SLNH
SLNH · Soluna Holdings, Inc. · Catalyst Calendar · as of August 18, 2026
Banded 0–3 / 3–6 / 6–12 months, earliest first. Numbering is gapless and runs straight through the bands, so a boundary never resets the count. An ID marks position in the current ordering and changes between revisions, so catalysts are named by title.
On August 3, 2026 the Governor of Texas directed the Public Utility Commission of Texas and ERCOT to conduct a comprehensive verification and audit of all data centers advancing through ERCOT's interconnection process, and stated that any project failing to comply will be denied connection to the Texas grid. ERCOT suspended the Batch Zero Large Load classification notifications due by August 7, 2026 in market notice M-A080326-01, and filed its request for good cause exceptions on August 10, 2026 at item 42 of Project 59142, with a draft proposed order attached, ahead of the Commission's open meeting of August 20, 2026, which is the dated point here. The filing is broader than the notice was: ERCOT is pausing the Batch Zero study process itself, not only the classification notifications, and says it will make a further filing for that meeting with the scope and timing. No completion deadline has been named, but ERCOT states twice that the verification and audit "will take several months" and that it expects to begin shortly after the August 20 meeting.
This lands on the pipeline, which is where the valuation premium sits, and the Q2 filing made the exposure larger rather than smaller. The development pipeline was restated at approximately 6.3 GW, and during the quarter the company redesignated capacity for AI workloads at Rosa, from 187 MW to 242 MW, at Hedy, from 120 MW to 198 MW, at Ellen, from 100 MW to 145 MW, and at Fei, from 120 MW to 240 MW. The Form 10-Q filed August 13, 2026 places Fei and a newly named Project Gladys, 150 MW, in ERCOT interconnection planning in terms, and Dorothy 3 at 300 MW sits behind them. Interconnection study completion remains the nearest observable proof that the pipeline is more than a list of sites.
It also lands on the same Commission this calendar is already tracking on two rules. The large-load interconnection standard (D13) and the co-location and net-metering rule (R1), adopted while this file carried it as pending, are both before the Commission now running the audit. Both statuses were checked against the Commission's docket on August 13, 2026: the co-location rule was adopted on March 26, 2026 and the large-load standard has not been adopted.
An exception on a short timetable, and study completions resume. An audit that thins a queue the operator has described at more than 438 GW is worth more to a developer with sites already advancing than to the speculative requests that make up much of that number.
Study completion and queue positions freeze behind an audit with no deadline, and the pipeline stays a list for longer. The liquidity edge of this has gone: with $113.4 million of unrestricted cash filed at June 30, 2026, a delay is now a schedule question rather than a solvency one. It is a schedule question over the assets carrying the valuation premium.
The scope is no longer entirely unestablished, and it moved against this company. ERCOT's August 10 filing states that "the Commission and ERCOT interpret data centers as including virtual currency mining facilities and other computational loads", that requests for information will go to every entity with a large load submitted for Batch Zero, and that community impact information will be sought from all data centers and virtual currency mining facilities of 25 MW or more that have not yet energised. Every unenergised project in this pipeline is above 25 MW. What is still not established is whether any specific site was submitted into Batch Zero, which is the fact that decides whether the request for information reaches it, and no filing read here names this company. The date is the Commission's own; the outcome for this issuer is unknown and is not guessed at.
Source: Office of the Governor of Texas, directive of August 3, 2026; ERCOT market notice M-A080326-01, August 3, 2026; Form 10-Q for Q2 2026, filed August 13, 2026, for the project statuses. Filed Neither the directive nor the notice appears in any SEC filing by this issuer, including the Form 10-Q filed the same day this was written.
Inferred, not announced: the prior proxy set the universal-proxy notice deadline under Exchange Act Rule 14a-19 at June 19, 2026, and that rule runs 60 days before the anniversary of the prior year's meeting, implying an anniversary around 18 August. A preliminary proxy statement was filed on August 14, 2026 and is not read here, so the meeting date and the resolutions are still to be taken from it.
Ordinarily routine: director elections, auditor ratification in KPMG's first such vote replacing UHY LLP, and say-on-pay. Two items make it worth reading. Say-on-pay follows a quarter in which stock compensation of $10.2 million exceeded revenue, June grants totalled at least 12,631,694 restricted shares, and the Compensation Committee raised the chief executive's base salary to $600,000 retroactive to January 1, 2026 with a target bonus of 100% of salary. Any proposal touching authorised shares or the incentive plan is a direct dilution signal.
One structural caveat limits what this meeting can do: the board sits in three classes on staggered three-year terms, so only one class stands for election and replacing a majority takes at least two successive meetings. Directors are removable only for cause and only on a 75% vote. The annual meeting is a weaker shareholder lever here than at a company with an annually elected board.
No dilution-related proposals on the ballot, implying management believes existing authorised headroom is sufficient. Strong say-on-pay support. Board additions with AI infrastructure or project-finance credentials.
A proposal to increase authorised common above 375 million, or to enlarge the incentive plan, would signal that the roughly 130 million share headroom remaining after the June quarter is expected to be consumed. A weak say-on-pay vote against the Q1 compensation charge would be a governance flag.
Source: Prior-year DEF 14A; 8-K of Nov 2025 (authorised shares raised from 75m to 375m); 8-K filed Jul 24, 2026, Item 5.02(e). Filed
Contractual: the Series A was issued as not redeemable prior to this date except in limited circumstances. The date arrives regardless of what anyone does. Redemption would cost $123.0 million of liquidation preference plus roughly $35 million of accrued arrears, far beyond the company's means, so this is a forcing function for disclosure rather than an action. Arrears grow about $2.8 million a quarter, which makes eventual resolution more visible: a negotiated exchange into common, a partial paydown from any large financing, or continued accrual.
A resolution on favorable terms, an exchange at a discount to face or a plan to resume dividends, removes an overhang the equity market largely ignores and the preferred market clearly does not. The preferred traded at $10.75 to $11.00 in June, near 44% of its $25 face, so settlement below face is plausible.
Arrears keep compounding at 9% of face, roughly $11.1 million a year, silently enlarging the claim ahead of the common. Any exchange into stock at $1.35 would issue an enormous number of shares. The protective provision requiring two-thirds holder approval before issuing anything senior constrains future financing structures.
Source: Series A prospectus supplement; Form 10-Q at Jun 30, 2026 (4,920,045 shares, ~$35.1m arrears); Forms 4, Jun 2026. Filed
Chair Kevin Warsh is expected to speak, and the address is watched for how the framework and communication approach evolve. Elevated relative to a normal symposium for two reasons: the July statement was notably short and offered neither a path nor explicit conditions, so this is one of few structured opportunities to read the reaction function before September; and it falls two weeks after the 13 August results, making it the first major macro event the post-earnings share price is marked against.
A dovish or neutral framing takes some September hike probability out of the market, supporting Bitcoin and micro-cap risk appetite into the autumn, the window in which a capital raise would most plausibly be needed.
Confirmation of the hawkish tilt implied by the three July dissents hardens September expectations. For a stock with a beta of 5.13 and a fifth of its shares short, that compounds directly into the financing cost of everything downstream.
Source: Symposium dates published by the Federal Reserve Bank of Kansas City, the organizer; FOMC statement and press conference, Federal Reserve, Jul 29, 2026. Filed
The company placed this within nought to three months as of July 14, 2026, which closes the window about October 14. Estimate The date is the company's own range measured from its own statement, not from the date of this document: a window written as "within three months" and left unanchored moves every time this file is restamped. As of July 14, 2026, Phase 2 modular data centers of 9 MW reached substantial completion and Phase 3 of 14 MW was under construction and described as ahead of schedule. Kati 1B totals 35 MW within the 83 MW Kati 1 site. This is incremental contracted Bitcoin hosting revenue: the highest-certainty revenue growth on the calendar and the item least dependent on a Kati 2 tenant.
Completion ahead of schedule with immediate customer deployment, demonstrating execution capability that supports credibility on the much larger AI builds. Kati 1 fully energised at 83 MW takes total operating capacity meaningfully higher.
Equipment or interconnection delay. The precedent exists: Dorothy 1A required transformer repair that took two modular data centers offline in early 2026, and Briscoe has needed remediation. Capacity completed into a weak Bitcoin tape may energise into softer hosting demand.
Source: Monthly business updates, 9 Apr and Jul 14, 2026. Filed
The July meeting held the target range at 3.50% to 3.75% on a 9 to 3 vote, with all three dissents favoring a rate hike. September carries a Summary of Economic Projections and dot plot. Three transmission channels apply. Rate expectations drive Bitcoin, which drives about 23% of revenue directly and hosting demand indirectly. Rates set risk appetite for the equity issuance the company depends on, and a beta of 5.13 means macro shifts land several times harder than on the index. Least discussed, the long end sets the discount rate for a fifteen or twenty year lease: the 30-year Treasury reached roughly 5.25% in late July, its highest since 2007.
A dovish turn lifts Bitcoin, small-cap risk appetite and speculative AI infrastructure at once, the three things this equity is levered to. Cheaper project debt improves Kati 2 economics directly.
The skew has turned. Following the July hold, market-implied odds moved toward a September increase rather than a hold. A hawkish September compresses Bitcoin and micro-cap valuations together, raising the cost of equity roughly a month after the 13 August disclosure.
Source: FOMC statement and meeting calendar, Federal Reserve, Jul 29, 2026 Filed. Implied probabilities as reported Jul 29 to 30, 2026 Market.
Texas allocates certain wholesale transmission charges by a large customer's load during four coincident-peak intervals across June to September, so flexible loads curtail heavily through summer and then stop. Four of five sites are in Texas, and the June and July updates disclosed increased curtailment at Dorothy 1A and 1B, Dorothy 2 and Kati 1A. Uptime and therefore hosting and mining revenue should recover in the December quarter. This is why the June and September quarters are seasonally weakest and the December quarter mechanically stronger, a pattern easy to misread as operational improvement.
December-quarter revenue rebounds visibly as curtailment ends and Kati 1B phases add capacity, producing a strong-looking sequential quarter. Demand response revenue may also be elevated for the summer period.
The June and September quarters print weak first. If the market has not internalised the seasonality, the August and November reports disappoint on a basis that is structural rather than a deterioration in the business.
Source: Monthly business updates, 9 Jun and Jul 14, 2026; Form 10-Q glossary. Filed
Statements at 2:00 pm Eastern, with a Summary of Economic Projections in December. The same channels as the September decision, inheriting a hawkish skew after three July dissents favored a hike. A specific overlay applies to December: it coincides with year-end tax-loss selling season, and the stock is far below its 52-week high of $5.14, making it a natural tax-loss candidate. Note that the long-end move is not under committee control, since 30-year yields reached 19-year highs in late July even as the front end was held, so a dovish decision does not automatically relieve project-financing costs.
A dovish December, combined with post-tax-loss January reversal dynamics, has historically been a favorable setup for beaten-down high-beta micro caps. Cheaper capital improves Kati 2 financing terms directly.
A hawkish December plus tax-loss selling plus a short interest last read at 34.41 million shares, about 14% of the enlarged count, into thin year-end liquidity is a poor combination. Any equity issuance completed into that window would be maximally dilutive.
Source: Meeting calendar published by the Federal Reserve Filed. 52-week range and short interest at the Jul 31, 2026 close Market.
Contractual: the Credit Agreement permits Tranche B draws from September 12, 2025 until this date, subject to conditions. Tranche B funds Kati construction and was cut from $18.5 million to $6.0 million in the April 1, 2026 amendment, with the released capacity redirected to the $12.5 million Tranche C for Briscoe. Only $6.0 million of committed project debt remains available before the window shuts. The agreement contemplates up to $64.5 million of Additional Tranche Loan Commitments, but only subject to lender and agent approval, which is not a commitment.
The remaining $6.0 million drawn before expiry, plus an approved Additional Tranche commitment for Kati or Dorothy 3. An incremental commitment would be third-party validation of a different kind: a project lender underwriting the asset base rather than an equity buyer underwriting a narrative.
The window closes with capacity undrawn because draw conditions are not met, or the company draws it purely for liquidity rather than construction. Either pushes the next dollar toward equity issuance. A 1.00% commitment fee is payable on undrawn Tranche B amounts.
