Castellum ($CTM) Stock Hub 2026: Q2 Revenue Holds Flat, Margins Compress and the Market Reaction on August 7

01 Q2 2026 results: flat revenue, lower margins and a stronger cash position

Castellum released second-quarter and first-half results after the close on August 6, 2026. The quarter was not a continuation of the 22.5% growth reported in Q1. Revenue of $13,864,676 declined 1.1% from $14,024,090 in Q2 2025 and fell 3.0% sequentially. Gross profit declined 6.7% to $4,719,419, gross margin compressed to 34.0%, and adjusted EBITDA fell to $25,792 from $500,317. The GAAP net loss widened to $1,026,597.

The half-year view is better than the isolated quarter: revenue rose 9.6% to $28,156,637, gross profit increased 1.7% to $9,781,640 and the net loss to common shareholders improved to $1,431,510 from $1,545,958. Cash increased by $1.09 million during Q2 and by $1.98 million from year-end to $16,865,160, with no interest-bearing debt. Management expects 2026 to set a company record for revenue, but it did not provide a numeric revenue, EBITDA or earnings range.

Direct links: official Q2 and first-half release · live audio webcast · Castellum investor relations.

02 Executive summary

Castellum is a small United States government services contractor selling cybersecurity, electronic-warfare support, software engineering and IT modernization primarily to the Department of Defense, with the U.S. Navy as its center of gravity. The company was assembled through seven acquisitions, then used equity and operating improvement to remove the debt burden created during that roll-up. The balance-sheet repair is now complete: cash reached $16.865 million at June 30, 2026, interest-bearing debt was zero and total liabilities were only $5.942 million.

Q2 shows that the next phase will be harder. Revenue was almost unchanged year over year at $13.865 million despite the ramp of long-term prime contracts. Gross margin fell 2.0 percentage points to 34.0%, adjusted EBITDA declined 94.8% to just $25,792 and the operating loss widened to $1.094 million. Management attributes the pressure to a higher subcontractor mix, completion costs on two firm-fixed-price contracts and planned spending on business development, investor relations and acquisition work.

The half-year figures keep the growth case alive. Revenue rose 9.6% to $28.157 million, cash increased by $1.98 million from year-end, total backlog held near $272 million and the qualified pipeline expanded to $953.5 million. Management expects record full-year revenue and says approximately 16% of backlog should convert over twelve months and 48% over twenty-four months. That is qualitative guidance, not a numeric forecast, and the Q2 release did not provide a refreshed funded/unfunded/options split.

The central question is no longer whether Castellum can survive its old debt structure. It is whether the company can convert a large contract base into higher-quality revenue, recurring operating profit and value per share without restarting the dilution cycle. Q2 strengthens the balance-sheet evidence but weakens the near-term margin evidence.

03 Market Data And Peer Comparison

Price and performance figures below are based on the completed session of Friday, August 7, 2026. Float, ownership, short interest, average volume and the consensus target are from Finviz, pulled on the same date. Company financial figures come from SEC filings and company releases, each carrying its own reference date.

Peer comparison, all figures at the August 7, 2026 close

The August 7 session is the single most important market data point on this page and it postdates the results. The shares fell 22.76% to close at $0.615, on volume of 5.35 million shares against a recent average nearer 1.4 million, in the first full session after the second quarter release of August 6. The hub sections below were written from the release itself; this section carries the market reaction to it.

On analyst coverage the honest position is a narrow one. The consensus target above is a Finviz aggregate of third-party estimates pulled on August 7, 2026. Individual houses, ratings and note dates were not verified for this update, so no coverage table is presented. A consensus figure without named notes behind it is a market-data point, not research, and it is neither a company figure nor a Merlintrader forecast.

04 Verified developments through August 6, 2026

Q2 revenue $13.865 million, gross margin 34.0%, adjusted EBITDA $25,792 and GAAP net loss $1.027 million. H1 revenue increased 9.6% to $28.157 million. Cash reached $16.865 million, the balance sheet remained debt-free, backlog was $271.7 million and the qualified pipeline was $953.5 million. Management expects record 2026 revenue but issued no numeric range.

Castellum scheduled the release for August 6 and the conference call for August 7 at 10:00 a.m. ET.

Glen R. Ives’s term was extended through December 31, 2027. The package included 773,630 options at $0.73 and a performance-based acquisition bonus.

