AESI · Atlas Energy Solutions Inc. · Stock research

Current Thesis

The behind-the-meter (BTM) power pivot is still on schedule operationally and still unpaid for in the tape. Q2 printed on 2026-08-03: revenue $293.2M against $280.9M consensus (a beat), EPS $(0.20) against $(0.12) (a miss), adjusted EBITDA $49.5M, adjusted FCF $34.9M, net loss $25.1M. The number that set the stock was forward — Q3 2026 adjusted EBITDA guided to $30–45M against the $49.5M just delivered, attributed on the call to planned customer breaks and a decision to hold sand price (Q2 posted price $17.70/ton) rather than chase volume. On the power side nothing new was signed: the 120 MW private-grid PPA of 2026-04-01 remains the only contract, with a 26 MW bridge facility completed, 180–200 MW targeted deployed by year-end 2026, first energization end-Q1 2027 and first material BTM revenue in Q2 2027. Shares closed $12.45 on 2026-08-14, -37.0% from the $19.75 52-week high and -34.0% over three months, with RSI(14) at 61.1 — a bounce off the 2026-07-24 close of $11.96 that has not approached the ~$15 shelf lost in July. The narrative leg an investor would be buying here is a second large PPA re-starting the announcement cadence; the leg being priced is West Texas proppant.

Bullish and bearish views on Atlas Energy Solutions Inc.

The model's bull view on Atlas Energy Solutions Inc. (AESI), in brief: Funnel concentration improved on the 2026-08-03 call: management said remaining power capacity could be contracted in 2–4 projects versus a prior assumption of 8–10, with counterparties asking for 15–20 year terms. The bear view: The guide cut the run-rate: Q3 adjusted EBITDA $30–45M against $49.5M in Q2 (2026-08-03 call). Both cases follow in full.

Bull Case

• Funnel concentration improved on the 2026-08-03 call: management said remaining power capacity could be contracted in 2–4 projects versus a prior assumption of 8–10, with counterparties asking for 15–20 year terms. Fewer, larger signatures are needed to restart the news flow the stock trades on.
• The first contract is being built, not just announced: the 26 MW bridge facility tied to the 120 MW BTM contract was completed in Q2 (2026-08-03 release); 180–200 MW deployment is targeted by year-end 2026, first energization end-Q1 2027.
• Top line beat again: Q2 revenue $293.2M vs $280.9M consensus, +10.4% sequentially and +1.6% YoY, with record Dune Express volumes and 6 million tons of Last Mile shipments (2026-08-03).
• Part of the tape still carries $25 targets: Stifel maintained Buy, target lowered to $25 (2026-08-06); Raymond James Outperform $25 (2026-06-02). Against Barclays Underweight $14 (2026-07-16) and Goldman Sell $14 (2026-06-04), the published dispersion is unusually wide.
• Funding is in hand for the guided build: liquidity $292.9M at 2026-06-30 ($168.2M cash plus $125M undrawn ABL) against ~$200M of guided H2 capex, 90%+ of it Caterpillar genset deliveries and 2027 order deposits; legacy sand/logistics capex stepped down to $5–7.5M per quarter.

Bear Case

• The guide cut the run-rate: Q3 adjusted EBITDA $30–45M against $49.5M in Q2 (2026-08-03 call). The low end takes the quarterly print back below where the power capex program was underwritten.
• Margin compression is already in the filing: gross profit fell to $25.9M in Q2 2026 from $52.1M a year earlier; H1 2026 net loss $72.4M, including a $2.55M litigation settlement charge in Q2 (Q2 10-Q, period ended 2026-06-30).
• Order silence is now over four months: no new PPA or genset conversion since 2026-04-01, with 120 MW of the 240 MW order placed 2025-11-03 still unconverted.
• Leverage against a shrinking legacy earnings stream: long-term debt net $914.2M at 2026-06-30, including $450M of 0.50% convertible notes due 2031-04-15 issued 2026-04-09; H1 capex was $183.1M. The common dividend has been suspended since the Q3 2025 results.

• The power P&L is a 2027 event: first material BTM revenue recognition guided to Q2 2027 on the 2026-08-03 call, so every 2026 print is graded on proppant.