Source: Form 10-Q at Mar 31, 2026, Note 8; 8-K, Apr 2, 2026 (Amendment No. 1). Filed
Repairs were ongoing as of July 14, 2026, delayed roughly two weeks by high winds, with an additional crew and crane deployed. Renewable energy credit sales for the June quarter were completed. Briscoe is the foundation of the vertical-integration thesis and of Dorothy 3. Management projected $6 million to $11 million of year-one adjusted EBITDA on $20 million to $24.4 million of annualised revenue against a $53.0 million headline price. The 8-K/A of June 17, 2026 complicates that: total cash consideration was approximately $55.9 million, and the asset's 2025 accounts show revenue of $8,950,927 against a net loss of $14,773,838.
Repairs complete, full output achieved, and disclosed contribution lands in or above the projected range. Visible compression of cost of revenue at the co-located Dorothy sites would be the cleanest evidence.
Further repair delay or cost. Also a presentation trap: because Briscoe's primary purpose is powering Soluna's own co-located data centers, intercompany power sales eliminate on consolidation and appear as lower cost of revenue rather than added revenue. Only third-party and renewable energy credit sales reach the top line, so the projected revenue may never appear as investors expect even if the asset performs.
The projected $20 million to $24.4 million of annualised revenue is unlikely to appear in consolidated revenue at that magnitude, because the intercompany portion eliminates on consolidation. This is an inference from the asset's stated purpose, not a company statement about accounting treatment.
Source: Monthly business update, Jul 14, 2026; press release, Apr 2, 2026; 8-K/A, Jun 17, 2026. Filed
Based on the statutory filing deadline for a non-accelerated smaller reporting company and prior-year practice. No date announced. Presumably accompanied by a conference call following the August precedent. This is the confirmation quarter: the June quarter reads as transitional because of the acquisitions, so September is the first clean period showing post-acquisition run-rate burn, full Briscoe contribution, and whether the liquidity trajectory is sustainable. It also captures the tail of summer curtailment and first contribution from Kati 1B phases.
Adjusted EBITDA approaching breakeven as Briscoe's cost benefit and Kati 1B capacity compound, with cash burn moderating and no emergency financing having been required. Material weakness remediated. Any Kati 2 lease progress disclosed alongside.
Two consecutive quarters of heavy cash consumption with no AI lease and a diminished cash balance shifts the frame from growth optionality to financing risk. A materially higher share count versus August confirms facility usage. Coverage-covenant pressure would be visible here first.
Source: Periodic reporting deadlines for non-accelerated filers; Form 10-Q cover page at Mar 31, 2026; Q1 filing precedent of May 15, 2026. Estimate
Texas Senate Bill 6 requires the Public Utility Commission of Texas to complete its review of wholesale transmission charge allocation and amend rules as necessary by that date, explicitly considering alternatives to the four-coincident-peak method, such as multiple seasonal peaks or peak energy intervals. This rewrites the economics of the Texas fleet directly: the coincident-peak mechanism is what the company curtails around each summer, so replacing it changes the uptime against transmission cost trade-off across all four Texas sites and every project in the pipeline.
A methodology spreading charges across more intervals could reduce the value of curtailing during four narrow windows, allowing higher summer uptime and therefore higher June and September quarter revenue. It could also raise costs for inflexible competitors relative to a flexible load, widening the relative advantage.
A methodology assigning more transmission cost to large loads regardless of curtailment behavior would eliminate an arbitrage the company currently harvests, raising costs with no offsetting revenue. The bill's stated purpose includes ensuring large energy users contribute appropriately to transmission cost recovery, which is not a framing that favors large loads.
Source: Texas Senate Bill 6, signed Jun 20, 2025, and Public Utility Regulatory Act amendments, published by the Texas Legislature; PUCT implementation dockets, published by the Public Utility Commission of Texas Filed. Law firm analyses of both, 2025 to 2026 Market.
The docket is Project 59142, and neither of this company's documents has ever carried a docket number for this rule, which is exactly why its status was never checkable: there was nothing to look up. The project is the Review of ERCOT's Interconnection Processes for Large Loads, and the rule under it is proposed new 16 TAC 25.194. It has not been adopted. The docket runs to August 13, 2026 across 45 items and contains no adoption order and no order of any kind, four months after the date on which reporting had suggested adoption could come. Comments on the staff memo closed April 17, 2026 as this card said. What the Commission did instead was approve the interconnection process rather than the rule: on June 18, 2026 it approved ERCOT Planning Guide Revision Request 145, Batch Zero Process for Large Load Interconnections, and Nodal Protocol Revision Request 1315 alongside it, establishing Batch Zero as a one-time transitional study process for loads of 75 MW or more. That is the process the Governor's audit has now paused. The rule sets entry cost and process for every new load of 75 MW or more, which covers Kati 2 Phase I at 100 MW, Kati 2 in full at 350 MW and Dorothy 3 at 300 MW. The draft requires financial security of $50,000 per megawatt of requested peak demand plus a study fee of at least $100,000: roughly $5 million for Kati 2 Phase I alone, roughly $17.5 million for the full 350 MW. Against the $113.4 million of unrestricted cash filed at June 30, 2026, that is a cost rather than the liquidity threat it was when this document could only reconstruct roughly $8 million. It is a cost that scales with the pipeline: at $50,000 per megawatt the full 350 MW of Kati 2 is roughly $17.5 million, and the same rule reaches every other project sized at 75 MW or more.
Siting risk is no longer theoretical anywhere. On July 14, 2026 New York's governor imposed a moratorium on new large data center development statewide, citing utility bills and resource use. There is no New York exposure here, but the precedent is that a jurisdiction can close to new capacity quickly.
Clear final rules let developers plan and price with certainty, and high entry costs disproportionately screen out weaker speculative entrants from an interconnection queue that grew from 63 GW in December 2024 to roughly 410 GW by March 2026. Existing energised sites and owned generation are relative advantages in a rationed queue.
The security posting is a direct cash demand at the wrong point in the liquidity cycle, and must generally be funded before a tenant lease produces revenue. Higher costs and longer study timelines push Kati 2 and Dorothy 3 revenue further right. New large loads interconnecting after December 31, 2025 also face mandatory curtailment protocols and remote-disconnection equipment requirements.
Source: Published analyses of draft 16 Texas Administrative Code Section 25.194, 17 Mar and May 2026; Senate Bill 6 analyses. Open
Prior-year precedent: the FY2025 Form 10-K was filed late March 2026. KPMG was appointed for the year ending December 31, 2026, so this is its first opinion. The highest-information single document of the year, and unusually loaded. Three things resolve at once: whether KPMG's first audit surfaces anything its predecessor did not; whether the FY2025 material weakness covering debt classification, lease classification and valuation, and overstated equipment deposits is remediated; and what the going-concern language says. Audits for FY2022 through FY2024 carried going-concern explanatory paragraphs, and after a year of heavy cash consumption that question is live again.
A clean opinion with no going-concern paragraph and the material weakness declared remediated is a real if unglamorous upgrade in the quality of the reported numbers, and a brand-name audit lends credibility with the project lenders and tenants the company needs.
A new auditor arriving after a disclosed material weakness is the classic setting for restatements, additional weaknesses or reclassifications. A return of going-concern language would be damaging beyond the accounting: for a company selling twenty-year leases, an auditor's doubt about twelve-month viability is a commercial problem.
Source: FY2025 Form 10-K, Item 9A; 8-K and monthly update, Mar to Apr 2026 (KPMG appointment); auditor consents referencing going-concern paragraphs for FY2022 to FY2023. Filed
Precedents: FY2025 results released March 30, 2026 alongside the 10-K, and Q1 2026 results May 15 to 18, 2026. Neither announced. The full-year release is where management historically frames the strategic narrative: the FY2025 version introduced the 4.3 GW pipeline figure and the $142 million raised. The FY2026 equivalent will either announce contracted AI capacity or explain its absence after a full year of Kati 2 development. The following quarter then shows whether the December quarter's post-curtailment recovery held.
A full year of revenue growth, adjusted EBITDA at or near breakeven with Briscoe and Kati 1 fully contributing, and contracted AI capacity to point to. The December quarter should benefit mechanically from the end of summer curtailment.
A second consecutive annual report with zero HPC revenue would be difficult to reconcile with a pipeline-based valuation. Watch the pipeline figure itself. It grew from 4.3 GW to approximately 6.3 GW at the Q2 report, and operating capacity grew too, from 171 MW to approximately 192 MW, so the unfavourable version of this test was not what happened in the June quarter. The test still binds on the annual figure: a pipeline that grows while operating capacity does not is a signal about what management is measuring. Continued heavy stock compensation and a much larger share count would confirm the dilution trajectory.
Source: FY2025 results release, Mar 30, 2026; Q1 2026 release, May 18, 2026, and Form 10-Q filed May 15, 2026. Filed
Contractual. Issued April 15, 2026 to YA II PN under a Securities Purchase Agreement as part-funding for the $16.5 million purchase of Spring Lane Capital's 85.4% Class B interest in Project Dorothy 1A, unsecured and carrying a 2,400,000 share warrant. A $12 million cash obligation, which the $113.4 million of unrestricted cash filed at June 30, 2026 covers comfortably, where the reconstruction this document previously carried of roughly $8 million did not. It must be repaid, refinanced or converted. It also carries a redemption trigger: if the company closes a financing, or series of financings within thirty days, exceeding $20 million in gross proceeds, 20% of principal plus accrued interest falls due, with the at-the-market program, both standby facilities and pre-existing warrant exercises carved out. A large parent-level raise to fund Kati 2 therefore triggers a cash call unless waived.
Repaid from operating cash flow or refinanced into cheaper project debt, demonstrating that the Dorothy 1A consolidation was self-funding. An early partial redemption from a large financing would show the note being retired ahead of schedule.
Repayment funded by equity issuance at a depressed price, or conversion into stock. Combined with the Green Cloud maturity, roughly $19 million of debt matures within about thirteen months, a meaningful call on a company generating negative free cash flow near $50 million a year.
Source: Press release, Apr 16, 2026; 8-K, Apr 17, 2026 (membership interest purchase agreement and note terms). Filed
Following the annual reconstitution calendar, with rank day typically in late April and implementation in late June. Soluna was added to the Russell 3000 and Russell 2000 Value indices in the June 2026 reconstitution. Mechanical passive flows in either direction, and a useful market-capitalization checkpoint. One caution on the mechanics: heavy share issuance can support index eligibility by raising market capitalization even as per-share value falls, so retention is not straightforwardly good news.
Retention, or promotion within the index family if market capitalization rises on a signed AI lease. Membership modestly widens the institutional buyer base from the current roughly 8%.
Deletion if market capitalization falls below the eligibility threshold, producing forced passive selling into an already high-short-interest, high-beta name. The effect is amplified in micro caps where passive holdings are large relative to average daily volume.
Source: Monthly business update, Jul 14, 2026 (June 2026 index addition); reconstitution calendar published by FTSE Russell, the index provider. Filed
Contractual. Originally a $12.5 million secured promissory note issued June 20, 2024 by a Soluna Cloud subsidiary at 9% interest, later modified; carrying value approximately $7.07 million at March 31, 2026, of which $5.49 million was current. Amortising through the period, so the residual at maturity should be smaller than current carrying value, but it is secured and the investor holds a lien on substantially all Company property. The company became a direct co-obligor under a March 2025 modification, making this a parent-level obligation rather than merely a subsidiary one.
Amortised down and retired on schedule, releasing the lien on Company property, which would materially simplify the collateral picture for future financings. The modification also contemplated penny warrants issuable when the investor removes its lien, so resolution would clarify a small dilution item.
The lien on substantially all Company property constrains raising secured debt elsewhere. The note is convertible into up to 2,500,000 shares at a $5.00 conversion price, far out of the money at $1.35, which is itself a reminder of how far the share price has fallen relative to terms creditors once accepted.
Source: Form 10-Q at Mar 31, 2026, Note 8. Filed
Ongoing and undated, most material first. These take “Why undated” in place of timing confidence. Six of the twelve were carried as dated catalysts under the previous numbering despite having no date; the concordance in Methodology maps them.
