BTU · Peabody Energy Corporation

Current Thesis

Two stories sit on one ticker. The one that gets Peabody quoted in AI-power coverage is Powder River Basin thermal — 82.0–88.0 million tons guided for 2026 into a US grid deferring coal retirements. The one that actually moves the P&L over the next two quarters is Centurion, the Australian premium hard coking coal longwall, which is still consuming money rather than producing it: the seaborne metallurgical segment printed $(17.0)M of adjusted EBITDA on 2.5Mt in Q2 2026, and on 2026-07-29 the full-year Centurion sales target was cut to 2.0–2.5Mt with 0.5–0.7Mt targeted for Q3. Group adjusted EBITDA was $24.0M in the quarter against $93.3M a year earlier, on revenue of $1,003.2M.

The narrative is maturing — coal-as-AI-power has been widely covered since the 2025 US policy shift and the company-specific leg has already been re-dated once at the 2026-07-29 print, so what remains is whether tons ship, resolved at the 2026-10-29 Q3 report.

Bullish and bearish views on Peabody Energy Corporation

The model's bull view on Peabody Energy Corporation (BTU), in brief: The met segment is a swing item, not a broken one. The bear view: The reported quarter was worse year-over-year on every line that matters. Both cases follow in full.

Bull Case

• The met segment is a swing item, not a broken one. Seaborne metallurgical adjusted EBITDA was $(17.0)M on 2.5Mt in Q2 2026 while absorbing Centurion commissioning costs (8-K, 2026-07-29). Management framed the remaining constraint as roof control in "a limited rock fault zone" and targets 1.5–2.0Mt of Centurion sales in H2 2026 — a bounded, dated claim that the 10-29 print grades directly.
• Seaborne thermal is funding the ramp. $52.1M of adjusted EBITDA on 3.0Mt with export realized pricing of $95.87/ton in Q2 2026 (8-K, 2026-07-29).
• The Q2 capital work removed a refinancing overhang. $250M of 2031 convertible notes issued, $241.2M of 2028 converts repurchased, roughly $350M of restricted cash and collateral released, and the revolving facility raised to $400M; cash was $526.3M and total liquidity $959.1M at 2026-06-30.
• Price has not run ahead of the story. The 2026-09-03 close of $27.64 sits 29.6% below the $39.25 52-week high on the adjusted series, with a three-month price change of -1.3% — the shares have not been paid for a ramp that has not yet been delivered.
• Group context is supportive. The Oil, energy & geopolitical cluster was tagged accelerating again on 2026-09-03, and thermal coal equities rallied in late August, with BTU reported up 7.3% in a single session during that move (Simply Wall St commentary, late August 2026).

Bear Case

• The reported quarter was worse year-over-year on every line that matters. Net loss attributable to common of $(90.6)M, or $(0.74) per diluted share in Q2 2026, versus $(0.23) a year earlier; adjusted EBITDA $24.0M versus $93.3M (8-K, 2026-07-29).
• The AI-power demand story is not visible in PRB economics. Powder River Basin adjusted EBITDA was $(7.1)M on 16.4Mt in Q2 2026, attributed to weather and maintenance downtime. A negative segment result on 16.4 million tons is the counter-evidence to the demand narrative the sector is being marked on.
• Centurion guidance has already moved once. The full-year target was set at 2.0–2.5Mt on 2026-07-29. A ramp that has been re-dated has no privileged claim on the next quarter, and Q3's 0.5–0.7Mt band is narrow enough to miss.
• Capex is large relative to current cash generation. 2026 capital expenditures guided at $340M against $24.0M of group adjusted EBITDA in Q2; the $0.075 quarterly dividend declared 2026-07-29 and paid 2026-09-03 is a further, if small, cash use.
• The most recent published sell-side mark is below the market. UBS maintained Neutral and lowered its price target to $26.50 on 2026-08-04 — under the 2026-09-03 close of $27.64.
• Australian cost and governance friction. On 2026-08-06 Peabody back-paid nearly $5M to 197 workers under a Fair Work Ombudsman enforceable undertaking; an 8-K dated 2026-08-27 records a consulting agreement transitioning the Chief Operating Officer, a leadership change during the ramp.

