Ameresco Insiders Bought the Dip Near the 52-Week Low. The Backlog Says Buy; the Cash Flow Says Wait.

This week, two of the people who know Ameresco best spent their own money on its stock. Director Joseph Sutton bought 9,700 shares at an average $20.87 on August 24 — roughly $202,000 worth. The same week, founder and CEO George Sakellaris, a 10 percent owner per filings, added 10,000 shares for about $214,000 on August 24–25 and roughly 5,000 more for about $113,000 in the days after. Combined, the senior insiders put about $530,000 into a stock trading a whisper above its 52-week low of $18.38, after a year that cut the shares from a high of $44.93 to about $22.

That is the opening of a fight, not an answer. Both camps can sign the same record: a record $6.73 billion project backlog, including $1.2 billion of data-center power awards booked in a single quarter, sitting on top of a business that burned through roughly $465 million of free cash over the last twelve months, with $1.82 billion of net debt under a $1.17 billion market value. The bulls see a multiyear data-center boom whose revenue simply has not landed yet. The bears see a capital-hungry developer whose cash keeps flowing the wrong way while executives leave. At $21.99, down about 25% this year, the question is which equation the current price is really paying for.

The shared record

As of August 29, 2026, the shares trade at $21.99, down about 4% on a day when Bloom Energy, Vertiv and GE Vernova all fell 3–4%. Market value is about $1.17 billion; enterprise value about $3.0 billion once net debt is added.

Q2 2026 (reported August 3): revenue of $515.5 million, up 9%, about 10% ahead of analysts' ~$465 million estimate; adjusted EBITDA of $62.8 million, up 12%; non-GAAP EPS of $0.20, in line; GAAP EPS of $0.18. Backlog reached $6.73 billion, up 32% year over year — $4.42 billion "awarded" and $2.30 billion contracted — with total revenue visibility of $11.49 billion. Management raised full-year 2026 non-GAAP EPS guidance to $1.15–$1.35.

The cash side is the other half of the record: Q2 operating cash flow was negative $107.2 million, trailing twelve-month free cash flow was negative $465 million, and twelve-month capital spending ran $339 million. Corporate debt is $384.8 million at a leverage ratio of 3.2x against a 3.5x covenant; energy-asset project debt is $1.55 billion backed by a $2.24 billion asset book. The CFO, Mark Chiplock, resigns September 25; the general counsel retired earlier in the month; the shares fell about 12.5% the day the news landed; guidance was reaffirmed.

Round One — the backlog is real; the cash is not

The data-center story is the most concrete thing Ameresco has ever had. The $1.2 billion of second-quarter data-center awards came from hyperscalers, data-center operators and landowners who need behind-the-meter power because the grid cannot deliver it. Five projects sit in the awarded backlog, and management says the pipeline of additional opportunities is growing. That is the bull's best punch: a leading-demand indicator strong enough that the CEO called the quarter an "inflection point" in customer trust.

The bear does not dispute the demand. It disputes the timetable. Of the $6.73 billion backlog, $4.42 billion is awarded rather than contracted. Management itself says converting an award to a contract takes six to 24 months, construction can run up to three years, and the revenue impact is limited in 2026, small in 2027, with the major contributions landing in 2028–2030. The equipment has not even been ordered. The bear's point is not that the backlog is fake; it is that today's price is being asked to wait up to four years for revenue that must first absorb capital.

Round One to the bull on direction, to the bear on timing. The data-center awards are the strongest demand signal this company has ever printed — and the stock's discount to that backlog is exactly what far-away cash deserves.

Round Two — who pays for the growth?

The bear's heaviest punch is the balance sheet. A $465 million trailing free-cash outflow, a quarter that burned $107 million of operating cash, $339 million of annual capital spending, and a corporate leverage ratio of 3.2x against a 3.5x covenant — 0.3x of headroom. The equity is a $1.17 billion claim under a $3 billion enterprise value, so the market values the whole company at roughly 11.5x this year's guided EBITDA. A data-center build-out that eats billions before it earns could force financing at exactly the moment the CFO is walking out the door, and a return on invested capital near 5% is below what the growth costs.