Specialty Systems received a directed subcontract under SAIC’s GSA ASTRO prime contract for Navy Aircraft Data Management and Control System modernization.

CTM JV won a position on a $250 million Navy multiple-award vehicle shared across 59 companies. Individual task orders remain separately competed.

Revenue grew 22.5% to $14.292 million, adjusted EBITDA reached $394,930 and the detailed backlog table showed $22.398 million funded, $38.300 million unfunded and $212.562 million of priced options.

Castellum and its subsidiaries completed the certification required for sensitive Department of Defense work involving controlled unclassified information.

The final $400,000 related-party note was retired, completing the deleveraging of the acquisition-built balance sheet.

Primary update: Castellum Q2 and first-half 2026 results.

05 What Castellum is, and how the business actually earns money

Castellum is a holding company. The contracts, the clearances and the past-performance record sit inside the subsidiaries, which is the structure most small government services firms adopt because in federal contracting the qualification travels with the operating entity, not with the parent.

Seven acquisitions in roughly four years, one divestiture, one dissolution. The company reports a single operating and reportable segment, so there is no segment-level revenue or margin disclosure to work with. At December 31, 2025 it employed 244 full and part-time employees plus nine independent contractors, of whom 58% hold degrees in science, technology, engineering or mathematics, 20% hold advanced degrees, and 97% hold security clearances. That last figure is the practical description of the asset: this is a cleared-labor business, and revenue moves with billable headcount and funded task orders almost mechanically.

Contract vehicles: the permission to bid

Very little in this sector is sold directly. Work flows through vehicles, pre-competed frameworks a contractor must hold a position on before it can bid a task order at all. Castellum holds or accesses SeaPort-NxG, the Navy’s principal services vehicle; GSA Multiple Award Schedule, to which Specialty Systems added special item number 54151S for IT professional services in October 2025; GSA ASTRO, reached through SAIC as prime; OASIS+; the Missile Defense Agency SHIELD multiple-award IDIQ; and the new Navy logistics LIIS CMDS vehicle. It also bids through Small Business Administration mentor-protege joint ventures, including CTM JV, C-K2 and Epic Specialty Systems. Holding a vehicle position is necessary and not sufficient: it is the difference between being allowed into the competition and being paid for work.

06 The roll-up, the goodwill and what the acquisitions actually cost

Seven businesses were bought with a combination of cash, seller notes and stock, and the accounting record shows what happened next. Goodwill on the balance sheet is $10,676,834 and has been unchanged at that figure since September 30, 2024. Finite-lived intangible assets, mainly customer relationships and trade names, have amortized from $6,793,750 at December 31, 2024 to $5,067,056 at March 31, 2026. Together the two lines are $15,743,890, which is 43.6% of the $36,142,641 of stockholders’ equity. Tangible book value is therefore about $20.4 million against a market capitalization near $69 million.

There has been one impairment. In 2023 the company recorded a goodwill impairment charge of $6,919,094, the single largest reason the 2023 net loss reached $17,800,178; the 10-K records that a decline in the stock price during the third quarter of 2023 was treated as the triggering event. It also states that no triggering events were identified during 2024 or 2025. That is a clean bill for the carrying value, and it is a judgment that depends on the share price and forecast cash flows, so a sustained decline could reopen it.

The debt used to fund those deals has now been cleared, and interest expense tells that story compactly: $1,732,265 in 2023, $706,054 in 2024, $385,912 in 2025, and effectively nothing in the first quarter of 2026, where the only charge was $2,301 on the final related-party note.

07 The balance-sheet repair, step by step

This is the strongest documented fact the company owns, and it happened in dated increments that can be checked against the filings.

What remains at March 31, 2026 is ordinary and small: total liabilities of $5,681,998, comprising accounts payable and accrued liabilities, accrued payroll of $2,952,153, operating lease obligations of $738,261, and a $10,000 derivative liability attached to a legacy warrant tranche from a 2022 financing. There is no revolver, no convertible, no earnout and no covenant package.

The qualifier that matters. The debt was not repaid out of operating cash flow. Operating cash flow was negative $1,948,377 for the whole of 2025 while financing activities contributed a positive $4,737,880. In the company’s own December 2025 summary it had raised over $25 million in gross proceeds through public offerings and warrant exercises. The balance sheet was repaired with equity, which is a legitimate way to do it, and the share-count section below records the price.

