Stock Analysis · Bulls vs Bears

Babcock & Wilcox Enterprises BW

🐂 Bulls vs Bears 🐻Report #110 Oct 2026
★ Willem owns this stock · 1% of the portfolio
63
high risk
🐂 6vs7 🐻
Bears lead
35%
▲ Chance of the biggest tailwind the forecasts miss
30%
▼ Chance of the biggest headwind the forecasts miss
The verdict

B&W is a shrunken, repaired boiler-service business, and the stock is now essentially an option on whether a related-party, thinly capitalised developer can finance a $2.4B gas plant in North Dakota. The aftermarket base is real and benefits from coal life extensions, but the stock's fall of ~75% since May tells you the market is already pricing heavy doubt about the counterparty, the schedule and the dilution.

The Bulls vs Bears podcast is being made
Two AI hosts debate this report, NotebookLM-style. It appears here automatically within a few hours.

Written by Claude (Anthropic’s AI) from public sources · Willem owns BW · AI can make mistakes — numbers in brackets point to the sources · Not financial advice

🐂 Bull arguments 6

  1. 1A practical way around the gas-turbine queueMedium
  2. 2Cost-plus contract structure shifts overrun risk to the customerMedium
  3. 3Balance sheet repaired, going-concern risk liftedMedium
  4. 4The coal-service base is getting a policy tailwindMedium
  5. 5Second project and the 1 GW Siemens reservationWeak
  6. 6Insiders bought into the decline; expectations are now lowWeak

🐻 Bear arguments 7

  1. 1The counterparty: a thin, insider-linked company with fading credit supportHigh
  2. 2Concentration and schedule drift on one projectHigh
  3. 3Thin and falling marginsMedium
  4. 4Serial dilution, with B. Riley earning fees on itMedium
  5. 5Governance and related-party webMedium
  6. 6The shortage window is temporary and competitors are scaling fasterMedium
  7. 7Legacy liabilities eat cashLow

Bull arguments show how well they survive the devil’s advocate (strong / medium / weak); bear arguments show the damage they could do after the counter-argument (high / medium / low).

Bull case — all positive points (most important first)

🐂 Bull #1A practical way around the gas-turbine queue

  • The point: Data centers want gigawatts of power now. The usual answer, gas turbines, is sold out for years. A gas-fired boiler driving a steam turbine uses a different, less congested supply chain, so B&W can deliver firm power sooner.
  • Evidence:
    • GE Vernova's gas-power backlog plus slot reservations rose from 100 GW to 116 GW in Q2 2026, and it plans only 24 GW/yr of output by 2028 [24].
    • Siemens Energy quotes lead times of "three years or more" [23].
    • Applied Digital's CEO said new turbine orders "might not be delivered until 2032" [22].
    • B&W signed the first 1.2 GW contract, and in August 2026 reserved 20 × 50 MW Siemens steam turbines, first delivery in 12–15 months, for an as-yet-unnamed second project [6,20].
    • POWER magazine (2026-09-10) reports data centers turning to industrial boilers and steam turbines as a trend, not a one-off [22].
  • Why it matters: Each 1 GW plant is roughly a $2B contract, about 3–4× B&W's entire pre-2026 annual revenue. B&W also expects an ongoing parts-and-service contract once each plant runs [9].
  • 😈 Devil's advocate: A boiler plus steam turbine is a 19th-century cycle, less efficient than a combined-cycle gas plant. EIA puts combined cycle at ~6,200–6,300 Btu/kWh [25]. B&W has published no heat rate; Interpretation: a steam-only gas plant likely burns ~30–50% more gas per MWh. At ~$2,000/kW (computed: $2.4B / 1.2 GW), it is also not cheap. The window closes as turbine makers add capacity in 2028–2030, and Wärtsilä/Caterpillar engines and Bloom fuel cells are scaling much faster [24,38,39]. Steam turbines themselves are tightening, and Siemens may sell its steam-turbine business [23].
  • Verdict: Medium The opening is real, but it is a 3–5-year window. One signed project, with a related-party buyer, does not yet prove repeat demand.