As of the July 14, 2026 update a letter of intent is signed with one potential tenant, the focus is finalising design, commercial terms and the lease agreement, detailed design is roughly 50% complete and a general contractor has been selected. No target date for a binding lease has been published, and that is the date this item turns on. The company has published a target for the campus itself, and this calendar did not carry it. Its investor presentation, its pipeline page stamped “as of August 1, 2026” and its cover carrying no date states, at page 23, that Kati 2 is planned as a two-phase campus with approximately 100 MW of critical IT load targeted for service in 2027 and at least 200 MW more in 2028. Estimate A target for energising capacity is not a commitment to a signed tenant, so this item stays open and stays undated: what the deck supplies is the schedule the build is aiming at, not a lease date. This is the single largest value driver and the event nearly everything else is downstream of. A binding lease would move the company from the announcement category into the execution category the market has been rewarding, and would unlock project financing that is currently theoretical.
A signed long-dated lease with a named investment-grade or hyperscaler-backstopped tenant for Phase I's 100 MW. On sector precedent, peers disclosed contracted revenue figures at or above their market capitalizations, so a comparable disclosure against a $192 million market capitalization would be a step-change rather than an increment. One caution on magnitude: through July 2026 the peer group derated 25% to 38% despite signed contracts, with one operator falling about 37% in a month having announced a $3.4 billion NVIDIA deal and another selling off after a 61% revenue beat. Signing may be necessary without currently being sufficient.
A letter of intent is ordinarily non-binding: it establishes intent to negotiate, not an obligation to lease. It can lapse quietly, with no announcement and no disclosure obligation, so silence is the most likely form of bad news. A competitor with a completed facility can take the tenant, and this would be a first-time AI landlord negotiating against counterparties who have signed with operators of far greater scale.
Source: Monthly business updates, 9 Jun and Jul 14, 2026. Open
The company disclosed in its June 9, 2026 update that it had engaged an investment bank experienced in financing AI infrastructure to lead capital raising for Kati 2. Project financing of this type ordinarily follows a signed lease rather than preceding it. This determines whether Kati 2 happens and how much of it SLNH shareholders own. A 100 MW critical-IT data center typically costs well over a billion dollars, many times the market capitalization. The joint venture already has Metrobloks holding all Class B interests, with the distribution waterfall returning capital plus a 14% internal rate of return plus $100,000 per Gross PPA MW before a 50/50 split.
Financing closed at the project level with a meaningful promoted interest and development fees retained, the structure that made this model work for peers. A named institutional partner would validate the asset independently of the tenant.
Conditions tightened materially in July 2026: a major peer was reported on 29 July to be seeking a higher interest rate on a $2.6 billion loan tied to Anthropic capacity, and credit-default-swap costs for technology borrowers have been rising. Two failure modes follow. Financing does not close and Kati 2 stalls; or it closes on terms leaving a small residual interest, in which case the 350 MW headline overstates what accrues to shareholders, a dynamic already visible in the $54.5 million of non-controlling interests at June 30, 2026, which is down from $65.8 million only because the company bought out the Dorothy 1 minorities. A large parent-level component would also trip the 20% redemption clause on the Dorothy 1A note.
Source: Monthly business update, Jun 9, 2026; 8-K, ~Jun 9, 2026 (joint venture agreement and waterfall); Form 10-Q at Mar 31, 2026; 8-K, Apr 17, 2026. Filed
Four live instruments: the at-the-market offering agreement with H.C. Wainwright dated April 29, 2025; standby equity purchase agreements with YA II PN dated August 12, 2024 and March 24, 2026; and a shelf registration filed March 9, 2026 registering up to $1.0 billion of securities, of which $500 million is carved out for an at-the-market sales agreement prospectus. Registered capacity is roughly three times the entire market capitalization, down from five times only because the capitalization rose. Shares went from 10.6 million at end-2024 to 102.6 million at end-2025 to 157,747,354 by May 12, 2026, to 225,821,479 at June 30 and 244,590,575 by August 10, 2026: the count has risen 55% since the figure this document carried until the Q2 filing, and the June quarter alone added more shares than the six weeks that funded the acquisitions.
Issuance into strength, at prices well above current levels and funding accretive project equity, is how this model is supposed to work. The 2025 raises did fund real assets: Dorothy 2, Kati 1 and ultimately Briscoe.
Issuance at $1.35, and the June quarter shows it happening: roughly $113.5 million of at-the-market proceeds in the six months to June 30 and about $23.6 million more after it. The facilities are standing and pre-negotiated, so issuance requires no new shareholder approval and can occur at any time without advance notice. This is why dilution belongs in the base case rather than in a risk scenario.
Source: Form 10-Q at Mar 31, 2026; Form S-3 filed Mar 9, 2026 and amended Mar 30, 2026; Form 424B5, Apr 1, 2026; Form 424B3, ~May 12, 2026. Filed
4,920,045 shares at a $25.00 liquidation preference and a 9.0% cumulative rate of $2.25 per share annually. Arrears accrue at approximately $11.1 million a year, about $2.8 million a quarter, whether or not dividends are declared. Disclosed at approximately $32.4 million at March 31, 2026, and the Q2 filing puts it at approximately $35.1 million at June 30, 2026, which is what that run rate projected; roughly $38 million by 30 September on the same basis. Combined with the $123.0 million preference, a claim of roughly $158 million ranks ahead of a $330.2 million common market capitalization, about 48% of the equity beneath it. That share fell because the common grew, not because the claim shrank: it was 81% before the June quarter's issuance.
The preferred trades near 45% of face, so a negotiated settlement well below the stated claim is plausible and would be accretive relative to the headline figure. Removing the overhang would clarify the equity story considerably. That level is independently corroborated: at the 31 July close one quote provider showed a market capitalization of $247.74 million against 157.75 million shares at $1.22, a gap implying about $11.24 per preferred share.
The claim grows quarterly and silently, and no equity screening metric reflects it. Any settlement in stock at a depressed price would be severely dilutive. The two-thirds consent requirement for senior issuance also constrains financing structures.
Source: Form 10-Q at Mar 31, 2026; Series A prospectus supplement; Forms 4, Apr to Jun 2026. Filed
The Credit Agreement requires a minimum trailing debt service coverage ratio of 1.60:1.00 and a minimum forward contracted ratio of 1.20:1.00, secured by first-priority liens on substantially all borrower and guarantor assets. A mandatory cash sweep applies a percentage of free cash flow to principal, and the current versus long-term debt split depends on management's own forecast of that sweep. Separately, the contract termination liability to Hewlett Packard Enterprise stood at $19.348 million at March 31, 2026 and is unchanged at $19.348 million at June 30, 2026, with no payments made since termination in March 2025. The June 30 covenant test did not happen. Amendment No. 3 of August 7, 2026 provided that neither ratio was required to be tested at that measurement date, conditional on prepaying the Tranche A-1 and A-3 loans in full, which occurred on August 10, 2026. The entire outstanding principal is classified as current as a result of the covenant matters, and Briscoe is first tested on September 30, 2026.
The June 30, 2026 test was not passed, it was not taken: Amendment No. 3 of August 7, 2026 removed the requirement to test either ratio at that date, conditional on a prepayment made on August 10, so no default arose. That is a lender still willing to amend. Rising contracted hosting revenue supports the forward coverage test. A negotiated settlement of the termination liability below carrying value would release an obligation the market largely ignores.
Coverage covenants tied to project cash flow are the transmission mechanism that turns a weak operating quarter into a financing problem. A limited waiver was already obtained in connection with diligence requests. The termination liability is unpaid and could crystallise through litigation or demand at any time, with no scheduled date, which is why it sits here rather than among the dated items.
Source: Form 10-Q at Mar 31, 2026, Note 6 and Note 8. Filed
Roughly 23% of Q1 2026 revenue was direct proprietary mining, exposed immediately to price and difficulty. The 71% hosting share is contractually insulated in the short term but not economically insulated: hosting customers are themselves leveraged Bitcoin operators who stop expanding, then stop paying, then fail. FY2025 revenue fell 21.8% year over year largely on weaker Bitcoin-linked economics, so the sensitivity is demonstrated rather than theoretical.
A strong tape lifts mining revenue directly, accelerates hosting expansions of the kind one customer has done four times, raises the value of existing energised capacity, and improves equity issuance terms simultaneously.
A sustained decline compresses mining revenue, stalls hosting demand, and depresses the share price used to fund the AI pivot, all at once. The current backdrop is already unfavourable: Bitcoin has traded near a 21-month low of about $58,115 and in a roughly $58,000 to $64,000 range after record spot exchange-traded fund outflows of approximately $4.5 billion in June 2026. There is no halving within this window; the next is expected around 2028.
Source: Form 10-Q at Mar 31, 2026; FY2025 results release, Mar 30, 2026; press releases Apr 21 and 28, 2026; market levels as reported Jul 2026. Market
Compliance with the $1.00 minimum bid requirement under Nasdaq Listing Rule 5550(a)(2) was regained on May 1, 2026, after maintaining a close of $1.00 or more from April 14 to 29, 2026. At the 12 August close of $1.35 the buffer is roughly 35%. Alongside: short interest of 34.41 million shares at the July 31, 2026 close, reported then as 21.8% of shares outstanding and 33.8% of float, at 1.85 days to cover. That percentage was struck on a share count the Q2 cover page has since superseded: the same 34.41 million shares are about 14% of the 244,590,575 now outstanding, and no more recent short interest reading is carried here. Also; average daily volume near 15.5 million shares; and a beta of 5.13.
Short interest at a third of float with less than two days to cover creates genuine squeeze potential on a positive surprise, a signed Kati 2 lease being the obvious trigger. High-beta micro caps can move several hundred percent on such news, as the move from $0.41 to $5.14 within twelve months demonstrates.
The same mechanics operate in reverse, and the buffer above $1.00, 35% at the 12 August close, is wider than it was but is not a cure: the stock fell 6.15% on 31 July alone, on roughly 15 million shares. A disappointing quarter plus a hawkish macro turn could put the listing requirement back in play within a single quarter, which would in turn constrain the equity issuance the company relies on.
Source: Press release, May 1, 2026; the exchange daily record for the Aug 12, 2026 close; market data at the Jul 31, 2026 close for short interest. Market
As of July 14, 2026: environmental due diligence, survey work and fiber studies in progress; schematic design and master planning begun; hiring in final stages for a Texas-based community engagement consultant. A definitive purchase agreement for 300 acres was signed by May 2026. Moderate near term and high eventually: Dorothy 3 is the purest expression of the vertical-integration thesis, 300 MW of AI and HPC capacity adjacent to Dorothy 1, Dorothy 2 and the owned Briscoe wind farm, restated from 300+ MW at the Q2 report, which also allocated Project Grace, the 2 MW technical validation effort with the Siemens PTI team, into the Dorothy 3 campus plan. It is at least two development stages behind Kati 2 and realistically a 2028 or later revenue contributor.
The process this campus has to pass through has been stopped. On August 3, 2026 the Governor of Texas directed the Public Utility Commission of Texas and ERCOT to conduct a comprehensive verification and audit of all data centers advancing through ERCOT's interconnection process, and stated that any project failing to comply will be denied connection to the Texas grid. ERCOT suspended the Batch Zero Large Load classification notifications due by August 7, 2026 in market notice M-A080326-01, gives no replacement date, and filed its request for good cause exceptions on August 10, 2026, in which it states the verification will take several months. Dorothy 3 is early enough that a pause costs schedule rather than a queue position, and it is also the site most exposed to the co-location and net-metering rule before the same Commission, so the audit and that rulemaking now sit with one regulator at one time. Whether this campus falls inside the audit's scope is not established here. The Form 10-Q filed August 13, 2026 does not mention the directive, the notice or the Commission at all.
Interconnection progress, a power purchase or firming arrangement, or an anchor tenant discussion disclosed earlier than the stage would suggest. Community and county support secured smoothly; the consultant hire suggests management treats local approval as a real gate.
Environmental, fiber or interconnection findings that impair the site. Local opposition, which has delayed data center projects across multiple jurisdictions. Directly exposed to the co-location rule, which is no longer a pending rulemaking: it was adopted on March 26, 2026 as 16 TAC 25.205 and now governs any net-metering arrangement between this campus and a co-located generation resource, subject to an ERCOT study and a Commission approval that may carry conditions. Pre-development spending continues with no revenue against a constrained cash position.