Balance sheet data from the Forms 10-Q and 10-K filed for each period. Interest-bearing debt combines notes payable and the related-party note; it excludes operating lease liabilities, which are not borrowings.

08 The income statement after Q2 2026

The second quarter interrupts the growth acceleration visible in Q1. Revenue was stable, but the gross-profit, operating-loss and adjusted-EBITDA lines all weakened. The first-half comparison remains positive because Q1 was strong.

Quarterly revenue, Q1 2025 to Q2 2026

The series remains lumpy rather than linear. Q2 revenue declined 3.0% sequentially and 1.1% year over year even as first-half revenue increased 9.6%.

Detailed historical income-statement baseline through Q1 2026

Before the Q2 release, the first quarter of 2026 was the latest reported period and established the following baseline. The comparison is with the first quarter of 2025.

The genuinely new line is the last but one. First-quarter operating cash flow of positive $1,290,696 compares with negative $2,502,640 a year earlier and negative $1,948,377 for the whole of 2025. Financing activities were a net outflow of $426,819, essentially the note repayment and the preferred dividend, and cash still rose $888,196 without any equity being sold. Whether that repeats is one of the two or three things worth watching on August 6. The annual record puts the quarter in context.

Two things about that table deserve to be said precisely. The 2023 net loss of $17,800,178 included the $6,919,094 goodwill impairment, so the improvement from 2023 to 2024 is partly a non-recurring charge dropping out. The improvement from 2024 to 2025 is not: it is operational, and it happened while gross margin fell.

One discrepancy is worth flagging because it circulates. The March 4, 2026 press release describes 2025 revenue as an “$8.1 million (15.2%) increase” over 2024. The two audited figures, $52,866,001 and $44,764,852, produce a difference of $8,101,149, which is an increase of 18.1%, not 15.2%. The filed statements are the authoritative version.

Quarterly revenue, Q1 2025 to Q1 2026

Bars are scaled to the largest quarter, $14.62 million in Q3 2025, which equals 100% of the track. Q1 to Q3 2025 and Q1 2026 are reported figures from the Forms 10-Q; Q4 2025 of $12,557,859 is the audited full year of $52,866,001 less the nine months of $40,308,142. The point of the chart is that the trajectory is not a straight line: the fourth quarter fell 14.1% from the third. Government services revenue moves with task-order funding and federal budget timing.

09 The margin question became more demanding in Q2

Gross margin fell to 34.0% from 36.1% in Q2 2025 and 35.4% in Q1 2026. Revenue was almost unchanged year over year, but gross profit declined by $341,028. That means the Q2 weakness was not primarily a top-line problem; it was a revenue-quality and cost-mix problem.

CFO David Bell identified three contributors: a higher share of subcontractor work, which generally carries less margin than direct labor; costs required to complete the remaining work on two firm-fixed-price contracts; and planned spending on business development, investor relations and acquisition activity. The first two affect gross margin, while the planned investments affect operating expense and adjusted EBITDA.

This distinction matters. A temporary closeout cost on two firm-fixed-price contracts could fade. A structurally higher subcontractor mix would not. The August 7 call should clarify how much of the 34.0% margin is temporary and whether management still sees a path to durable higher-margin performance in the second half.

Contract-mix evidence established before Q2

Revenue rose 22.5% in the first quarter while gross profit rose 11.1%. Gross margin fell from 39.0% to 35.4%. On the earnings call for the fourth quarter that gap will be the first thing a professional analyst asks about, and the filings already contain most of the answer.

Quarterly gross margin

Bar widths are scaled so that 40.0% fills the track. Each margin is gross profit divided by revenue for that quarter, computed from the Forms 10-Q and 10-K: 4,554,616 on 11,664,365; 5,060,447 on 14,024,090; 5,490,568 on 14,619,687; 4,263,227 on 12,557,859 for the implied fourth quarter; and 5,062,220 on 14,291,961.

The 10-Q disaggregates revenue by contract type, and the shift between the two first quarters is large.

Revenue by contract type, share of the quarter

Bar width equals the share of total quarterly revenue. Dollar amounts from the Q1 2026 Form 10-Q: cost plus fixed fee $9,031,745 versus $5,988,283; time and material $4,373,628 versus $4,935,016; firm fixed price $886,588 versus $741,066. Totals $14,291,961 and $11,664,365.