🐂 Bull #2Cost-plus contract structure shifts overrun risk to the customer

  • The point: B&W's history of near-death came from fixed-price contracts: European waste-to-energy plants, then B&W Solar. The new mega-contract is mostly reimbursable.
  • Evidence:
    • Of $2.4B, ~$2.0B is "reimbursable costs incurred plus mark-up". Fixed-price elements can only be added by a signed amendment, and either side may decline [3,7].
    • Delay penalties (liquidated damages) are taken from the fixed fee, are capped per unit, and are the "exclusive remedy" except for gross negligence or wilful misconduct [3].
    • Management says the project is "on budget" [11].
  • Why it matters: B&W can grow revenue several-fold without betting the company on its cost estimates.
  • 😈 Devil's advocate: Cost-plus only protects you if the customer can pay. Base Electron is a thinly capitalised special-purpose company (see Bear #1). The fixed portion (~$435M) still carries the boilers and turbines, the hardest part. The mark-up, delay-damage caps and overall liability cap are all redacted. Exhausting the delay-damage cap is itself an event of default [3]. Labour-shortage cost overruns already hit "a specific construction project" in Q2 2026 [4,11].
  • Verdict: Medium The structure is genuinely safer than history, but the risk has moved from cost risk to credit risk.

🐂 Bull #3Balance sheet repaired, going-concern risk lifted

  • The point: In 2024–early 2025 B&W carried a going-concern warning. That is gone.
  • Evidence:
    • 2025 asset sales and a May 2025 bond exchange (40% of notes swapped into 8.75% second-lien 2030 notes) [1].
    • ~$260M of equity raised in H1 2026, including $230M gross at $18.50 in May [4,18].
    • The 6.50% December 2026 notes were redeemed in August 2026 [35]. The only bond left is ~$129.5M of 8.75% notes due 2030 [1,4].
    • B. Riley's guarantee of the Axos credit line was released in February 2026 [34].
    • Cash and restricted cash were $382.8M at 30 June, and the revolver was undrawn at 30 June per the deck [5,6].
  • Why it matters: Customers, sureties and LC banks need a solvent contractor before awarding multi-year jobs. This was a precondition for winning data-center work.
  • 😈 Devil's advocate: The repair was paid for by shareholders, not operations.
    • Operating cash flow was −$68.9M in 2025 and only +$0.4M in H1 2026, and the H1 figure was flattered by a $100M rise in unpaid supplier bills [1,4].
    • ~$48.5M of the cash is LC collateral, and ~$61M went to the August note redemption after quarter-end [4,35].
    • The pension remains underfunded by ~$168M, and the 7.75% preferred costs ~$15M a year in cash [4].
    • Debt still costs 8.75%, and the Axos facility matures in January 2028 [1].
  • Verdict: Medium Strong. Solvency is no longer the issue near term, but the business still does not fund itself.

🐂 Bull #4The coal-service base is getting a policy tailwind

  • The point: US coal plants are retiring much more slowly than planned, and every extra year of running means outage work, parts and emissions retrofits for B&W.
  • Evidence:
    • EIA: only 2.6 GW of coal retired in 2025 vs 8.0 GW planned, the fewest since 2010 [26].
    • DOE emergency orders forced several plants to keep running [26,27].
    • EPA repealed the 2024 mercury-rule (MATS) amendments (effective April 2026) and partially repealed power-plant CO₂ rules, effective 2026-11-16 [28].
    • B&W parts revenue rose 17% in 2025 on higher coal-plant utilisation [1,10].
    • The new ~$130M scrubber project at a US coal plant (September 2026) is direct evidence of utilities spending to keep coal units [21].
    • The TerraSpark 1.6 GW coal-plus-carbon-capture project in West Virginia (FEED stage, DOE grant up to $18.5M) names B&W as boiler partner [36].
  • Why it matters: Aftermarket is the part of B&W with real recurring revenue, and it doesn't depend on Base Electron.
  • 😈 Devil's advocate: On 2026-09-11 the D.C. Circuit ruled the first DOE "must-run" order (Campbell, Michigan) unlawful, which weakens the forced-extension lever [27]. Rolling back CO₂ rules also removes the regulatory driver for B&W's carbon-capture technologies [28]. Coal plant life extensions delay decline; they don't reverse it, and 6.4 GW is still scheduled to retire in 2026 [26].
  • Verdict: Medium This is a real, multi-year support for the core, but it is a slowing decline rather than growth.

🐂 Bull #5Second project and the 1 GW Siemens reservation

  • The point: B&W is not a one-contract story if the second project lands. Management guided a second data-center full notice to proceed in 2026, and Applied Digital's group holds an option for another 1.2 GW [7,11].
  • Evidence:
    • B&W has started work with Siemens on 20 steam-turbine sets, 1 GW in total (2026-08-11) [20].
    • The pipeline is ">$14B", including 4–6 GW of generation [6].
    • There is a coal-to-gas data-center partnership with Denham Capital (September 2025) [5].
  • Why it matters: A second, unrelated owner would cut concentration and prove the product sells on its own merits.
  • 😈 Devil's advocate: No second project had been announced by 2026-10-10, with less than three months left in the guided year. The pipeline is unweighted. B&W is now carrying reservation risk on 20 turbines without a named customer; the turbines' terms, deposits and cancellation clauses were not found. Management itself said "we're not going to book all of that business" [10].
  • Verdict: Weak to-Medium, pending evidence. This is the single most important signpost for the next 3 months.