Source: Monthly business updates, 9 Jun and Jul 14, 2026; Q1 release, May 18, 2026. Filed
As of the Q2 report of August 13, 2026: capacity was redesignated for AI workloads at Rosa, 187 MW to 242 MW, Hedy, 120 MW to 198 MW, Ellen, 100 MW to 145 MW, and Fei, 120 MW to 240 MW, with Hedy, Ellen and Fei advancing through power term sheets to a combined 583 MW aimed at AI and HPC. Annie is sized at 74 MW with capital formation and construction design commenced. Project Gladys is new to this file, 150 MW co-located with a 226 MW wind farm, and the Form 10-Q places both it and Fei in land acquisition, power contract negotiation and ERCOT interconnection planning. Grace has been allocated to Dorothy 3 and leaves this list. These projects constitute most of the pipeline figure restated at approximately 6.3 GW that supports the valuation premium, and interconnection study completion is the nearest observable proof point that the pipeline is more than a list of sites.
The process these projects are advancing through has been stopped. On August 3, 2026 the Governor of Texas directed the Public Utility Commission of Texas and ERCOT to conduct a comprehensive verification and audit of all data centers advancing through ERCOT's interconnection process, and stated that any project failing to comply will be denied connection to the Texas grid. ERCOT suspended the Batch Zero Large Load classification notifications due by August 7, 2026 in market notice M-A080326-01, gives no replacement date, and filed its request for good cause exceptions on August 10, 2026, in which it states the verification will take several months. No completion deadline for the audit has been named. Whether these particular projects fall inside the audit's scope is not established here. The Form 10-Q filed August 13, 2026 does not mention the directive, the notice or the Commission at all.
Studies completed and agreements executed, converting pipeline megawatts into queue positions with defined capacity and cost, which is the point at which pipeline starts to mean something specific. A new project-level capital partner, as secured twice in 2025, would validate the model again. Geographic diversification beyond Texas would reduce the concentration in one grid and one rulemaking program.
Studies returning poor results or high upgrade costs. Land agreements not closing. Each project sized at 75 MW or above also faces the $50,000 per megawatt security requirement. Pre-development capital continues to be consumed with no revenue. The risk here is quiet attrition: projects fading from the update list rather than being formally canceled.
Source: Monthly business update, Jul 14, 2026; FY2025 results release, Mar 30, 2026. Filed
No announced upcoming conference appearance at the date of this document. Recent precedent: a virtual fireside chat at an investor research conference on April 14 to 15, 2026, plus a recurring investor question and answer format with the chief financial officer. Low individually, modestly relevant collectively, because institutional ownership near 8% against fresh Russell 3000 and Russell 2000 Value membership leaves room for a sponsorship-building effort. The 13 August earnings call is the more consequential version of this.
Appearances at data-center or power-infrastructure conferences rather than crypto events would support the repositioning from Bitcoin miner to infrastructure developer. New sell-side coverage from a firm without a banking relationship would matter considerably more.
Minimal direct downside. The indirect signal is that promotional activity outpacing contracted progress tends to read poorly, and disclosure has leaned on magazine covers, index inclusion and media features.
Source: Press release, Apr 10, 2026; monthly business updates, 9 Jun and Jul 14, 2026. Open
Recent releases: 10 March, 9 April, 7 May, 9 June and July 14, 2026. The cadence matters more than a routine disclosure normally would, because it is the only regular channel through which Kati 2 tenant progress becomes visible. The June and July updates are where the letter of intent and the general contractor selection were disclosed.
A tenant named, a letter of intent converted, a financing mandate closed, or a new capital partner disclosed. Historically these have appeared here first rather than in a standalone release.
Kati 2 language repeats unchanged month over month, which is how a stalling negotiation reads in practice. Watch for the disappearance of the letter of intent reference: an omission, not a statement, would be the signal.
Source: Company monthly business updates, March to July 2026. Filed
Closed items, kept for the record, ordered by the date each completed rather than the date it began. A resolved catalyst takes the next free R number; it does not carry a previous number over. Note what this list contains and what it does not: asset acquisitions, vertical integration, a listing cure and capital-structure cleanup, but not one AI or HPC revenue contract.
Adopted, and this document carried it as possibly still pending for nearly five months. The Commission signed the Order Adopting New 16 TAC 25.205 on March 26, 2026, at item 66 of Project 58479, with the Texas Register acknowledgement filed the same day. This card previously read that the status "was not confirmable at the date of this document, so this may already be adopted". It was confirmable, on the Commission's own public docket, and it had been adopted. The rule requires Commission approval of net-metering arrangements between a large load customer and an existing generation resource in the ERCOT region, with ERCOT to study a proposed arrangement within 120 days and the Commission to approve, approve with conditions, or deny within 60 days of the study being filed.
The curtailment question is answered, and the answer is neither the one this card feared nor the one the falsification test expected. Adopted 25.205(k)(1) imposes a mandatory obligation, and it falls on the generation resource rather than on the load: where the Commission approves an arrangement with conditions on a resource that was making dispatchable capacity available beforehand, the conditions must require it to keep making at least that much available as ERCOT directs ahead of an anticipated emergency. Behind-the-meter load reduction is permissive: 25.205(k)(2) says the Commission may impose conditions including requiring retail customers served behind the meter to reduce load during certain events, and may set ramp rate limits or a maximum duration for load curtailment. The order's own reasoning closes the gap the text leaves open: the Commission states that a resource's full capacity "is made available first and foremost through the full curtailment of the large load customer that is co-located with the existing generation resource". So the rule does not command curtailment and its mandatory limb is satisfied by it.
Whether it binds this company is not established, and the test is specific enough to be settled by someone with the registration facts. The rule reaches an existing generation resource, defined as one registered with ERCOT as stand-alone as of September 1, 2025. Two exemptions apply: where the modeled facility already included a co-located large load at the generator's energisation, and where a majority interest was owned as of January 1, 2025 by a parent of the participating large load customer. The second cannot help here, because Soluna acquired Briscoe on April 1, 2026 and held nothing in it at the start of 2025. On the first, ERCOT filed its list of stand-alone generation resources at item 15 of this docket on September 19, 2025 and Briscoe does not appear on it under any spelling this project could find, which would point toward the exemption applying. That reading is not adopted here: the extraction of a 25-page tabular filing is lossy and an absence read off it is weaker evidence than a presence.
What resolves is the adoption, not the exposure. The rule is now law and its application to this company is a live, undated question, which is why it sits with the Dorothy 3 development milestones rather than here. Approval requirements add timeline and cost to Dorothy 3, and a mandatory curtailment condition arriving through the generation limb would undercut the uptime guarantees an AI tenant requires, which is the core objection to this company as an AI landlord.
Source: Public Utility Commission of Texas, Project 58479, item 66, Order Adopting New 16 TAC 25.205, March 26, 2026, and item 15, ERCOT stand-alone generation resource list, September 19, 2025, both read from the Commission's Interchange on August 13, 2026. Filed
Acquired 100% of the equity in Briscoe Wind Farm, LLC, an approximately 149.85 MW nameplate wind project co-located with the Dorothy campus in West Texas, for $53.0 million, funded with cash plus $12.5 million of new Tranche C debt. Treated as an asset acquisition with $5.0 million of transaction costs capitalized. Management projected $6 million to $11 million of year-one adjusted EBITDA on $20 million to $24.4 million of annualised revenue. The 8-K/A of June 17, 2026 put total cash consideration at approximately $55.9 million and disclosed the asset's 2025 revenue of $8,950,927 against a net loss of $14,773,838.
Creates the vertical integration underpinning Dorothy 3 and the contracted-uptime pitch to AI tenants. It is also the largest single cash outflow in the company's history and the main driver of the liquidity question at the August results.
Source: 8-K and press release, Apr 2, 2026; Form 10-Q Note 16; 8-K/A with audited and pro forma financials, Jun 17, 2026. Filed
Compliance with Nasdaq Listing Rule 5550(a)(2) regained after maintaining a minimum closing bid of $1.00 or greater from 14 April through April 29, 2026. The exchange closed the matter and the company stated it was in full compliance with all continued listing requirements. This removed a delisting overhang that had constrained institutional participation through 2025, when the stock traded largely below $1.00. The cure is not permanent: at the 12 August close of $1.35 the buffer is roughly 35%, and the requirement remains a standing condition.
Source: Press release, May 1, 2026. Filed
Two transactions. Dorothy 1A: acquired Spring Lane Capital's 85.4% Class B interest for $16.5 million, closing April 15, 2026, with $6.0 million paid at closing and $10.5 million due no later than July 1, 2026, part-funded by a $12 million note issued to YA II PN under a Securities Purchase Agreement dated April 15, 2026, maturing May 15, 2027 and carrying a 2,400,000 share warrant. Dorothy 1B: acquired the remaining 49% membership interest from Navitas West Texas Investments for $8,765,490, closed May 19, 2026 and funded with balance sheet cash.
Together these complete equity consolidation of the 50 MW Dorothy 1 campus, increase the shareholders' share of Dorothy economics and clear title for the Dorothy 3 development. They leave the $12 million maturity in May 2027 and the financing-triggered 20% redemption clause. The completion date is the date the second transaction closed, not the date the first began.
Source: Press release, Apr 16, 2026 and 8-K, Apr 17, 2026; Form S-3, Apr 2026; 8-K and Exhibit 99.1, May 20, 2026; Form 424B3, ~May 12, 2026. Filed
A definitive limited liability company agreement replaced the earlier non-binding memorandum of understanding, governing Soluna MB KK II JVCo, LLC between Soluna HPC KK II HoldCo, holding all Class A interests and acting as manager, and DC Kati Venture LLC, managed by Metrobloks, holding all Class B interests. Soluna contributed the Phase I property, a purchase agreement for the Phase II property and approximately $3.5 million of operating expenses, with total commitments reported around $21 million to $22.5 million including roughly $19 million for the Phase II property. Phase I is 100 MW critical IT; Phase II adds 250 MW. The company's own presentation totals the campus differently. Page 23 is headed “300MW+ Tier III AI Campus Plan” and describes roughly 100 MW in 2027 with “at least 200 MW more” in 2028, against the 350 MW the split above implies. Estimate Not reconciled here: the figures above come from the filed agreement and the deck's wording is open-ended, so “at least 200 MW more” does not exclude 250. The distribution waterfall returns capital plus a 14% internal rate of return plus $100,000 per Gross PPA MW, then splits 50/50.
Converting a memorandum to a definitive agreement is real progress on the partner side. It is not a tenant, a lease, or financing, which remain open. This is the vehicle, not the revenue.
Source: 8-K, ~Jun 9, 2026 (agreement effective Jun 3, 2026); monthly update, Jun 9, 2026. Filed
All 62,500 outstanding Series B Convertible Preferred shares were converted into 6,510,416 common shares during 2026, with $2.1 million of accrued dividends paid at full conversion. A Withdrawal of Designation was filed in Nevada on June 23, 2026, terminating the designation and removing the provisions from the Articles of Incorporation. Only common stock and the Series A preferred now remain.
A genuine simplification that removed roughly $1.8 million of Series B arrears. But it was achieved by moving the claim into the common share count: 6.5 million new shares plus $2.1 million of cash. The far larger Series A layer, at roughly $158 million including arrears at June 30, 2026, is untouched.
Source: 8-K, Jun 23 to 25, 2026; Form 10-Q at Mar 31, 2026. Filed
Michael Picchi appointed Chief Financial Officer effective April 1, 2026, previously chief financial and accounting officer at a data center developer. Ryan Carver, formerly of Microsoft, appointed Chief Development Officer in July 2026 to lead the AI and HPC data center platform lifecycle. KPMG appointed independent registered public accounting firm for FY2026, replacing UHY LLP, which audited FY2025 and FY2024. Notably, Picchi purchased 100,000 common shares in the open market at $1.63 on May 22, 2026.
Relevant credibility for the tenant negotiations and the financing that follow: a former hyperscaler executive is a meaningful hire for a company trying to lease to hyperscalers. The open-market purchase is the most substantive insider signal of the period. The first KPMG opinion arrives with the FY2026 annual report.