Setup & Price Structure

• Reference levels: last close $12.45 (2026-08-14); 52-week high $19.75 (-37.0%); three-month return -34.0%; RSI(14) 61.1. The June-to-July break ran $16.73 (2026-06-29) to $11.96 (2026-07-24), removing the ~$15 breakout shelf and the $13 area on the way.
• Where the base attempt sits: the $11.96 close from 2026-07-24 has held on a closing basis through 2026-08-14, and the Q2 print did not force a new low. RSI at 61 with price 37% under the high describes a bounce inside a downtrend; structural repair only begins on a weekly close back above the broken $15 shelf.
• The narrative is dead. Dated by the failed breakout ($16.73 on 2026-06-29 → $11.96 on 2026-07-24), the absence of any power order since 2026-04-01, and three target cuts inside four weeks (2026-07-15, 2026-07-16, 2026-08-05). The AI-power cohort narrative itself remains live; this name has stopped being paid for membership in it. A second investment-grade PPA plus a reclaim of $15 is what would move the label back toward accelerating.
• Crowding and positioning observables: $450M of converts placed 2026-04-09 at roughly a $14.51 conversion price (68.9275 shares per $1,000) creates a standing hedged-supply reference between spot and the broken shelf; 125,003,451 shares outstanding as of 2026-07-30; published targets span $14 to $25; the dividend suspension removed the income holder base; no Form 4 activity has surfaced in recent filings. Retail-sentiment coverage has thinned since the April PPA headlines — the visible flow is sell-side maintenance notes rather than new attention.

Catalyst Calendar (next 30 days)

• 2026-08-21, 2026-08-28, 2026-09-04, 2026-09-11 — weekly Baker Hughes US rig count (Fridays). The direct read on the frac activity that drives ~90% of revenue and the Q3 volume assumption.
• No company-confirmed event falls inside the 30-day window as of 2026-08-16. The next scheduled binary is the Q3 2026 print, ~2026-11-03 (est.), which grades the $30–45M EBITDA guide and the 180–200 MW year-end deployment target.

Elapsed catalysts

• Unscheduled, any week — second private-grid PPA or conversion of the remaining 120 MW of the 2025-11-03 genset order. The only discrete event that revives the pivot narrative; precedent is the 2026-04-01 announcement, guided then to ~$55M annualized adjusted FCF with sub-3.5-year payback. (passed 147d ago)

What Would Change Our Mind

• What flips it constructive: a second BTM contract at or above 100 MW with an investment-grade counterparty, paired with a weekly close back above the $15 shelf lost in July. That combination would show the pivot is being capitalised again rather than described.
• What flips it on fundamentals: Q3 adjusted EBITDA arriving at or above the $45M top of guidance with posted sand price held near $17.70/ton would refute the reading that legacy is deteriorating faster than power arrives.
• What breaks it: a weekly close below $11.50 forfeits the July base (2026-07-24 close $11.96) and extends the failed breakout into the lower third of the 52-week range. Secondary conditions: the ~2026-11-03 (est.) Q3 print arriving with no second PPA and adjusted EBITDA at or under the $30M low end, or any financing beyond the guided ~$200M H2 capex — an ABL draw, a new note, or equity — disclosed in an 8-K.

Correlation Notes

• Roughly 90% of revenue is proppant and logistics, so the daily driver is oil, frac-spread count and West Texas sand pricing; the Q3 guide was explicitly volume- and price-discipline-driven rather than power-driven.
• Correlation to the AI-power cohort (GEV, VRT, TLN, CEG) is episodic and headline-triggered. Since 2026-06-29 the stock fell 34% over three months while that cohort narrative stayed active — co-movement only appears on Atlas-specific order news.
• Caterpillar is the single supplier for the entire power build (2026-03-10 framework agreement, 1.4 GW reserved, ~$840M, deliveries 2027–2029); CAT commentary on genset lead times reads through directly to the 180–200 MW year-end target.
• The $14.51 convert conversion reference sits above spot and below the broken $15 shelf, so rallies into that band carry hedging-related supply independent of fundamentals.