Cost-plus-fixed-fee work went from 51.3% of revenue to 63.2%, and grew 50.8% in dollars, while time-and-material revenue actually fell 11.4%. On a cost-plus contract the fee is set as a percentage of allowable cost and the contractor keeps no efficiency gains, so a mix moving toward cost-plus mechanically compresses reported gross margin even when execution is unchanged. That is the most likely single explanation for the 3.6 points, and it is consistent with the CFO’s own language on May 8, that gross margin “was somewhat affected by changes in contract mix and type.”

Two other explanations point the same direction and are not mutually exclusive. Full-and-open competitive awards typically price tighter than small-business set-asides, and the company won its first full-and-open contract, the $66.2 million NAWCAD Lakehurst award, in October 2025, with revenue ramping in 2026. And subcontract work such as ADMACS under SAIC carries less margin than prime work, because the prime takes a layer.

If the mix shift is deliberate, taken in exchange for larger, longer prime work and past performance that unlocks bigger bids, it can be a sound trade at this size. If it is competitive pressure, it caps how profitable the business can become at any revenue level. The Q2 release confirmed additional margin pressure; the full 10-Q and August 7 call are needed to update the contract-type mix behind it.

10 Backlog after Q2: the headline held, but the new funded split is missing

Total backlog was $271.7 million at June 30, 2026, down 0.6% from $273.3 million at March 31 and above the $265.0 million year-end comparison used in the Q2 release. The qualified pipeline increased to $953.5 million from $938.0 million in Q1. Backlog held broadly stable while the opportunity set grew.

The most detailed current composition remains the March 31 Form 10-Q: funded backlog of $22,397,810, unfunded backlog of $38,300,351 and $212,561,804 of priced options. Funded backlog represented only 8.2% of the $273,259,965 total. The August 6 release updates the total but does not publish the June 30 split, so it would be incorrect to assume that the funded portion moved in the same direction as the headline.

Management continues to expect approximately 16% of backlog to convert over twelve months and 48% over twenty-four months. Sixteen percent of $271.7 million is about $43.5 million, below the current trailing revenue rate. Existing backlog therefore does not by itself support record full-year revenue; the difference must come from new task orders, funding increments and awards entering revenue during the period.

Detailed March 31 backlog baseline and definitions

Total backlog stood at $273,259,965 at March 31, 2026. Against annual revenue of $52.9 million that looks like more than five years of work already sold. The company’s own 10-Q supplies the correction, in a table that is rarely quoted.

Backlog composition at March 31, 2026

Bar width equals the share of the $273,259,965 total. Exact figures from the Form 10-Q for the quarter ended March 31, 2026: funded $22,397,810 (8.2%), unfunded $38,300,351 (14.0%), priced options $212,561,804 (77.8%).

The company defines each component in the filing. Funded backlog is work for which money has actually been appropriated, less revenue already recognized. Unfunded backlog is orders under existing contracts for which funding has not been appropriated. Priced options represent 100% of the potential revenue value of scheduled future option periods the customer may exercise at its option, also unfunded.

So 77.8% of the headline is option periods the government has the right, but not the obligation, to exercise, and a further 14.0% is authorized in principle without money behind it. The portion the government has actually funded is $22.4 million, about 1.6 quarters of work at the first quarter revenue rate. That structure is normal and it is how most federal services contractors report; the point is that treating the headline as revenue visibility overstates the funded position by roughly twelve to one.

Read that way, the first quarter looks better rather than worse. The funded component rose 82.0% in three months, from $12.3 million to $22.4 million, which is exactly what should happen when new prime contracts move from award into execution and contracting officers begin obligating money. The headline growth of 5.8% understates the quality change.

Two caveats belong with it. First, the May 8 press release describes the December 31, 2025 base as “$265 million” while the audited Form 10-K reports $258,194,285; the filed figure is the one used above. Second, backlog excludes contracts under protest and excludes task orders under indefinite-delivery vehicles except to the extent that task orders have actually been awarded, which means positions on the SHIELD and LIIS CMDS vehicles do not enter backlog until work is won under them.