Source: Monthly update, Feb 2026; press release, Jul 16, 2026; Q1 release, May 18, 2026; Form 4, May 22, 2026. Filed
This card named six things that would be disclosed for the first time. All six were, and the answers did not all run the same way. Post-acquisition liquidity: $113.4 million of unrestricted cash and $33.1 million of carrying debt at June 30, 2026, so net cash of $80.2 million where this document had reconstructed roughly $8 million from disclosed deal terms and expected net debt. That estimate was wrong by about $105 million, and the correction is recorded in the log rather than quietly absorbed. The share count at a recent date: 244,590,575 at August 10, 2026, against 157,747,354 at May 12. Briscoe's contribution: wind energy generation revenue of $366 thousand against $2.253 million of cost before depreciation, with $1.5 million of maintenance on gearboxes and main bearings, so the wind farm reduced gross profit in its first full quarter rather than lifting it. The sequential revenue streak: it held, to a fifth consecutive quarter.
The other two answers were negative. The material weakness was not remediated: management concluded that disclosure controls and procedures were not effective at June 30, 2026, on the same balance sheet classification and presentation weakness carried since the FY2025 annual report. Covenant compliance was not demonstrated either, it was waived: on August 7, 2026 Amendment No. 3 to the Generate credit agreement provided that neither the Debt Service Coverage Ratio nor the Forward Contracted DSCR was required to be tested at the June 30, 2026 measurement date, conditional on prepayment in full of the Tranche A-1 and A-3 loans, which occurred on August 10, 2026. The entire outstanding principal is classified as current as a result of the covenant matters. Briscoe is first tested on September 30, 2026.
Liquidity and dilution were one question, and the filing settles which side it fell. This document's own assumption was that a comfortable cash balance would be reassuring only in combination with a share count that had not moved much to produce it. The cash flow statement shows $113.5 million of at-the-market proceeds in the six months, and the count rose 55%. The cash is there because the shares were sold, which is the framing this file carried rather than a departure from it.
Source: Press release, Aug 13, 2026; Form 10-Q for the quarter ended June 30, 2026, filed Aug 13, 2026, cover page, Notes on debt and equity, and Item 4. Filed
Every category reviewed appears as a row, including those that produced catalysts, which read across by title and ID. A category checked and found empty is recorded rather than omitted.
| Category | Status | Basis |
|---|---|---|
| Earnings | Covered above | Q2 2026 results and first quarterly earnings call (R8), resolved on August 13, 2026, Q3 2026 results and Form 10-Q (D11), FY2026 Form 10-K and first KPMG audit opinion (D14), FY2026 results and Q1 2027 results (D15). |
| Product launches | None | Checked against company press releases and monthly business updates through July 30, 2026 and found empty. The category does not apply to an infrastructure business with no product cycle: the nearest equivalents are capacity energisations, carried as Kati 1B Phase 3 completion (D5), and the MaestroOS control software, which is internal tooling rather than a sold product. |
| Investor days | None | Checked against company press releases and monthly business updates through July 30, 2026 and found empty. No capital markets day is announced. For a company asking the market to underwrite a pipeline restated at approximately 6.3 GW that absence is itself notable. The first earnings call was held on August 13, 2026, which is a first step toward closing it and not a capital markets day. |
| Regulatory decisions | Covered above | PUCT transmission cost allocation review (D12), PUCT co-location and net-metering rule (R1), resolved on adoption, PUCT large-load interconnection standard (D13). |
| Lawsuits | Not established | Never examined. No litigation search was performed and no legal proceedings disclosure was checked. Also named in what was not checked, and carried forward in Methodology. |
| Macro events | Covered above | Jackson Hole Economic Policy Symposium (D4), FOMC September decision (D6), FOMC October and December decisions (D8). |
| Industry conferences | None | Checked against company press releases and the investor relations events page through July 30, 2026 and found empty. No appearance is announced. Carried as a standing condition, investor conference appearances and investor relations outreach (S12), because appearances have historically been confirmed only days ahead. |
| Management changes | Covered above | Management build-out (R7): chief financial officer effective April 1, 2026, chief development officer July 2026, auditor changed to KPMG for FY2026. |
| Buybacks | Not established | Never examined. The company is a structural net issuer with $1.0bn of registered capacity, which makes a repurchase improbable, but improbable is not the same as checked. Also named in what was not checked. |
| Dividends | Not established | Never examined for a common dividend. Series B preferred dividends of $2.1m paid on conversion and Series A arrears of roughly $35.1m at June 30, 2026 accruing unpaid are both recorded above. Also named in what was not checked. |
| Major contracts | Covered above | Kati 2 letter of intent converts to a binding lease (S1), carried as standing because it has no date. The most consequential item in the document, and it has not occurred. |
| Signed AI or HPC lease | None | Checked against the Form 10-Q for the quarter ended June 30, 2026 and the results release of August 13, 2026, and found empty: zero HPC revenue again this quarter and one non-binding tenant letter of intent, against peers holding multi-billion-dollar contracted backlogs. The Kati 2 joint venture signed June 3, 2026 is a development agreement with Metrobloks, not a tenant lease, and is not counted here as one. |
| Bitcoin halving | None | Checked against the protocol issuance schedule and found none within the twelve-month horizon; the next is expected around 2028. This removes one recurring sector catalyst that would otherwise appear. |
| Credit rating action | Not established | Never examined. No rating agency was checked, so the absence of a rating here is a gap rather than a finding. Also named in what was not checked. |
| Index events | Covered above | Russell 3000 and Russell 2000 Value addition completed June 2026; the forward event is retention or deletion (D17). |
| Company sale or takeover | None | Checked against the 8-K current reports listed in Sources through July 30, 2026 and found empty. No process or approach is disclosed. The $158m preferred claim and the variable interest entity structure would complicate any acquisition of the parent. |
| Stock split | None | Checked against the 8-K current reports listed in Sources through July 30, 2026 and found empty. A 1-for-25 reverse split occurred October 16, 2023, relevant only as precedent should the minimum bid requirement come back into play. |
| Series B preferred events | Covered above | Class eliminated (R6) on June 23, 2026. No further conversions or dividends are possible from this class. |
Sits across every other catalyst rather than beside them. No item here is meaningful in per-share terms without tracking the denominator.
Loss per share improved from $(0.93) to $(0.18) year over year in the June quarter while the absolute loss grew, because the share count grew faster than the loss. The same happened in the March quarter, $(1.21) to $(0.24).
| Date | Shares outstanding | Driver and basis |
|---|---|---|
| Dec 31, 2024 | 10,607,020 | Post 1-for-25 reverse split baseline. Filed |
| Dec 31, 2025 | 102,531,089 | At-the-market settlements, warrant exercises, standby facility draws, restricted stock, July and December offerings. Filed |
| Mar 31, 2026 | 111,717,040 | 8.15m warrant exercises, 1.37m restricted stock, partial Series B conversion, and 335,976 commitment fee shares. Filed |
| Apr 6, 2026 | 113,181,690 | Plus 1.5m. The same prospectus reported a last sale price of $0.9259. Filed |
| Apr 24, 2026 | 141,347,055 | Plus 28.2m in the eighteen days from 6 April, bracketing the Briscoe and Dorothy 1A closings. Filed Form 424B3 |
| May 12, 2026 | 157,747,354 | Plus 16.4m more; roughly 46m added in six weeks, about 41% of the starting count. Filed 10-Q cover |
| Jun 30, 2026 | 225,821,479 | Plus 68.1m more across the quarter; 225,986,784 issued less 165,305 in treasury. Filed 10-Q balance sheet |
| Aug 10, 2026 | 244,590,575 | Plus 18.8m in six weeks, on roughly $23.6m raised under the at-the-market program after quarter end. Filed 10-Q cover |
| Authorised | 375,000,000 | Increased from 75m at the special meeting of Nov 7, 2025; roughly 130m of headroom remains after the Q2 issuance. Filed |
| Instrument | Capacity | Basis |
|---|---|---|
| Shelf registration, total | $1.0bn | Filed Form S-3, Mar 9, 2026, amended Mar 30, 2026 |
| of which ATM sales agreement prospectus | $500m | Filed Form 424B5, Apr 1, 2026. Supersedes an earlier 2025 filing of $87.65m |
| Resale registrations, April 2026 | 31.66m sh | Filed 26,512,815 under the 2024 standby agreement; 2,459,400 warrant and Harmattan merger shares; 2,700,000 Generate warrant shares. Resale only, producing no cash for the company |
| Warrants outstanding | 27,549,840 sh | Filed Form 424B3 |
| Generate warrant package | 2,700,000 sh | Filed a pre-funded warrant for 700,000 shares plus common warrants for up to 2,000,000 more, being 1,350,000 and 650,000, issued with the $12.5m Tranche C loan |
| Placement agent warrants | $1.313m | Filed carried in mezzanine equity |
| Green Cloud conversion | 2,500,000 sh | Filed at $5.00, far out of the money |
| June 2026 restricted stock grants | 12,631,694 sh | Filed five insiders, about 8% of shares outstanding |
| Company ATM illustration | 760,178,588 sh | Filed shares issuable were the ATM fully sold at $0.77 |
These catalysts are correlated, not independent. Almost every forward item is downstream of one event, and the sequence runs: liquidity survives the acquisitions, the letter of intent becomes a binding lease, project financing closes, construction and energisation follow, and only then does AI revenue appear. Counting the items as separate shots on goal overstates the number of ways this resolves well.
The first gate is easy to miss, and it is not really a cash gate. Pre-development spending, the $50,000 per megawatt security deposit contemplated by the proposed interconnection rule, and ordinary operating burn all require cash. The Q2 filing settles how much there is, and this document had it badly wrong: $113.4 million of unrestricted cash at June 30, 2026, against a reconstruction from disclosed deal terms of roughly $8 million. The reconstruction netted the acquisition outflows against the opening balance and did not model the issuance that funded them. But between 31 March and May 12, 2026 the share count went from 111,717,040 to 157,747,354, roughly 46 million shares, about 41% of the starting count, in six weeks. That is how the acquisitions were funded. The company does not run out of money so long as it can sell stock, and with $1.0 billion registered it has ample capacity to do so.
The question at that first gate is therefore never whether there is cash, but at what price, and how much of the eventual upside still belongs to today's shareholders. Read the August cash balance and share count together; either alone is misleading.
Working from disclosed transaction terms, the June quarter cash impact is roughly $43m for Briscoe net of the $12.5m Tranche C draw, or nearer $48m if the $5.0m of capitalized transaction costs sits outside that figure since the filing is not explicit, $4.5m net for Dorothy 1A after the $12m note, $8.8m for Dorothy 1B, and $2.1m of Series B dividends: about $59 million to $64 million out against $68.6 million at the start, implying $5 million to $10 million unrestricted before operating burn or construction spending. The Form 10-Q filed August 13, 2026 puts unrestricted cash at $113.4 million, so that implication was wrong by about $105 million. The outflows above were not the error; what the arithmetic omitted was the $113.5 million of at-the-market issuance over the same six months. With roughly $19 million of debt maturing by June 2027, a possible $5 million to $17.5 million interconnection security requirement, and $500 million of at-the-market capacity registered, further issuance remains the base case rather than a risk scenario. The question is price and size, not whether. This note said the single most useful figure on 13 August would not be revenue but the share count on the Form 10-Q cover page. That was right, and it read 244,590,575.
One structural quirk cuts the other way. The $12 million Dorothy 1A note requires redemption of 20% of principal plus accrued interest if the company closes a financing, or series of financings within thirty days, exceeding $20 million in gross proceeds, with the at-the-market program, both standby facilities and pre-existing warrant exercises carved out. A large parent-level raise to fund Kati 2 therefore triggers a cash call unless waived. Success at the financing gate has a toll attached.