The bull's answer is structural, and it is the best comeback in the duel. Most of the scary debt is not the kind that endangers the equity. Roughly $1.55 billion is project-level debt against a $2.24 billion energy-asset book — a 69% advance rate on assets with contracted, recurring cash flows. The discretionary corporate debt is $385 million. And the market has already seen the monetization path: HASI, a disciplined infrastructure investor, paid $400 million in May for 30% of Ameresco's biogas business — a $1.8 billion valuation on a single segment — with $100 million of it going straight to Ameresco, which said it would use the proceeds partly to deleverage. Management has said it is open to the same partnership structure for data-center spending rather than diluting shareholders. An outside investor putting real money on that mark is harder to dismiss than a slide deck.

Round Two to the bull, narrowly. The debt is genuinely less frightening than the $3.5 billion headline suggests; but 0.3x of covenant headroom with an interim CFO is a real constraint, and the next financing round is the moment of truth.

Round Three — what the "raise" is worth

The bear's cleanest point: the number they raised is not an operational one. Full-year non-GAAP EPS rose to $1.15–$1.35 on an expected tax benefit of 25–40%, tied to a planned change in how the company accounts for transferable tax credits. Meanwhile the 2026 adjusted-EBITDA range of $250–270 million sits below the $270–295 million the company issued in March. The optics improved while the profit guide quietly came down. GAAP EPS was only $0.18 in Q2, and the first quarter was a $0.33 loss. At $22 the stock costs about 18x the guided non-GAAP number — and roughly 45x the same earnings measured against enterprise value.

The bull's answer: do not let the label bury the direction. The 10% revenue beat, the record awards and the mix shift — Energy Assets revenue up 21%, O&M up 29%, gross margin up to 17.7% — are operational, and they are why the tax tailwind is worth something. Recurring revenue growing faster than the project pipeline is the model improving, not just the tax rate doing the work.

Round Three is even: the bear takes the label, the bull takes the underlying. That leaves the duel resting on exactly one question — the pace of cash conversion over the next four quarters.

What $21.99 demands

Reverse the price. At $22, the whole company is valued near $3.0 billion — an enterprise value roughly 11.5x guided 2026 adjusted EBITDA for a business with heavily negative free cash flow. The equity gets whatever is left after the debt is served. The market is not paying for the data-center backlog; it is pricing slow conversion, cash trapped in projects, and an equity claim squeezed between the two.

Variable	Bull assumption	Bear assumption	Tripwire
Cash conversion	Billing milestones turn cash toward break-even through 2027	The model burns cash throughout the build-out; equity waits years	Deeply negative rolling free cash flow with leverage still at 3.2x at the Q2 2027 report
Backlog to contract	The $1.2B of data-center awards convert to contracts on schedule	Awards stall; "awarded" stays an unpaid promise	No meaningful award-to-contract conversion within two quarters
Funding	Project debt and partners like HASI carry the capital need; no dilutive raise	Covenant crunch or a below-book equity raise with an interim CFO	No permanent CFO in place by early 2027, or a waiver/raise to fund projects

Run the scenarios and the stakes get visible. Bull path: if adjusted EBITDA reaches about $400 million by 2029 as the backlog converts and cash turns positive, 10x EBITDA puts enterprise value near $4 billion; net out the ~$1.8 billion of net debt and the equity is roughly double today's $1.17 billion. Bear path: EBITDA stays around $260–300 million with continued cash burn, 9x gives an enterprise value near $2.6 billion, less $1.8 billion of net debt leaves equity near $0.8 billion — meaningfully below today. The equity is a levered wager on one variable, and the variable is timing.

Ruling

At $21.99, the bull case has the better evidence-to-expectation ratio — with conditions. The insiders' ~$530,000 spending spree is not the reason, it is the confirmation: a 10% owner putting his own money in at the lows, an outside infrastructure investor marking the biogas book at $1.8 billion, a record backlog, and a stock below book value. The bear needs the cash picture to keep deteriorating over several quarters, and at this price it is not being paid much to wait for that.

But this is not a clean buy-the-dip. It is a leveraged claim that has to survive the cash-crunch years before the data-center backlog pays for itself. The verdict: bull on the business and bull on the stock at this price, with the burden of proof on cash conversion across the next four reported quarters, and only for capital that can sit through the volatility of 2027 and 2028.

The ruling flips to bear if, by the Q3 2027 report, rolling free cash flow is still deeply negative with corporate leverage at 3.2x; if the data-center awards have not begun converting to contracted backlog within two quarters; or if growth needs an equity raise or a covenant waiver before a permanent CFO is in place. The earliest way to keep score is the next 10-Q: watch whether operating cash flow — the number that decides this duel — moves toward the company's own billing milestones by year-end.