cut: I found no downgrade or downgrade review since the 2 March announcement; the consent release said “Neither AES nor Parent currently expects that the ratings of the Notes will be downgraded by any Rating Agency”. Against it: AES’s own debt schedule at 30 September 2025 lists its holding company’s senior notes at Baa3 at Moody’s and BBB- at S&P and Fitch, one notch above junk at both agencies that decide the put.
A second clause needs only one agency, from a different pair. Contracts the merger agreement lists as “Specified Ratings Credit Support Contracts” would demand new collateral, letters of credit or “cash collateral” if either S&P or Fitch drops AES below BBB minus; the agreement gives no amounts. Cash is where the two forecasts part.
Two forecasts, one price
AES management wrote two forecasts. The Standalone Case, for 2026 to 2030, assumed AES stayed public. The board used it, and J.P. Morgan and Wells Fargo built their fairness opinions on it. The Acquisition Case was written for bidders, to show what AES could do “with access to capital” and “a flexible dividend policy”.
Two sentences in the proxy matter more than the tables. “The Board did not review the Acquisition Case.” And: “The Company did not provide the Standalone Case to Parent.” The bidders worked from a management case the board never reviewed, and the board judged the price on one the bidders never saw. The bidders’ case, the proxy adds, “was prepared before the Standalone Case was finalized”.
The two differ most in AES Clean Energy, the US renewables arm. The Standalone Case ramps it to 2.8 gigawatts placed in service a year from 2028. The Acquisition Case ramps to 5.5 gigawatts a year from 2030 and assumes AES “regularly recycles capital by selling minority stakes in operational projects to fund growth, increasing Levered Free Cash Flow, but reducing Adjusted EBITDA.” Levered free cash flow here is the cash reaching AES’s holding company from its businesses, after its investments, corporate costs, interest and net borrowing.
So the bidders’ case builds about twice as fast and sells pieces of what it builds.
The fairness opinions bracket $15.00 widely. J.P. Morgan’s and Wells Fargo’s six ranges run from $9.07 to $20.25 a share, and $15.00 sits inside five of them. Ranges that wide show only that $15.00 was defensible on the board’s case, the one forecast the banks used.
The two cases reach about the same EBITDA by 2029, so the price barely turns on which forecast is right; what pays for the build does.
If the cheque only buys the shares, what pays for AES’s plan, and are its 2031 notes worth holding for the put until 31 January 2027?
Below the paid line:
What pays for the board’s plan once the public dividend stops, counted from AES’s filings
The bond clause priced at my odds, in total and per note, with one sensitivity table
What this changes for the next all equity utility sale, and four dated checks to watch