What would prove this read wrong, stated in advance. Each trigger names an observable event settleable from a document: a filing, a date passing, a disclosed figure crossing a stated threshold, or a price. An unquantified threshold is not a test. All ten now carry one: eight are settleable from a filing or a date, and two rest on a stated judgment, tagged as an estimate where the threshold is not a filed figure.
| # | If this happens… | …this was wrong | Status |
|---|---|---|---|
| 1 | The Q2 2026 Form 10-Q shows unrestricted cash above $35 million and shares outstanding at or below 157,747,354 on the cover page. | The roughly $8 million liquidity estimate, for reasons other than issuance. A high cash balance with a much larger share count confirms the framing instead. | Untriggered, and settled. Cash was $113.4m, above the threshold, but the cover page shows 244,590,575 shares, so the second limb fails and the test reads as it was written to read: the estimate was wrong because of issuance. |
| 2 | An 8-K or press release discloses a binding Kati 2 lease with a named tenant before December 31, 2026. | The low timing confidence on the lease, and the ordering that places project financing, Dorothy 3 and pipeline progression behind it. | Untriggered |
| 3 | Shares outstanding on the Q2 or Q3 2026 Form 10-Q cover page are no higher than 157,747,354. | The conclusion that dilution is the base case rather than a risk scenario. | Untriggered, and settled for Q2 at 244,590,575 on the August 10, 2026 cover page, 55% above the threshold. Still live for Q3. |
| 4 | Q2 2026 total revenue as filed exceeds $9.394 million, the Q1 2026 figure reported in the Form 10-Q of May 15, 2026. | The inference that summer curtailment breaks the four-quarter sequential growth streak. | Triggered. Q2 2026 revenue as filed was $15.060m against $9.394m. The inference is withdrawn; the streak reached five quarters. |
| 5 | Consolidated revenue attributable to Briscoe reaches $5.0 million or more in a single quarter, the quarterly equivalent of the $20 million annualised projection. | The caveat that intercompany power sales eliminate on consolidation and so will not appear in reported revenue. | Untriggered |
| 6 | The PUCT adopts the co-location rule and the adopted text contains no mandatory behind-the-meter curtailment obligation. | The treatment of that rule as the most serious regulatory threat to the model. | Triggered on its own terms, and the test was drawn too narrowly. The rule was adopted March 26, 2026 and the adopted text contains no mandatory behind-the-meter curtailment obligation: 25.205(k)(2) is permissive. But the mandatory limb at (k)(1) falls on the generation resource, and the order says that capacity is made available "first and foremost through the full curtailment of the large load customer". The treatment is not withdrawn. The test asked what the text says; the exposure is in how the text is satisfied. |
| 7 | A firm other than the company's at-the-market sales agent publishes an initiation of coverage. | The implication that the bullish view is confined to a conflicted source. | Untriggered |
| 8 | The FY2026 Form 10-K carries an audit opinion with no going-concern explanatory paragraph and Item 9A reports the material weakness remediated. | The reporting-quality discount applied throughout this document. | Untriggered |
| 9 | The company announces a settlement, exchange or redemption of the Series A preferred for aggregate consideration below $77.5 million, being roughly half of the $158 million claim of $123.0 million liquidation preference plus roughly $35.1 million of arrears at June 30, 2026. The threshold stays at $77.5 million rather than tracking the arrears, because a test whose bar moves every quarter can never be settled. Estimate the threshold is a judgment anchored on the preferred trading near 45% of its $25 face, not a filed figure. | The treatment of the preferred claim as an overhang that resolves expensively for the common. | Untriggered |
| 10 | Forms 4 disclose open-market purchases of common stock by two or more insiders other than the chief financial officer, totalling 200,000 shares or more within any rolling ninety days. Estimate the count and the share threshold are judgments set against the single 100,000 share purchase of May 22, 2026, not filed figures. | The reading that insider conviction is confined to one officer and that the liquidity position is under strain. | Untriggered |
All ten tests now carry a settleable trigger. Three stood as open defects at the previous revision, one for want of a baseline this document did not hold and two for want of any threshold at all. The baseline is now sourced here from the quarterly filing, and the two judgment thresholds are stated as judgments and tagged, which is what the rule requires of a threshold that is not a filed figure. A stated threshold that turns out to be wrong is still a test; an unstated one never was.
Every ID in one table, gapless within each class, matching the cards above in count and order. Links point at title slugs so they survive renumbering.
Reading the table honestly. Three unresolved items carry Critical impact: the Kati 2 letter of intent converting to a binding lease (S1), Kati 2 Phase I project financing (S2), and the standing dilution machinery (S3). All three are the tenant lease itself, its financing, or the dilution required to fund it. Two more resolved during 2026 and sit in the resolved sequence: the PUCT co-location and net-metering rule (R1), adopted March 26, and the Q2 2026 results and first quarterly earnings call (R8), reported August 13. The regulation governing the lease structure has therefore stopped being a pending question and become a fixed condition, which concentrates the remaining Critical items further rather than reducing them. That concentration is the document's real message, and the reason the gating chain sits in the capital overlay rather than as a footnote.
Maps the numbering used before this revision to the current IDs, so earlier log entries remain resolvable. The earlier scheme used bare numbers for dated items, S for standing conditions and C for completed events. Six items it carried as dated catalysts have no date and are now standing conditions.
| Was | Now | Catalyst |
|---|---|---|
| 1 | R8 | Q2 2026 results and first quarterly earnings call |
| 2 | S1 | Kati 2 letter of intent converts to a binding lease |
| 3 | D3 | Series A preferred becomes callable |
| 4 | D2 | 2026 annual meeting of stockholders |
| 5 | D9 | Generate Tranche B draw window closes |
| 6 | D7 | ERCOT summer curtailment season ends |
| 7 | D5 | Kati 1B Phase 3 completion, 14 MW |
| 8 | D10 | Briscoe repair completion and first full quarter |
| 9 | S13 | Monthly business updates |
| 10 | D6 | FOMC September decision |
| 11 | D4 | Jackson Hole Economic Policy Symposium |
| 12 | S12 | Investor conference appearances and outreach |
| 13 | D12 | PUCT transmission cost allocation review |
| 14 | R1 | PUCT co-location and net-metering rule |
| 15 | D13 | PUCT large-load interconnection standard |
| 16 | D11 | Q3 2026 results and Form 10-Q |
| 17 | S2 | Kati 2 Phase I project financing |
| 18 | D8 | FOMC October and December decisions |
| 19 | S9 | Dorothy 3 development milestones |
| 20 | D14 | FY2026 Form 10-K and first KPMG audit opinion |
| 21 | D16 | Dorothy 1A promissory note matures |
| 22 | D18 | Green Cloud secured note matures |
| 23 | D15 | FY2026 results and Q1 2027 results |
| 24 | D17 | Russell reconstitution, retention or deletion |
| 25 | S10 | Pipeline project progression |
| 26 | S11 | Possible authorised-share increase or restructuring vote |
| S1 | S6 | Bitcoin price and network difficulty |
| S2 | S3 | Standing dilution machinery |
| S3 | S4 | Series A preferred arrears accrual |
| S4 | S5 | Generate covenants, cash sweep, and the HPE liability |
| S5 | S7 | Nasdaq listing compliance and trading mechanics |
| S6 | S8 | Customer concentration, structure, related parties |
| C1 | R2 | Briscoe Wind Farm acquired |
| C2 | R4 | Dorothy 1 campus fully consolidated |
| C3 | R5 | Kati 2 definitive joint venture signed |
| C4 | R3 | Nasdaq minimum bid compliance regained |
| C5 | R6 | Series B preferred eliminated |
| C6 | R7 | Management build-out |
Tags record where a claim came from and nothing else. They are not color-coded, because color is reserved for meaning and identity and provenance is neither. A Filed fact can be terrible news.
| Tag | What it asserts |
|---|---|
| Filed | Stated in an SEC filing or company release, cited by form and date. This also covers a dated official publication by a named non-SEC issuer, such as an exchange notice, a central bank statement or a schedule published by its organizer, with the issuer named in the source line. What the tag asserts is unchanged: a specific body published this, on this date, and the reader can go and look. It never covers a third party's characterisation of someone else's facts, which is Market. |
| Estimate | Derived or inferred here. The arithmetic is shown. |
| Open | Expected but unconfirmed. Nothing filed either way. |
| Market | Price, volume, float, short interest, and published targets and ratings, stamped with the close or publication date. A third party's target or rating is market data, attributed to the firm that issued it and never adopted here. |
| Level | Means |
|---|---|
| High | Date is company-announced, contractually fixed, or statutorily set. |
| Medium | Date inferred from filing cadence, prior-year practice, or a stated deadline window. |
| Low | Date is a judgment call and could move by a quarter or more. |
Confidence rates timing, never outcome. A High-confidence catalyst can be a coin flip; a Low-confidence one can be near-certain in direction.
Impact is independent of confidence. A High-impact, Low-confidence item is the most important kind here: it matters enormously and could land at any time. Standing conditions take a reason for being undated in place of a confidence rating, because there is no date to rate.
Dated items are ordered by expected date, earliest first, with ties broken by the previous ordering. Standing conditions are ordered by materiality. Resolved items are ordered by the date each completed, not the date it began, which is why the Dorothy 1 consolidation (R4) sorts after the Nasdaq compliance cure (R3) despite starting first: its second transaction closed on May 20, 2026 and the cure completed on May 1, 2026. Ordering the resolved class this way means a newly resolved item always lands at the end and existing numbers stay put.
Two figures were removed from the dilution and capital overlay in this revision, for two different reasons.
The count of standby facility shares already issued by 1 April, part of the breakdown of the April resale registrations, is a fact belonging to the research report's evidence rather than to this document's. It is the mirror of several figures removed from that report in the same pass, and applying the rule in one direction only would be worse than not applying it. The overlay still carries the total resale registration of 31.66m shares and its three components, so what is lost is how much of the standby tranche had already been drawn at that date.
The weighted average exercise price of the outstanding warrants is a different and more serious case: a figure with no traceable source in either document. It appears in neither predecessor file, so it was never a companion import; it entered from research during an earlier update and was carried forward as though it had always been present. A revision that examines no new source may reorganize and restate what is in front of it, but it may not introduce a claim that was not already there. The overlay still carries the total of 27,549,840 shares issuable on warrants, so what is lost is the price at which they would be exercised, and with it any sense of how far the warrant book sits in or out of the money.
Separately, the market capitalization cited in the standing conditions was carrying a more precise figure than this document holds, and reverts to the $192 million stated on its own evidence. The falsification table has since regained that test on this document's own evidence, and now lists ten with no deferrals outstanding.
What the overlay still covers on this document's own evidence: the share count progression from December 31, 2024 to May 12, 2026, including the 6 April reading and the eighteen-day April window; the authorised total and remaining headroom; the shelf and at-the-market capacities; the resale registrations and their components; the Generate and Yorkville warrant packages; the June 2026 restricted stock grants; the company's own illustration of full at-the-market issuance; and the gating chain that ties liquidity to dilution. Neither removal has been backfilled from the companion, and restoring either requires going to the underlying filing, which is new research.
Defects that remain unresolved, with the text at issue quoted and the reason each remains open.
Three null categories rest on material that was never examined. Lawsuits, buybacks and common dividends are recorded above as "Not established" rather than as absences, because no litigation search, repurchase history or common dividend history appears anywhere in this document. A category that produces a finding or is stated explicitly as empty satisfies the coverage requirement; a category never examined does neither. Resolving these requires new research.
Two regulatory items were carried as possibly resolved, and both are now settled. The co-location and net-metering rule (R1) and the large-load interconnection standard (D13) both carried adoption windows that had opened before the date of this document. Both were checked against the Commission's own docket on August 13, 2026: the co-location rule was adopted on March 26, 2026 and the large-load standard has not been adopted. The earlier note said neither status was confirmable. Both were confirmable; one had simply never been looked up, and the other had no docket number recorded in this document to look up.
The adopted rule remains a dated catalyst, and that is now a decision rather than an unknown. Moving it to resolved would renumber the dated catalysts behind it and the whole resolved sequence as well, because that sequence is ordered by resolution date and this adoption predates every item currently in it. Thirteen identifiers would move. The item is left where it is, stating its own adoption date, until that ordering question is decided.
Primary filings and company releases first, with form type and date.