On conversion the company says it expects to recognize approximately 16.0% of the remaining performance obligations over the next twelve months and approximately 49.0% over twenty-four months, against 18% and a similar two-year figure at December 31, 2025. Sixteen percent of $273.3 million is about $43.7 million, which is below the trailing twelve-month revenue of $55.5 million. Backlog conversion alone does not cover the current run rate; the balance has to come from new task orders and awards not yet in backlog. Note also that the 10-Q phrases the percentage against remaining performance obligations while the press release phrases it against the backlog headline, and those are not the same base.

Alongside backlog the company reports a qualified pipeline of $938 million at March 31, 2026, up from $817 million at year end. Pipeline is identified opportunity, not award; it is a measure of business-development activity, useful as a trend and not as an asset.

11 The contracts, one by one, and what each headline actually is

Castellum issues a press release for essentially every award, and the headline values are large relative to a company worth about $69 million. Each is set out below with the company’s own characterization, because the wording carries the meaning.

Set out this way the picture is clearer than the headlines and it is not a bad picture. Three of these are genuine prime contracts, and the company sums them to just over $219 million of potential value across a five-year runway: PMA-290 at $103.3 million, the Lakehurst mission operations award at $66.2 million and the software support recompete at $49.8 million. That arithmetic checks: $219.3 million.

What the picture does not support is adding all eight rows together. A ceiling shared among unnamed awardees, a vehicle split fifty-nine ways and a $4 million subcontract are not comparable to a $66 million prime contract, and none of them becomes revenue until a task order is funded.

ADMACS, read carefully

The June 24 award has been read more expansively than the filing supports, so it is worth separating what is known from what is inferred. ADMACS is the Navy’s Aircraft Data Management and Control System, a command-and-control application that coordinates carrier-based flight operations and aviation readiness. Specialty Systems is lead system integrator on the modernization, responsible for technical planning, software modernization, containerization, automated testing, DevSecOps integration and delivery of modernization artifacts, in a phased model that starts with a representative vertical slice.

The case for treating it as more significant than $4 million: a directed subcontract means the government specified this contractor rather than leaving the choice to the prime, lead-integrator status on a carrier flight-deck application is a credential that supports future bids, and a phased start is how larger modernization work usually begins. The case against over-reading it: $4 million is about 7.6% of one year’s revenue, Castellum is a subcontractor rather than the prime, and there is no disclosed commitment beyond the initial phase. What would settle it is a disclosed period of performance or follow-on scope, and neither exists today.

One structural gap in the award record

Every award listed above is Navy or Missile Defense Agency. No Air Force or Army award has been announced in any Form 8-K during 2025 or 2026. The customer base is not merely concentrated in the Department of Defense; it is concentrated within one service branch and a small number of commands, principally NAWCAD Lakehurst and Naval Air Systems Command. That is the flip side of deep past performance: the same incumbency that makes recompetes winnable also makes the revenue base narrow.

12 Customer concentration: the number that got worse while everything else improved

The disclosure is in the notes, in the company’s own words. For the years ended December 31, 2025, 2024 and 2023, three customers represented 73%, 47% and 52% of revenue respectively. The same three represented 73% of accounts receivable at December 31, 2025, against four customers at 65% a year earlier.

The quarterly disclosure, which is more current and which most summaries skip, is slightly less extreme: for the three months ended March 31, 2026 and 2025, three customers, described in the filing as all parts of the U.S. government, represented 71% and 63% of revenue. Receivable concentration was 73% at March 31, 2026 against 65% at December 31, 2025.

Winning larger prime contracts naturally concentrates revenue, and the ultimate counterparty is the United States federal government, which does not default. Concentration among agencies of a sovereign customer is different in kind from concentration among three commercial firms. The credit risk is negligible.

The risk is not credit, it is renewal and funding. If one of those three relationships ends through a lost recompete, a descoped program or a budget line cut, the revenue effect is immediate and large, and a company running near breakeven has little cushion to absorb it. Going from 47% to 73% in a single year means the business became structurally more fragile at the same time as it became less unprofitable.

Management’s February 2026 statement that there are “no recompete risks on any of our current contracts” during 2026 addresses the near term directly and is reassuring for this year. It says nothing about 2027 and beyond, when the cycle resumes.

13 Capital structure, dilution and what is still outstanding

For a micro-cap the share count is usually the most important number on the page, and here it is the clearest explanation of why an improving business has a falling stock.