🐂 Bull #6Insiders bought into the decline; expectations are now low

  • The point: The stock is ~$5.46, down from ~$22 in May. Insiders have bought, sell-side targets sit far above, and the bar is low [30,31,32].
  • Evidence:
    • Open-market buys: CEO Young bought 7,000 shares at $9.65 (2026-08-12); director Moeller bought 5,000 at $9.50 (2026-08-13) [30].
    • No 10b5-1 or open-market officer sales were found.
    • Analyst targets (public coverage only; reports paywalled): Needham Buy $20 (2026-08-31); Northland $23 (May); Craig-Hallum Buy $19 (March). MarketBeat consensus is Hold, with an average PT ~$15.40 [31].
  • Why it matters: If the counterparty doubts clear, the stock has room to re-rate.
  • 😈 Devil's advocate: The insider buys are token-sized (~$50–70k against the CEO's ~1.7M-share holding). B. Riley's group sold 1.16M shares at $9.00 in February 2026 to repay debt secured by pledged shares [29]. Price targets mostly pre-date the slide.
  • Verdict: Weak This is a sentiment cushion, not a business argument.

Bear case — all negative points (most important first)

🐻 Bear #1The counterparty: a thin, insider-linked company with fading credit support

  • The point: The $2.4B contract is with Base Electron, Inc., not with Applied Digital.
  • Evidence:
    • Applied Digital's 8-K says Base Electron was "formed by certain executive officers, directors and/or affiliates of Applied Digital" [13].
    • Applied Digital owns ~10%, carried at $2.0M, and says Base Electron "does not have sufficient equity at risk to finance its activities without additional subordinated financial support" [15].
    • Base Electron's funding so far is a $58.5M on-demand loan from Applied Digital (facility up to $100M) [15].
    • Applied Digital's February 2026 guarantee to B&W was uncapped but terminable:
      • if Base Electron listed its shares, or
      • if Base Electron raised ≥$50M while current on B&W invoices, or
      • if Applied Digital paid a $50M fee (the 8-K summary says "by" 1 August 2026, while the exhibit as read by our researcher says "after"; the conflict is unresolved), or $100M otherwise [13,14].
    • In Applied Digital's 10-Q for the quarter to 2026-08-31, its "variable interests in Base Electron consist of its equity investment and related party promissory note". No guarantee is listed, and it is "not contractually required to provide financial support" [15].
    • Interpretation: the guarantee has probably been ended or replaced. A third-party search summary claims it was terminated on 2026-05-29 against a $37.0M payment. UNCONFIRMED: we could not find this in a primary filing.
    • Wolfpack Research (short seller, 2026-03-12) alleges Base Electron has ties to B. Riley, B&W's ~19% holder, including a shared address and Bryant Riley as a director. UNCONFIRMED: only the web summary was read [16].
    • A securities class action (Cho v. B&W, N.D. Ohio; class period 2025-11-05 to 2026-03-11) alleges the counterparty links and the guarantee's exit terms were not disclosed. These are allegations [17].
  • Why it matters: Under cost-plus, B&W's profit depends entirely on the owner paying its invoices. Receivables from Base Electron were already $73.9M at 30 June [4]. If financing fails, B&W keeps what it has billed, but the backlog that drives the whole equity story disappears.
  • 🛡️ Counter-argument:
    • Applied Digital is a large, well-funded data-center developer with a strong reason to get the power built for its North Dakota campus.
    • A search summary (UNCONFIRMED) says Applied Digital signed a 15-year power purchase agreement (PPA) and credit-support agreement with Base Electron on 2026-10-04. Base Electron can terminate it if financial close has not happened by 2027-03-31. A PPA from a creditworthy buyer is exactly what project lenders need.
    • B&W's contract lets it suspend work if unpaid, and cancellation clauses generally recover costs plus profit on work done [1].
  • Severity: High Even with a PPA, financing is not closed, the owner is insider-linked, and B&W's direct guarantee appears gone.