Primary. Form 10-Q for the quarter ended June 30, 2026, filed August 13, 2026, and the results release of the same date; Form 10-Q for the quarter ended March 31, 2026, filed May 15, 2026, which is the source of the $9.394 million Q1 revenue baseline used in the falsification tests; FY2025 Form 10-K; 8-K current reports of 20 January, 31 March, 2 April, 15 April, 17 April, 18 May, 20 May, 9 June, 17 June (8-K/A), 14 July, 24 July and July 30, 2026; Forms 3 and 4 filed April to June 2026; Schedule 13G filings of February and May 2026; the prior-year DEF 14A; the Series A prospectus supplement; Form S-3 filed March 9, 2026 and amended March 30, 2026, a further Form S-3 of April 29, 2026, Form 424B5 of April 1, 2026 and Forms 424B3 of April and May 2026. Company press releases and monthly business updates dated 10 March, 30 March, 2 April, 9 April, 10 April, 16 April, 21 April, 28 April, 1 May, 7 May, 18 May, 20 May, 9 June, 23 June, 14 July, 16 July and July 30, 2026.
Commission dockets, read directly rather than through analyses. Public Utility Commission of Texas Interchange, read August 13, 2026: Project 58479, item 66, Order Adopting New 16 TAC 25.205, March 26, 2026, and item 15, ERCOT stand-alone generation resource list, September 19, 2025; Project 59142, Review of ERCOT's Interconnection Processes for Large Loads, items 1 to 45, including item 42, ERCOT's requests for good cause exceptions of August 10, 2026; Project 58317, SB 6 Implementation, item 26, the Governor's August 3, 2026 letter as filed with the Commission. Filed
Secondary. S&P Global Market Intelligence for market data at the July 31, 2026 close, and the exchange daily record for the August 12, 2026 close that every price-dependent figure here is struck on. Texas Senate Bill 6 and PUCT implementation dockets as described in published law firm analyses between December 2025 and May 2026. Federal Reserve meeting calendar and the FOMC statement of July 29, 2026, with implied probabilities and Jackson Hole dates as reported July 29 to 30, 2026. Peer lease disclosures, the July 2026 AI infrastructure derating, the reported CoreWeave loan repricing, the New York data center moratorium and US Treasury yield levels as reported between December 2025 and July 2026.
Two ratings exist on the stock. Both are the published views of the firms named, reported here as market data and adopted by nothing in this document, which issues no rating and derives no target of its own.
| Firm | Rating | Target | Basis |
|---|---|---|---|
| H.C. Wainwright (M. Colonnese) | Buy | $4.00 | Market initiated or assumed Jul 20, 2026 |
| Weiss Ratings | Sell (D−) | n/a | Market upgraded from Sell (E+), Jun 12, 2026 |
| Wall Street Zen | Hold | n/a | Market upgraded from Sell, Apr 25, 2026 |
| MarketBeat consensus | Hold | $4.00 | Market one Buy and one Sell |
That firm's published estimates project losses throughout: $(0.11) for Q2 2026, $(0.08) for each of the following two quarters, $(0.45) for FY2026 and $(0.27) for FY2027. The $(0.11) estimate is the only published benchmark for the 13 August print and will function as the de facto bar. On 157.75 million shares the $4.00 target implies a market capitalization near $631 million, roughly 19 times trailing revenue, a level consistent with the same analyst's loss estimates only on the assumption that the pipeline converts. Analyzing what a published target assumes is in scope; adopting it is not.
The sole positive rating comes from the company's share-selling agent. Per the Q1 2026 Form 10-Q, the at-the-market offering agreement through which the company sells newly issued shares into the open market is with H.C. Wainwright & Co., LLC, dated April 29, 2025. On July 20, 2026 that firm published the only Buy rating on the stock.
Nothing about this is improper, and research and banking coexist under disclosure rules at every firm. To that firm's credit its own estimates candidly project losses through 2027. But a bullish note from the firm earning commissions on the company's share issuance is not independent corroboration, and a higher share price directly increases the proceeds available under the facility. Read that firm's own disclosure section before assigning weight.
Selling has concentrated in the Series A preferred rather than the common: a director sold 18,000 preferred shares on 17, June 18 and 22, 2026 at $10.75 and $11.00; the chief executive sold 1,000-share lots in April and May under a Rule 10b5-1 plan adopted December 16, 2025; the chief accounting officer sold 4,838 common shares at $1.70 in June. On the buying side, the chief financial officer purchased 100,000 common shares at $1.63 on May 22, 2026, an open-market purchase and the most substantive insider signal of the period. Insiders hold roughly 35% of shares outstanding. The disclosed institutional register is BlackRock at 5.3%, filed July 30, 2026 under Exchange Act Rule 13d-1(b), together with three financing counterparties each reporting exactly 9.99% and one holder at 1.531%; the identical 9.99% figures are a ceiling on conversion and exercise rather than blocks of common stock, so they do not add to a single institutional percentage. The chief executive's June 1, 2026 grant of 3,018,802 restricted shares is compensation, not a purchase. Filed Forms 4, Apr to Jun 2026
Newest first. The original build entry is never removed or rewritten.
This calendar said no target date has been published, and the company has published one. Its presentation targets approximately 100 MW of critical IT load at Kati 2 for service in 2027. The sentence is narrowed rather than removed, because the date this item actually turns on is a binding lease, and no lease date has been published. A target for energising capacity is not a commitment to a signed tenant, so the item stays open and stays undated, and the published target is now recorded beside it.
The phase split disagrees with the deck as well. This calendar has 100 plus 250; the deck heads the campus a 300 MW plan with at least 200 MW more after the first 100. Recorded as a disagreement rather than resolved, since the deck's wording is open-ended.
The deck carries no cover date and is used only because its pipeline page is stamped as of August 1, 2026, which is the stamp cited. That precedes this calendar's August 18 build, so nothing here is newer information. No catalyst changes state, no date moves, nothing is repriced.
Two provenance statements named only the superseded half of this document's market data. The sources list said market data was taken at the July 31, 2026 close, and the Nasdaq compliance card said the same, while every price-dependent figure in this file is struck on the August 12, 2026 close of $1.35, which that card itself cites two sentences earlier. Both lines now name the exchange daily record for the August 12 close alongside the July 31 data, which is short interest and the 52-week range. Nothing was repriced and no figure moved. The companion research report already stated both halves and is untouched.
A preliminary proxy statement exists, and the annual meeting card said none did. The card read "No proxy statement was on file as of the date of this document"; a Form PRE 14A was filed on August 14, 2026, the day this calendar was stamped. It is named and not read here: the meeting date and the resolutions are still to be taken from it, and the card's timing, confidence and provenance are unchanged.
The filing record was re-read today and the sources note was ten days behind it. That note said nothing had been filed after July 30, 2026, which was true of the sweep it describes and has not been true since August 4. Six filings have appeared since: a Form 4 of August 4, a Form 8-K and a Schedule 13G of August 5, the Form 8-K and the quarterly report of August 13 that this document absorbed, and the preliminary proxy of August 14. Four of the six were not read. The issuer newsroom was reached and carries nothing after the second-quarter results of August 13.
Not repriced, and the stamp moves without the basis. The 13, 14 and 17 August sessions have closed since this file was last stamped and it stays on the August 12, 2026 close of $1.35. No catalyst resolved, none was added and no expected date moved.
The timing field of the PUCT large-load interconnection standard (D13) stated a status rather than a date. It read "Not adopted as of August 13, 2026; Project 59142", which is when the docket was read and not when the rule is expected. The catalyst summary table has carried the timing as 2026, may be final throughout and was never brought into line, so one catalyst stated its timing two ways and nothing compared them. The card now carries the window. The verified status and the docket number are unchanged and remain in the card, which is where they were already stated.
The same card's timing confidence had been raised to High on the strength of that status check. Confidence rates how sure the date is and never rates the outcome. The date here is an unadopted rule with no scheduled adoption, four months past the point at which reporting had suggested it could come, and the status check did not make it surer. It returns to Medium, which is what the summary table had continued to state.
The earlier-numbering concordance sent a reader to the wrong catalyst. Its row for the co-location and net-metering rule still pointed at the identifier now held by the large-load interconnection standard. The co-location rule moved into the resolved sequence when it was found to have been adopted, and that row did not move with it, so a reader following an earlier log entry through the table arrived at a different rule than the one named. The row now points at the resolved card.
Nothing here changes what is known about the company, and the price is not restruck. All three were disagreements inside the file: two fields restating one card, and one row of a table whose whole purpose is to keep earlier references resolvable. The catalyst count, the standing count and the resolved count are unchanged, and the pricing basis stays at the Aug 12, 2026 close.
The adopted co-location rule left the dated sequence. Its dated event happened on March 26, 2026, and section 12.6 is explicit that a dated catalyst whose date brings an announced decision resolves and moves to the resolved sequence. It did not append. That sequence is ordered by resolution date and this adoption predates every item already in it, so the rule entered at the front and every existing resolved identifier moved down one, alongside the six dated catalysts that sat behind it. Thirteen identifiers in all, swept together with the cross-references, the section 8 concordance and the band labels, in the same pass that moved the card. The exposure did not resolve and is not in the resolved card: it sits with the Dorothy 3 development milestones, which is where an undated, live question belongs.
The two Commission rules carried at Provenance Open were checked against the Commission's own docket, and neither status was what this file assumed. Both had been recorded as owed and unchecked in the entry above, on the reasoning that re-checking them is a fetch rather than a reading. It is, and the fetch was made. The co-location and net-metering rule was adopted on March 26, 2026 as 16 TAC 25.205, at item 66 of Project 58479, which is nearly five months before this document first described it as possibly still pending. The large-load interconnection standard has not been adopted, and its docket is Project 59142, a number neither of this company's documents has ever carried: that is why the status was never checkable, since there was nothing to look up. Its docket runs to today across 45 items with no order of any kind.
The adopted co-location text answers the curtailment question in a way that fits neither the fear nor the test. Adopted 25.205(k)(1) puts a mandatory obligation on the generation resource to keep making dispatchable capacity available as ERCOT directs; 25.205(k)(2) says the Commission may, not must, require behind-the-meter load reduction. The order's own reasoning then states that a resource's capacity is made available "first and foremost through the full curtailment of the large load customer". Falsification test 6 is therefore triggered on its literal terms and the exposure is not withdrawn, which is a finding about how that test was drawn rather than about the rule.
Whether the adopted rule binds this company is not established, and the test is now specific. It reaches a generation resource registered with ERCOT as stand-alone as of September 1, 2025, with exemptions where the modeled facility already included a co-located large load at energisation and where a majority interest was owned by the load's parent as of January 1, 2025. The second cannot apply, since Briscoe was acquired on April 1, 2026. On the first, Briscoe does not appear on ERCOT's stand-alone list filed September 19, 2025, which points toward exemption; that reading is not adopted, because an absence read out of a lossy extraction of a tabular filing is weak evidence.
The audit catalyst is corrected on two points and its scope moved against this company. ERCOT did not merely commit to seek a good cause exception: it filed the request on August 10, 2026 at item 42 of Project 59142, with a draft proposed order, and it is pausing the Batch Zero study process itself rather than only the classification notifications. It states the verification "will take several months" and expects to begin shortly after the August 20 open meeting. On scope, the filing records that the Commission and ERCOT interpret data centers to include virtual currency mining facilities and other computational loads, and that community impact information will be sought from all such facilities of 25 MW or more not yet energised. What is still unestablished is whether any specific site here was submitted into Batch Zero.
One thing this entry does not do. The co-location rule's dated event has now happened, so on this file's own convention the item no longer belongs in a forward-looking dated sequence. Moving it renumbers the dated catalysts behind it and sweeps the section 8 concordance again. Re-ordering is judgment and it is not taken here; the card states its own adoption date so that no reader is misled while the question is open.
Q2 2026 reported, and the largest figure in this file moved by about $105 million. Revenue of $15.06 million, up 60.3% sequentially and 144.6% year over year, or 73.1% excluding a $4.4 million presentation change that grosses up pass-through electricity in both revenue and cost of revenue. The reconstruction of post-acquisition liquidity from disclosed deal terms, roughly $8 million of unrestricted cash and a swing into net debt, was wrong: the filed figure is $113.4 million of unrestricted cash against $33.1 million of carrying debt, so $80.2 million of net cash. Every card that reasoned from the $8 million estimate is corrected here rather than left standing. The estimate's method is not defended: it netted disclosed outflows against a disclosed opening balance and did not model the at-the-market issuance that funded them, which the cash flow statement now puts at $113.5 million for the six months.