Setup & Price Structure

The reference close is $27.64 on 2026-09-03, 29.6% under the $39.25 52-week high, with RSI(14) at 60.2 and a three-month price change of -1.3%. That combination describes a mid-range tape with an upward tilt off the summer lows — not a base being broken out of, and not a name extended against its own annual range.

Crowding and positioning observables, stated as observables:

• Retail-sentiment coverage clustered after the move, not before it. Benzinga's 2026-08-25 piece screened BTU alongside TK and PSX as overbought energy names on RSI — momentum coverage arriving at the top of a sector rally.
• The market is trading above the latest published sell-side mark, the UBS $26.50 target of 2026-08-04, which means the incremental bid is paying for a ramp the covering analyst has not yet underwritten.
• Equity-linked supply exists. The $250M of convertible notes due 2031 issued in Q2 2026 is dilution-linked paper on the cap table, offset in cash terms by the $241.2M repurchase of the 2028 converts.
• No imminent earnings date. The next scheduled company print is 2026-10-29 — 55 days after the reference close — so there is no binary compressing positioning inside the next month.
• The cluster read is unstable. The Oil, energy & geopolitical status trail runs 07-26 accelerating → 08-07 dead → 08-11 accelerating → 09-03 accelerating. A group that flipped twice in five weeks is weak corroboration for a single name.

Catalyst Calendar (next 30 days)

• 2026-09-30 — Q3 2026 quarter close. The window for the 0.5–0.7Mt Centurion Q3 sales target ends here; the number itself is not disclosed until the October report.
• No company-scheduled event falls between 2026-09-04 and 2026-10-04. The only inputs inside the window are seaborne index prints (premium hard coking coal, Newcastle thermal), US power-burn and gas data, and any unscheduled 8-K.
• ~2026-10-29 (est., per Yahoo Finance and Zacks earnings calendars) — Q3 2026 results. Outside the 30-day window and the next genuine binary: Centurion Q3 tons against the 0.5–0.7Mt band, whether seaborne met adjusted EBITDA turns positive from $(17.0)M, whether PRB recovers from $(7.1)M, and the next dividend declaration.

What Would Change Our Mind

The read breaks on execution first. A third re-dating of Centurion — Q3 sales printing under the 0.5Mt low end on 2026-10-29, or the 2.0–2.5Mt full-year target cut again — removes the only company-specific leg that is not already priced by the sector, because the thermal segments alone produced $79.0M of combined adjusted EBITDA in Q2 against $340M of guided 2026 capex. A second PRB quarter of negative adjusted EBITDA would do similar damage to the AI-power framing that is carrying the group multiple.

On price, the gradeable condition is a weekly close below $24, which gives back the late-August coal-equity re-rating and leaves the shares a full leg beneath the UBS $26.50 mark of 2026-08-04 — the market declining to pay for the ramp ahead of the print. Confirmation in the other direction would be a weekly close above the low $30s with the met segment turning positive at the October report.

A softer failure mode: the 2026-10-29 print comes and goes with tons in-band but costs unchanged, the story stays where it is, and mainstream coal-for-AI coverage keeps arriving without a widening bid. That would date the narrative as saturated rather than broken, and the level above still applies.

Correlation Notes

• Met-coal peers are the closest read-through. AMR and HCC are levered to the same premium hard coking coal index that sets roughly the top of Peabody's seaborne met realizations; divergence between BTU and those two isolates Centurion execution from the commodity.
• The oil-levered names in the same cluster are a different driver. HP, NOG and ARIS trade on crude and US shale activity; BTU's Q2 seaborne thermal realization of $95.87/ton and its PRB volumes respond to Newcastle/API indices, Asian power demand and US gas prices, not Brent.
• US gas is the direct substitution variable for the 82.0–88.0Mt PRB guide: PRB burn competes on delivered cost against Henry Hub-linked generation, so a sustained gas break pressures the segment that the AI-power narrative rests on.
• Queensland weather disrupts met shipments in the December–March period, and AUD/USD moves the cost line reported in dollars.