🐻 Bear #2Concentration and schedule drift on one project

  • The point: One project dominates the company, and its finish date has moved twice in five months.
  • Evidence:
    • Base Electron is ~94% of backlog [4].
    • Commercial operation target: "2028" (10-K, 2026-03-16) → "2029" (Q1 10-Q, 2026-05-11) → "2030" (Q2 10-Q, 2026-08-10, checked directly) [1,2,4].
    • The conditional-use permit was only submitted in August 2026 [5,11].
    • 58% of total backlog is expected to be recognised after 2027 [4].
    • Management called the project "ahead of expectations" on the same day the 10-Q showed the 2030 date. No explanation was found [11].
  • Why it matters: Every quarter of delay pushes revenue and EBITDA out. It also lengthens B&W's exposure to the counterparty and to inflation in the ~$435M fixed portion.
  • 🛡️ Counter-argument: The 2030 date may refer to the last of four units, with the first earlier. Q2 revenue recognition was higher than expected because of manufacturing milestones [11].
  • Severity: High This is unexplained drift on the only large job, and no analyst has publicly pressed management on it [11].

🐻 Bear #3Thin and falling margins

  • The point: B&W is becoming a low-margin pass-through builder.
  • Evidence:
    • Gross margin was ~15% in Q2 2026 vs ~30% in Q2 2025.
    • Operating margin was 3.7%, and adjusted EBITDA margin 6.8% [4,6].
    • The 10-Q blames project mix plus a skilled-labour shortage that caused "lower than expected productivity on certain job sites" [4].
    • The mark-up on the $2.0B variable portion is undisclosed [3].
  • Why it matters: On cost-plus, EBITDA dollars rise, but the fixed costs that remain hurt more: the 8.75% notes, ~$15M of preferred dividends, pension funding and corporate overhead. Interpretation: at a 5–8% margin, a 1.2 GW plant may add only tens of millions a year of EBITDA.
  • 🛡️ Counter-argument: Cost-plus margins are low but predictable. The parts business, at higher margin, keeps growing. 2026 EBITDA guidance was raised, not cut [6].
  • Severity: Medium

🐻 Bear #4Serial dilution, with B. Riley earning fees on it

  • The point: Equity has been the main source of cash.
  • Evidence:
    • ATM sales: $67.5M in November 2025, then more in Q1 2026 [1,2].
    • A 12.4M-share offering at $18.50 in May 2026 [18].
    • Shares rose from ~130.4M to ~149.0M in H1 2026 [4].
    • Warrants for ~10.5M shares at $4.11 are held by Applied Digital and Base Electron, and are still in the money [4].
    • B. Riley is lead underwriter, ATM agent (3% fee) and financial adviser (3% of debt deals) [4,19].
    • A $50M buyback was authorised in July 2026 [35].
  • Why it matters: Operating cash flow is near zero. Any delay at Base Electron, or the working capital needed for a second project, likely means more equity at today's much lower price.
  • 🛡️ Counter-argument: With bonds redeemed and ~$300M+ of cash, there is no pressing maturity before January 2028. The buyback signals management sees no immediate need.
  • Severity: Medium High.

🐻 Bear #5Governance and related-party web

  • The point: B. Riley is shareholder, lender-adjacent, banker and adviser. The CEO is its former President. The main customer is insider-formed at Applied Digital and allegedly linked to B. Riley.
  • Evidence:
    • B. Riley relationships: [1,4,19].
    • Applied Digital 8-K: [13].
    • Wolfpack allegations (UNCONFIRMED): [16].
    • Class action: [17].
    • Material weaknesses in internal control were unremediated at 31 March 2026 [2].
    • The general counsel departed in January 2026 [1].
  • Why it matters: It raises the risk that disclosures are incomplete, and that capital decisions serve B. Riley's fee and liquidity needs.
  • 🛡️ Counter-argument: B. Riley's guarantee has been released, and its stake fell from ~22% to ~19%. The board added an independent director in January 2026. The Q2 10-Q reports no new litigation [4].
  • Severity: Medium

🐻 Bear #6The shortage window is temporary and competitors are scaling faster

  • The point: B&W's opening exists only while gas turbines are scarce.
  • Evidence:
    • GE Vernova is lifting output to 24 GW/yr by 2028, with actions toward 30 GW in 2030 [24].
    • Caterpillar is tripling large-engine capacity, mostly in 2027–2029 [39].
    • Wärtsilä has >2.4 GW of US data-center engine orders [38].
    • Bloom guides ~$4B of 2026 revenue (summary only) [39].
  • Why it matters: By the time B&W proves its first plant (2029–2030), customers may be able to buy more efficient combined-cycle plants again.
  • 🛡️ Counter-argument: The "enhanced combined cycle" design lets owners bolt on a gas turbine later and convert the boiler into a heat-recovery unit, so early B&W plants are not stranded [11].
  • Severity: Medium

🐻 Bear #7Legacy liabilities eat cash

  • The point: Before any growth, B&W must fund a ~$168M pension deficit (PBGC waiver conditions), ~$15M/yr of preferred dividends and 8.75% interest on ~$130M of notes [1,4].
  • Why it matters: These fixed costs consume much of the core business's EBITDA.
  • 🛡️ Counter-argument: Pension funding is scheduled, and higher interest rates have shrunk the deficit (from $176M to $168M in H1) [4].
  • Severity: Low Medium.