Two of the calendar's own falsification tests were settled by the filing and one is triggered. Test 4, that Q2 revenue as filed would exceed the $9.394 million of Q1, is triggered: $15.060 million. The inference it was set against, that summer curtailment breaks the sequential growth streak, is withdrawn. Tests 1 and 3 both required a share count at or below 157,747,354 and both fail on that limb at 244,590,575, so each stands untriggered and, in the words test 1 was written with, a high cash balance with a much larger share count confirms the framing instead. Nothing was rewritten to make a test read better.
The Q2 results catalyst resolved and the Texas audit takes its place, so no identifier moved. The Q2 2026 results and first quarterly earnings call (R8) happened on August 13, 2026 and moved to the resolved sequence, and the Texas audit of ERCOT data center interconnections (D1) enters as a dated catalyst at August 20, 2026. One item leaving the dated sequence at position one and one entering at position one cancel exactly, so the 2026 annual meeting of stockholders (D2) and every dated catalyst behind it keep the identifiers they carried, through to the Green Cloud secured note maturity (D18). The section 8 concordance therefore needed a single row changed, the one mapping the earlier identifier 1, which now points at the resolved Q2 results card rather than at the first dated catalyst. That card kept its anchor, so any reference by title still resolves.
A Texas regulatory review neither file knew about, and this calendar was cleared of it once in error. An earlier sweep cleared this file on the reasoning that its Commission references concern a different proceeding, the transmission cost allocation review. They do, and the conclusion was still wrong: the pipeline conditions hold projects inside the process being audited. The pipeline (S10) and Dorothy 3 (S9) conditions now record the suspension while staying standing, since neither has a date and neither gains one. The directive is still absent from EDGAR, and that was re-tested rather than carried: the Form 10-Q filed the same day names ERCOT eighteen times and the Public Utility Commission, the Governor, the directive and market notice M-A080326-01 not once. What the 10-Q does add is that Fei and a newly named Project Gladys are in ERCOT interconnection planning in terms.
Two items before the same Commission are still carried with statuses this document has never confirmed, the large-load interconnection standard (D13) and the co-location and net-metering rule (R1). Re-checking them is a docket fetch rather than a reading and was not performed at that point, which was recorded here rather than left implicit. It was performed later the same day and both are now settled, in the paragraphs below: the co-location rule was adopted on March 26, 2026, and the large-load standard has not been adopted. The gap and its closure are both kept, in that order, because a reader should be able to see that the file knew what it had not checked.
Repriced to the 12 August close. $1.21 → $1.35, up 6.30% on the session against the one before, on volume of 9,370,569 shares, about 0.90 times the mean of the nine preceding sessions. The 13 August session had closed when this was written and the exchange had not yet published it, so the 12 August close is the latest completed session this document can read. The buffer above the $1.00 minimum bid moves from 21% to 35%. The share count moved further than the price did: 157,747,354 → 244,590,575, so market capitalization goes $190.9M → $330.2M, and headroom under the 375,000,000 authorised falls from roughly 217 million shares to roughly 130 million. The 52-week range of $0.41 to $5.14 was re-measured over the window ending 12 August rather than carried, and both ends survive it.
A renumber map was published in this entry and is withdrawn. Section 7.2 replaced the map with the sweep at v1.96 and is explicit that the entry carries none: a renumbering is apparatus, and a map published beside a sweep that was actually performed records nothing the titles do not already carry. The sweep was performed and the cross-references were checked. Nothing is stranded by the withdrawal: the map ran from the previous numbering of the dated catalysts and no preserved entry below cites an identifier from it, while the earlier-numbering concordance in section 8, which is permanent and serves the retired scheme, is untouched.
The Kati 1B Phase 3 window was measured from nothing. It read "Expected within 0 to 3 months", and a window written that way has no anchor in the document, so it moves every time this file is restamped. The company placed it within nought to three months as of July 14, 2026, which closes it about October 14, 2026. Estimate That is the company's own range measured from the company's own statement.
The monthly business update was a dated catalyst that can never resolve. A dated catalyst resolves once and moves to the resolved sequence; this one recurs about the 9th to the 14th of every month by construction. It is now a standing condition, which is what that band is for.
Nothing else changed. No figure moved, no catalyst resolved, and the pricing basis is unchanged. Both items were found by reading every timing this project's elapsed scan cannot parse, one at a time.
Repriced to the 4 August close. The session ended at 16:00 Eastern, stamped by the exchange, on volume of 11,971,770 shares, about 1.10 times the mean of the nine preceding sessions. $1.22 → $1.21, which moves the buffer above the $1.00 minimum bid requirement from 22% to 21%. The conversion and exchange prices the dilution overlay measures against move with it. No catalyst was added or removed.
Filing sweep re-run to August 4, 2026 against Central Index Key 0000064463. The index paginates: 1,547 filings across the current view and one overflow file it references. Four filings since the previous sweep: a Form S-8 of 28 July, a current report and a Schedule 13G of 30 July, and a Form 4 of 4 August.
Third-party coverage extended in the companion report. A Zacks Research upgrade to Hold of July 21, 2026 is added, the most recent action on the stock, and the published consensus is accordingly one Buy, one Hold and one Sell. Three of the four contributors are quantitative ratings services rather than broker research and are marked as such; the sole broker, H.C. Wainwright, is the company at-the-market sales agent under an agreement the Q1 2026 Form 10-Q records. No catalyst changed class and no identifier moved.
The statement that no late-filing notification had ever been made was false. This calendar recorded that no NT filing of any kind exists and treated it as informative for an issuer that had changed accounting basis and filer status. Three exist: a Form NT 10-K of December 30, 1996 and Forms NT 10-K/A of March 27, 2000 and March 27, 2001. They are old and none bears on the current accounts, but the claim was about the whole record. It rested on the 1,000 filings the current index view returns; the complete index holds 1,546, and the three sit in the portion that view does not reach.
The FY2025 audit opinion has been read and the carried defect is closed. It is unqualified, with no going-concern explanatory paragraph and no emphasis of matter; going concern is the second of two critical audit matters, and management plans are recorded as alleviating the doubt the conditions raised. Dated March 27, 2026 at Albany, New York.
No catalyst changed state. The filing record was swept to today by issuer central index key against the complete index; nothing has been filed since the Form 8-K of July 30, 2026, and the 31 July close of $1.22 remains the latest.
The falsification table carried seven tests and three open defects. All three are closed and the table carries ten. The Q2 revenue test is restored with its baseline sourced here from the quarterly filing of May 15, 2026 rather than from the companion document, which is what put it in the deferrals. The Series A settlement test and the insider buying test now carry stated thresholds, tagged as estimates because each is a judgment rather than a filed figure.
Net change since the previous entry. No repricing: July 31, 2026 was a Friday and the as-of date is the following Sunday, so the close the figures come from has not moved and the pricing basis is unchanged.
No catalyst moved. Nothing resolved, nothing was added and nothing changed its expected date between the two entries. The concordance in Methodology is unaltered and remains correct.
Filing sweep. Swept 8-K, 6-K, 10-Q, 10-K, 20-F, 40-F, S-1, S-3, 424, 13D, 13G, the full NT series, Forms 3 and 4 and DEF 14A, most recent first, reaching August 2, 2026. Nothing has been filed since the 8-K of July 30, 2026. Three late-filing notifications do exist, a Form NT 10-K of December 30, 1996 and Forms NT 10-K/A of March 27, 2000 and March 27, 2001, none of them recent. They sit in the older portion of the filing index, which the complete record reaches only when the index is read in full: 1,546 filings, not the 1,000 the current view returns. The earnings call at August 13, 2026 (D1) is unaffected and remains the next dated item.
The previous entry follows unchanged.
This revision supersedes the 26 July build. Net changes from that version:
Liquidity, corrected twice. A third acquisition was added: $8,765,490 in cash for the remaining 49% of Project Dorothy 1B on May 19, 2026, referenced only qualitatively at C2 in the original. Briscoe was restated at approximately $55.9 million total cash consideration against the $53.0 million headline (C1), with the asset's 2025 net loss of $14,773,838 on revenue of $8,950,927 now shown against the $6–11 million EBITDA projection (#8). Estimated post-acquisition unrestricted cash falls from roughly $16–19 million to roughly $5–10 million.
Liquidity and dilution merged into one question. The gating note and Section 8 now show that roughly 46 million shares, about 41% of the starting count, were issued between 31 March and May 12, 2026, which is how the acquisitions were funded. Registered capacity corrected from $87.65 million to a $1.0 billion shelf including a $500 million ATM prospectus, and split into primary capacity versus resale overhang: roughly 31.7 million shares registered for resale, 27,549,840 issuable on warrants, and the company's own illustration of 760,178,588 shares were the ATM fully sold at $0.77. The June 2026 grant cluster totalled at least 12,631,694 restricted shares, about 8% of shares outstanding, with director awards vesting on separation (#4).
Structure and ownership. Kati 2 economics quantified: capital returned plus a 14% IRR plus $100,000 per Gross PPA MW, then 50/50 with Metrobloks (C3, #17). Schedule 13G holder Robert L. Bugbee went from 9.3% to 8.5% while buying 4.2 million shares (S6). A sixth pipeline project, Grace, was added at #25. A systematic review across 8-K, 6-K, 10-Q, 10-K, 20-F, 40-F, S-1, S-3, 424, 13D, 13G, Forms 3 and 4 and the full NT series confirmed no NT, 13D, 6-K, 20-F or 40-F filings exist; the last three cannot, as Soluna is a domestic filer, and nothing has been filed since July 30, 2026. The review reconciled the resale figure (26.5m SEPA plus 2.7m Generate warrants plus 2,459,400 YA and HEL equals 31.66m) and the June grant cluster (12,631,694 across five insiders), and added four details: Briscoe's nameplate is approximately 149.85 MW (C1); the board is classified into three staggered classes, with removal only for cause on a 75% vote, which weakens the annual meeting as a lever (#4); Soluna director Matthew Lipman is also President and a director of Harmattan Energy (S6); and UHY LLP audited FY2025 and FY2024, with the incorporated EXPERTS paragraph carrying no going-concern language.
Resolved and added. Catalyst #10 resolved: the FOMC held at 3.50–3.75% on 29 July on a 9–3 vote with three dissents favoring a hike, and implied odds moved toward a September increase. New catalyst #11, Jackson Hole, 27–August 29, 2026, shifting former #11–#25 to #12–#26. Added: the 30 July earnings-date confirmation (#1), the CEO salary increase to $600,000 (#4), the Dorothy 1B terms (C2), and Bitcoin's 21-month low near $58,000 (S1).
Sector and macro. The July 2026 derating of 25–38% across the AI infrastructure peer group, in which signed contracts did not protect share prices (#2); tightening AI project-finance conditions including the reported CoreWeave loan repricing (#17); the long-end rate channel, with the 30-year Treasury near 5.25%, its highest since 2007 (#10, #18); and the New York data center moratorium of 14 July as siting precedent (#15).
Market data. All price-linked figures reset from the 24 July to the July 31, 2026 close of $1.22 (−6.15%), market capitalization $192 million. S3 gains an independent corroboration that the Series A preferred trades near 45% of face.
The baseline version, superseded by the entry above. It carried 25 dated catalysts (#1–#25) across three horizons, 6 standing conditions (S1–S6), 6 settled events (C1–C6), 12 explicit null categories, the gating-dependency chain, the dilution overlay, the disclosed-conflicts section and 10 falsification tests. Market data was at the July 24, 2026 close of $1.26, market capitalization $198.8 million.
It was built from the Form 10-Q for the quarter ended March 31, 2026, the FY2025 Form 10-K, 8-K current reports and Forms 4 filed March to July 2026, the prior-year DEF 14A, the Series A prospectus supplement and monthly business updates through July 14, 2026, plus Texas SB 6 and PUCT docket analyses and the published FOMC calendar. Its central judgment, that almost every forward catalyst is downstream of a single signature at Kati 2, and that the catalysts are correlated rather than independent, is unchanged.