Common shares outstanding

Bars are scaled to the largest value, 94,698,939 shares. Balance-sheet figures from the Forms 10-K for each year end: 41,699,363; 47,672,427; 77,076,129; 94,612,750. The May 7, 2026 figure of 94,698,939 is the cover page of the Q1 2026 Form 10-Q.

The count roughly doubled in two years while revenue grew about 17%, from $45.2 million to $52.9 million. Revenue per share therefore fell even as the business grew, which is the arithmetic behind a stock that halved while the income statement improved. The encouraging detail is at the bottom of the chart: the share count has risen by 86,189 shares since December 31, 2025, no shares were issued during the first quarter, and there has been no registered offering, private placement or at-the-market program in 2026. The phrase “at-the-market” does not appear in the FY2025 Form 10-K. The dilution was concentrated in 2024 and 2025, when it was funding the debt payoff.

The instruments that could still dilute

The company’s own figure, disclosed as of March 6, 2026, is a maximum potential dilution of 19,950,231 shares, comprising the preferred conversions, 13,352,500 options and 5,653,981 warrants. Updating for the first-quarter forfeitures and the July grant gives roughly 20.4 million, which is about 21.6% of the shares outstanding.

Two observations balance that. Almost all of the options and every warrant are struck far above the current price, so at today’s share price they are economically worthless and exert no near-term dilution pressure. Conversely, the CEO’s July grant had to be priced at $0.73, and that low strike makes it the tranche most likely to become dilutive if the shares recover. One structural item, stated without alarm: the certificate of incorporation authorizes 3,000,000,000 common shares against 94.7 million outstanding. Large authorizations are common after a reverse merger and authorized capital is not issued capital, but there is no charter constraint on future issuance.

The shelf

A Form S-3 universal shelf was filed on January 10, 2025 and declared effective on January 24, 2025, allowing up to $100,000,000 of equity and debt. Two offerings were executed under it, both with Maxim Group as sole placement agent: 4,500,000 units at $1.00 in March 2025 for about $4.5 million gross, with 1,755,543 attached warrants exercised at $1.08 and 2,744,457 expired; and 4,166,667 units at $1.20 in June 2025 for about $5.0 million, with 3,673,666 warrants exercised at $1.22 and 493,001 expired. Roughly $9.5 million of the $100 million capacity was consumed, leaving most of the authorization unused. That is the open question hanging over the equity: the capacity exists, the company has said it wants to complete an acquisition, and stock issued at $0.73 is considerably more dilutive per dollar raised than stock issued at $1.00 or $1.20. That is arithmetic, not a prediction.

14 Management, governance and the incentive that was written into the CEO’s contract

President and CEO Glen R. Ives took the role in July 2024 and is the executive associated with the turnaround. The board has now committed to him beyond it: the second amendment to his employment agreement, effective July 1, 2026, extends the term eighteen months to December 31, 2027, raises base salary to $375,000 and then $386,250, and grants 773,630 options at $0.73 vesting quarterly.

The explicit acquisition bonus of up to 100% of base salary, payable only if the acquired business meets board-approved net sales projections in the twelve months after closing, tells you the board has made mergers and acquisitions an incentivized objective rather than an aspiration. The performance condition protects against paying for a deal that underdelivers; it also means any eventual announcement should be read knowing that management is compensated for making one.

The rest of the team: David T. Bell continues as Chief Financial Officer and Treasurer, Andrew Merriman is Chief Operating Officer at $290,000 from January 1, 2026, and Tammy L. Martin became General Counsel on the same date at $290,000, replacing Jay O. Wright, who resigned as EVP Strategy, General Counsel, Secretary and director at the end of December 31, 2025 with no disclosed dispute. The board is five members after the May 19, 2026 annual meeting: Alarie, Campbell, Champoux, Ives and McMillen. Champoux is the independent chairman, so the chair and chief executive roles are separated, which is the better arrangement and not universal at this size. RSM US LLP was ratified as auditor for fiscal 2026.

Insider activity

Ninety-five Form 4 filings sit in the company’s EDGAR history, and the 2025-2026 pattern is asymmetric. Departing insiders sold in size: Jay O. Wright sold roughly 2.9 million shares across many filings during 2025 at prices between about $1.04 and $1.47, with his final sale on December 22, 2025, nine days before his resignation took effect. Mark C. Fuller, then a director, sold approximately 1.5 million shares in 2025 at prices from $1.048 to $1.493. C. Thomas McMillen, a continuing director, sold 55,000 shares at $0.8018 on March 20, 2026 and 30,000 at $0.7043 on March 27.