Tailwinds NOT in the forecasts

Management gives only a 2026 adjusted EBITDA target ($80–105M) and no 2027+ targets. Consensus is thin (2026 EPS ~$0.27 on aggregators) [6,31].

  • Second data-center project at full notice to proceed, with an unrelated, creditworthy owner — 35% chance within 12 months · Impact if it happens: High
    • What it is: A ≥1 GW order using the reserved Siemens turbines.
    • Why it's not priced in: Nothing is in backlog or guidance, and the share price implies doubt about even the first project.
    • What would trigger it / the sign to watch: An 8-K naming the owner. Base Electron's own 1.2 GW option being exercised would count for less, because of concentration.
    • Why this probability: Management guided it for 2026 and has reserved hardware, but the 2026 deadline is nearly gone with no LNTP announced. Below a coin flip.
  • Base Electron reaches financial close with project lenders — 45% chance by 2027-03-31 · Impact if it happens: High
    • What it is: A project-finance package backed by Applied Digital's PPA, if confirmed, so B&W's invoices are funded by real lenders.
    • Why it's not priced in: The stock trades at ~$813M market cap, versus a $2.4B contract. Interpretation: the market discounts heavy failure risk.
    • What would trigger it / the sign to watch: County permit approval; an Applied Digital or Base Electron financing announcement; the B&W Q3 call (~2026-11-03).
    • Why this probability: A 15-year PPA from a large data-center operator is financeable, but a first-of-its-kind gas-boiler plant with a new sponsor and a slipping date makes the March 2027 deadline tight.
  • Coal-fleet retrofit wave (scrubbers, gas conversions, life-extension rebuilds) — 40% chance of ≥2 more $100M+ coal-related awards within 24 months · Impact: Medium
    • What it is: More projects like the ~$130M September 2026 award.
    • Why it's not priced in: Guidance is EBITDA-only, and consensus treats the core as flat.
    • What would trigger it / the sign to watch: Utility integrated resource plans that cancel coal retirements; bookings excluding Base Electron above ~$150M a quarter.
    • Why this probability: Retirements are slowing, but the D.C. Circuit ruling and utility capex priorities make large retrofit awards lumpy.
  • BrightLoop / TerraSpark becoming real orders — 12% chance within 36 months · Impact: Low–Medium
    • What it is: A first commercial BrightLoop sale, or TerraSpark moving from design study to a boiler order.
    • Why it's not priced in: There is no revenue, and the CO₂ policy driver weakened in September 2026.
    • What would trigger it / the sign to watch: Massillon start-up (late 2027); TerraSpark financing.
    • Why this probability: Both are first-of-kind and lack a policy push.

Headwinds NOT in the forecasts

  • Base Electron misses financing, is restructured, or suspends work — 30% chance within 18 months · Impact if it happens: High
    • What it is: The PPA's 2027-03-31 financial-close deadline, if confirmed, passes, or the owner pauses. Backlog collapses back toward ~$0.2–0.4B.
    • Why it's not priced in: Guidance and the backlog figure assume the project proceeds.
    • What would trigger it / the sign to watch: Permit delays; a rising Base Electron receivable; any change to the demand loan at Applied Digital.
    • Why this probability: Insider-formed sponsor, thin equity, slipping date and an unclosed financing are each moderate risks on their own, and together they are meaningful.
  • New equity raise below $6 — 40% chance within 12 months · Impact: Medium
    • What it is: ATM or follow-on issuance to fund working capital or a second project.
    • Why it's not priced in: There is no announced need.
    • What would trigger it / the sign to watch: Q3 operating cash flow; a new or expanded ATM; turbine deposit outflows.
    • Why this probability: Cash covers near-term needs, but B&W's track record and B. Riley's fee incentive point to more issuance if a second project signs.
  • Adverse ruling or settlement in the securities class action — 20% chance of a material cash cost (>$10M net of insurance) within 36 months · Impact: Low–Medium
    • What it is: The case survives a motion to dismiss and settles.
    • Why it's not priced in: The 10-Q discloses no new material litigation.
    • What would trigger it / the sign to watch: Lead-plaintiff appointment; the ruling on the motion to dismiss.
    • Why this probability: Most such cases settle within insurance or are dismissed.
  • Labour-driven cost overruns on fixed-price work — 35% chance of a disclosed loss provision of ≥$10M within 24 months · Impact: Medium
    • What it is: The Q2 productivity problem repeats on fixed-price construction or on the ~$435M fixed part of Base Electron.
    • Why it's not priced in: EBITDA guidance was raised.
    • What would trigger it / the sign to watch: Gross margin below ~14%; commentary on boilermaker availability.
    • Why this probability: The shortage is industry-wide and peaks just as Base Electron construction starts in 2027.

