Ameresco director Brian Cox buys $100,085 in stock

Ameresco shares were recently bought by director Brian C. Cox: 4,625 shares for $100,085 at $21.64 (near $22.07 market level), with the stock down nearly 13% over the prior week. The company reported Q2 2026 revenue of $515.5M vs $465M consensus and non-GAAP EPS of $0.20 (at consensus), alongside adjusted EBITDA of $62.8M vs $61M consensus, and raised full-year non-GAAP EPS guidance. Ameresco also secured a record $1.8B in new awards (including $1.2B from data centers), lifting backlog to $6.73B, while Cantor Fitzgerald reiterated an Overweight rating (PT $44) and Oppenheimer kept Perform.

Analysis

The real signal is not the insider buy; it is that the market may still be underestimating how quickly data-center power demand can shift AMRC from a cyclical efficiency contractor into a quasi-infrastructure growth story. If that mix holds, the multiple can re-rate before the cash flow fully shows up, because sell-side models usually lag backlog-to-revenue conversion by 2-3 quarters.

Near term, the main winner is AMRC itself, but the second-order beneficiaries are electrical infrastructure and grid-support vendors that sit upstream of hyperscale buildouts. The loser set is smaller retrofit/ESCO competitors with weaker data-center exposure, because hyperscaler work tends to concentrate with names that can deliver scale, permitting, and execution; that can widen dispersion inside the industrial services basket over the next 6-12 months.

The key risk is that awards are not earnings: backlog quality, project margins, and working-capital intensity matter more than headline size. If the company needs to finance growth faster than it converts cash, the equity can stall even with strong bookings. The 1-3 month catalyst path is the next guidance update and any evidence that data-center awards are converting into margin expansion rather than just revenue growth; the 6-18 month thesis breaks if backlog rolls off or hyperscaler capex pauses.

Contrarian view: the consensus may be too focused on the top-line surprise and not enough on execution risk in a low-visibility project mix. The stock can stay cheap if investors assume the data-center pipeline is lumpy or non-repeatable. A cleaner expression may be to buy AMRC only on post-earnings weakness, not ahead of it, unless there is confirmation that free cash flow inflects with the backlog.

Key Decisions for Investors

• Buy AMRC on pullbacks into the next earnings cycle, with a 3-6 month horizon; thesis works if management confirms data-center awards are converting to margin lift and not just backlog inflation.
• Use a limited-risk call spread in AMRC for 3-6 months rather than stock if liquidity is acceptable; target upside on a re-rating while capping risk if project timing slips.
• Watch for a relative-strength long AMRC vs. a broader industrials ETF (XLI) only after the next quarter shows free-cash-flow conversion; otherwise the name-specific execution risk is too high for a clean pair.
• Set an alert on gross margin and operating cash flow in the next print; if revenue beats again but cash conversion deteriorates, fade the move and reduce exposure.
• If hyperscaler capex commentary weakens across ETN/EME/PWR, use that as a falsifier for the AMRC data-center thesis and cut risk quickly.