Investment Risk Score: 63/100

Risk area Weight Score (0–100) Why
Business & competition 25% 60 The core is a shrinking-fleet aftermarket. The data-center edge is a time window, not a moat, against faster-scaling engines, fuel cells and returning turbines.
Financial health & dilution 20% 65 Bonds are redeemed and cash is solid, but operating cash flow is ~zero. Preferred, pension and 8.75% notes are fixed drains. There is a history of serial equity issuance.
Execution (delivering the plan) 20% 70 First-of-kind 1.2 GW build. The date slipped twice in five months. Labour overruns in Q2. The second project is late versus guidance.
Dependency & concentration 15% 85 ~94% of backlog sits with one thinly capitalised, insider-formed owner, and B&W is single-sourced on Siemens steam turbines.
Regulation, policy & geopolitics 10% 45 Policy currently helps (coal extensions, permitting), but the D.C. Circuit ruling shows it can reverse, and the project still needs local and air permits.
Sentiment & expectations 10% 40 The stock is down ~75% from May, so little good news is expected. Some downside remains from the class action and overhang.

Overall = 63/100 (high). Weighted: 15 + 13 + 14 + 12.75 + 4.5 + 4 = 63.25. The score is driven by concentration and credit risk on one related-party project rather than by solvency, which has improved materially. First deep dive, so no change to report.

Conclusion

  • What kind of company this is: A 160-year-old boiler-service business, ~$585M of legacy revenue, bolted to a single, very large, mostly cost-plus construction job. Owning BW today is a bet that Base Electron gets financed and built, and that B&W can repeat it with other owners before turbine supply catches up.
  • The 2–3 things that will decide the outcome:
    1. Base Electron's financial close and permit, with a 2027-03-31 deadline (UNCONFIRMED).
    2. Whether a second, unrelated data-center owner signs.
    3. Whether B&W can fund working capital without another big equity raise.
  • Signposts — confirming the bull case:
    • an 8-K with full notice to proceed for a second project, with a named, unrelated owner;
    • Base Electron financing closed and the county permit granted;
    • quarterly bookings excluding Base Electron at ≥$200M;
    • positive operating cash flow without growth in payables.
  • Signposts — breaking the bull case:
    • Base Electron suspension, re-scoping, or another schedule slip;
    • a new ATM or offering below ~$6;
    • the Base Electron receivable rising while revenue from it falls;
    • a loss provision on a fixed-price job.
  • Balanced verdict:
    • The strongest bull version: B&W has the only scaled boiler-based answer to a real gas-turbine bottleneck. Its contract terms protect it from the overruns that sank it before, its balance sheet is the best in a decade, and the stock already prices in heavy failure.
    • The strongest bear version: the "$2.4B contract" is with a thinly capitalised company formed by Applied Digital insiders. The direct guarantee appears gone, the completion date has drifted to 2030, the margin is single-digit, and the company has funded itself with shareholder dilution arranged by its largest holder.
    • Interpretation: the evidence today leans bearish on the counterparty and schedule, and neutral-to-positive on the legacy aftermarket. The next six months (permit, financing, second project) should resolve most of the uncertainty.

Open questions & data gaps

  • Status of Applied Digital's guarantee: the termination date, the payment made, and what replaced it. We could not confirm it from a primary source, and it matters greatly.
  • The 2026-10-04 PPA and credit-support agreement: its terms and the 2027-03-31 financing deadline are UNCONFIRMED (search summary only).
  • Why the commercial-operation date moved from 2028 to 2030, and whether it refers to the first unit or the last.
  • Undisclosed figures: the mark-up on $2.0B of variable charges, the delay and performance damage caps, and the termination-for-convenience terms. All are redacted.
  • The Siemens 1 GW reservation: deposits, cancellation terms and who bears the cost if no project signs.
  • Wolfpack's claimed B. Riley ties to Base Electron: the full report was not read.
  • Heat rate and cost per kW of B&W's configuration versus combined cycle. No sourced figure exists.
  • The Q4 2025 and Q1 2026 transcripts were partially or third-party-sourced only. The parts margin is never disclosed.
  • Whether the $3.0M PBGC instalment due 2026-09-15 was paid. It is minor.
Sources
  1. Form 10-K FY2025 — SEC EDGAR (B&W) — 2026-03-16 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000018/bw-20251231.htm
  2. Form 10-Q Q1 2026 — SEC EDGAR (B&W) — 2026-05-11 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000030/bw-20260331.htm
  3. Design-Build Agreement, Base Electron / The Babcock & Wilcox Company (exhibit to Q1 10-Q, dated 2026-02-26) — SEC EDGAR — 2026-05-11 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000030/baseelectron-bwdesignxbu.htm
  4. Form 10-Q Q2 2026 — SEC EDGAR (B&W) — 2026-08-10 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000068/bw-20260630.htm
  5. Company Overview investor presentation, August 2026 (8-K Ex. 99.2) — B&W — 2026-08-10 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000066/bwiroverview-august2026x.htm
  6. Q2 2026 results press release (8-K Ex. 99.1) — B&W — 2026-08-10 — https://www.sec.gov/Archives/edgar/data/1630805/000163080526000066/e-2026q2earningsrelease.htm
  7. Q4/FY2025 results and Base Electron full-notice-to-proceed release, plus 8-K — B&W — 2026-03-04 — https://www.sec.gov/Archives/edgar/data/1630805/000110465926023138/tm267910d1_ex99-1.htm
  8. Q3 2025 results release — B&W — 2025-11-04 — https://www.sec.gov/Archives/edgar/data/1630805/000110465925106447/tm2530246d1_ex99-1.htm
  9. "B&W Announces AI Data Center Power Generation Solution and Signs LNTP with Applied Digital" — B&W — 2025-11-04 — https://www.sec.gov/Archives/edgar/data/1630805/000110465925106447/tm2530246d1_ex99-3.htm
  10. BW Q1 2026 earnings call transcript — AllMind — 2026-05-11 — https://allmind.ai/earnings/event/bt_BW_2026_Q1
  11. BW Q2 2026 earnings call transcript — Earnings Whispers — 2026-08-10 — https://earningswhispers.com/transcript/BW/Q22026
  12. BW Q4 2025 earnings call transcript — MarketBeat — 2026-03-04 — https://www.marketbeat.com/earnings/reports/2026-3-4-babcock-wilcox-enterprises-inc-stock/
  13. Applied Digital Form 8-K (Guarantee in favour of The Babcock & Wilcox Company) — SEC EDGAR — 2026-03-04 — https://www.sec.gov/Archives/edgar/data/1144879/000149315226008849/form8-k.htm
  14. Guarantee agreement, Exhibit 10.1 to Applied Digital 8-K — Applied Digital IR / SEC — 2026-02-26 (filed 2026-03-04) — https://ir.applieddigital.com/sec-filings/all-sec-filings/content/0001493152-26-008849/ex10-1.htm
  15. Applied Digital Form 10-Q, quarter ended 2026-08-31 (Notes 7 and 12) — SEC EDGAR — October 2026 (cover date 2026-10-06) — https://www.sec.gov/Archives/edgar/data/0001144879/000114487926000059/apld-20260831.htm
  16. "B&W's much-hyped $2.4 billion deal appears to be with an opaque RILY-backed corp…" — Wolfpack Research (short seller; web summary only) — 2026-03-12 — https://www.wolfpackresearch.com/items/babcock-%26-wilcox%E2%80%99s-%28bw%29-much-hyped-%242.4-billion-deal-appears-to-be-with-an-opaque-rily-backed-corp%3B-and-apld-has-already-contradicted-disclosures-about-the-deal
  17. "BW Investor Alert: … Securities Fraud Lawsuit" (class action notice; allegations) — Levi & Korsinsky via Business Wire — 2026-05-12 — https://www.businesswire.com/news/home/20260512384258/en/
  18. B&W prices common stock offering — B&W / Business Wire — 2026-05-15 — https://www.sec.gov/Archives/edgar/data/1630805/000110465926063192/tm2614762d1_ex99-1.htm
  19. Prospectus supplement (424B5) — SEC EDGAR (B&W) — 2026-05-14 — https://www.sec.gov/Archives/edgar/data/0001630805/000110465926061195/tm2614623-1_424b5.htm
  20. "B&W Signs Agreement with Siemens Energy to Commence Work on 20 Steam Turbines for Data Center Power Generation" — Business Wire — 2026-08-11 — https://secure.businesswire.com/news/home/20260811901159/en/
  21. "B&W Receives Award to Begin Work on Approximately $130 Million Air Quality Control Project" — B&W — 2026-09-10 — https://www.babcock.com/home/about/corporate/news/babcock-and-wilcox-receives-award-to-begin-work-on-approximately-130-million-air-quality-control-project
  22. "To meet power demand, AI data centers turn to industrial boilers and steam turbines" — POWER magazine — 2026-09-10 — https://www.powermag.com/ai-data-centers-turn-to-industrial-boilers-and-steam-turbines-for-power/
  23. "Steam turbines and the gas turbine bottleneck" — Turbomachinery Magazine — 2026-09-11 — https://www.turbomachinerymag.com/view/steam-turbines-gas-turbine-bottleneck-siemens-energy
  24. GE Vernova Q2 2026 results press release — GE Vernova — 2026-07-22 — https://www.gevernova.com/news/press-releases/ge-vernova-reports-second-quarter-2026-financial-results-raises-2026-financial
  25. Capital Cost and Performance Characteristics for Utility-Scale Power Generating Technologies (AEO2025) — U.S. EIA / Sargent & Lundy — AEO2025 edition (exact publication date not confirmed) — https://www.eia.gov/analysis/studies/powerplants/capitalcost/pdf/capital_cost_AEO2025.pdf
  26. "Retirement delays of U.S. electric generating capacity may continue in 2026" — U.S. EIA Today in Energy — 2026-02-23 — https://www.eia.gov/todayinenergy/detail.php?id=67206
  27. "Court rules that Trump can't force a Michigan coal plant to keep running" — Canary Media — 2026-09-11 — https://www.canarymedia.com/articles/fossil-fuels/court-rules-against-trump-michigan-coal-plant
  28. "EPA Partially Repeals Power Plant Carbon Rules…" — Beveridge & Diamond — September 2026 — https://www.bdlaw.com/publications/epa-partially-repeals-power-plant-carbon-rules-punts-questions-of-broader-authority-to-new-proposal/
  29. Form 4 / Schedule 13D/A summary, BRC Group sale of 1,155,382 BW shares at $9.00 — Stock Titan (summary of SEC filing) — February 2026 — https://www.stocktitan.net/sec-filings/BW/form-4-babcock-wilcox-enterprises-inc-insider-trading-activity-965b7418bef1.html
  30. BW insider transactions (Form 4 aggregation) — Quantisnow — viewed 2026-10-10 — https://www.quantisnow.com/insiders/BW
  31. BW analyst ratings and price targets — MarketBeat — viewed 2026-10-10 — https://www.marketbeat.com/stocks/NYSE/BW/forecast/
  32. BW stock data (close 2026-10-09, 52-week range) — stockanalysis.com — 2026-10-09 — https://stockanalysis.com/stocks/bw/
  33. "As losses endure, Babcock & Wilcox has a dilution and cash flow problem (Downgrade)" — Seeking Alpha (lead only) — 2026-03-06 — https://seekingalpha.com/article/4879407-as-losses-endure-babcock-and-wilcox-has-dilution-cash-flow-problem-downgrade
  34. 8-K: Tenth Amendment to Axos Credit Agreement — SEC EDGAR (B&W) — 2026-03-03 — https://www.sec.gov/Archives/edgar/data/1630805/000110465926022894/0001104659-26-022894-index.htm
  35. 8-K: share repurchase authorisation and full redemption of 6.50% Senior Notes — SEC EDGAR (B&W) / Business Wire — 2026-07-13 — https://www.sec.gov/Archives/edgar/data/1630805/000110465926082840/tm2620294d1_8k.htm
  36. "A new coal plant in the U.S.? Once unthinkable, now a strong maybe" (TerraSpark) — POWER magazine — June 2026 — https://www.powermag.com/a-new-coal-plant-in-the-u-s-once-unthinkable-now-a-strong-maybe/
  37. "Doosan Enerbility lands first North American steam turbine order" — Turbomachinery Magazine (search summary only) — 2026-03-19 — https://www.turbomachinerymag.com/view/doosan-enerbility-lands-first-north-american-steam-turbine-order
  38. Wärtsilä secures new order to power AI-driven data-center growth in the US — Wärtsilä (search summary only) — 2026-06-29 — https://www.wartsila.com/media/news/29-06-2026-wartsila-secures-new-order-to-power-next-wave-of-ai-driven-data-center-growth-in-the-u-s--3767555
  39. "Caterpillar's order backlog hit a record $63 billion" — Motley Fool (search summary only); Bloom Energy Q2 2026 highlights — Nasdaq (summary only) — 2026-08-03 / 2026-07-28 — https://www.fool.com/investing/2026/08/03/caterpillars-order-backlog-hit-a-record-63-billion/ ; https://www.nasdaq.com/articles/bloom-energy-q2-earnings-call-highlights
Disclaimer. AI-generated research and podcast (Claude by Anthropic, voices by Google Gemini) for educational purposes — not financial advice and not a recommendation to buy or sell. Willem (MR.INVEST) owns this stock. Check the company’s own filings before making any decision.
← All Bulls vs Bears reports