Continuing management has bought, but in token amounts: the July 1, 2026 purchases at $0.612 were 1,262 shares for the CEO and 1,202 each for the COO, CFO and General Counsel. The fairest reading is that the large sellers were people who were leaving, and departure sales are the least informative category of insider activity. The scale gap remains awkward. The CEO’s own direct holding is 200,878 shares, worth about $148,000 at the current price: his alignment is overwhelmingly through options rather than stock he paid for.

15 Ownership, short interest and how the stock trades

The most informative item is an absence. Castellum’s EDGAR history contains two Schedule 13Ds from October 2022 and a single Schedule 13G from February 2023, and nothing since. There is no reporting 5% institutional holder disclosing a current position. Combined with average daily volume of about 4.5 million shares on an 83.3 million share float, that produces a stock whose price is set by a small number of retail participants, which is why it can move several percent in a session on no company news at all.

Retail discussion of $CTM on message boards is persistent and is driven by the arithmetic of headline contract values against a $69 million market capitalization: a $250 million number attached to a company this size is inherently attention-getting. Those are the observations of non-professional traders, not company statements or analyst work. They help explain why the price moves between reports; they are not evidence about the business.

The block below is a snapshot of the Stocktwits flow, with its date. These are opinions of retail traders and non-professional investors, not analyst research, and they measure attention and how one-sided positioning has become rather than anything about the business.

16 Listing status, internal controls and litigation: an unusually clean file

Stocks trading below a dollar frequently carry listing problems. Castellum does not, and the distinction between a disclosed risk and an event that has occurred matters here.

• No deficiency notice has ever been received. The common stock has traded on the NYSE American since October 13, 2022, having previously been quoted on the OTC Pink marketplace. The FY2025 Form 10-K discusses the exchange’s discretion to require a reverse split where a stock sells at a low price for a substantial period, and states plainly: “We have not received any such notification from the NYSE American,” while noting it could in the future.
• No equity shortfall. Stockholders’ equity of $36,142,641 is comfortably above listing thresholds. There is no compliance plan and no listing-related Form 8-K in either 2025 or 2026.
• No material litigation. Both the 10-K and the Q1 2026 10-Q state that neither the company nor its subsidiaries is party to any legal proceeding believed to be uncovered by insurance or otherwise material.
• No restatement, no going-concern language, no material weakness. Disclosure controls and internal control over financial reporting were both concluded effective at December 31, 2025, with the audit report from RSM US LLP dated March 9, 2026.

One nuance belongs with that list. As a smaller reporting company Castellum is exempt from the auditor attestation on internal control over financial reporting, so the effectiveness conclusion is management’s own rather than independently attested. That is standard at this size and is not a red flag by itself. And a sub-dollar share price does create a live possibility that the exchange asks for a reverse split at some future point; nothing of the kind has been requested.

17 Valuation arithmetic after the Q2 release

At the latest August 6 quote of $0.7963 and 94,698,939 shares, Castellum’s market capitalization was approximately $75.41 million. Subtracting $16.865 million of June 30 cash and adding no interest-bearing debt produces a simplified enterprise value of about $58.54 million.

The revenue multiple looks undemanding relative to larger profitable federal-services peers, but the EBITDA multiple shows why the stock cannot be evaluated on revenue alone. The Q2 print added cash and preserved the debt-free structure, yet it reduced evidence of operating leverage. A sustained margin recovery would improve the earnings denominator; continued revenue growth near a 34% gross margin with rising operating expense would not.

18 Analyst coverage: one firm, and a relationship worth disclosing

Data services show a single Buy rating on $CTM, displayed by Finviz Elite with an average price target of $3.50 against a share price of $0.73. That figure requires context before it means anything.

The consensus is one analyst. Maxim Group initiated coverage in April 2025 and no other firm publishes on the stock, so an “average price target” drawn from a single estimate is not a consensus in any useful sense. Maxim was also sole placement agent on both 2025 offerings, earned fees on the associated warrant exercises, and hosted the company at its own Growth Summit in October 2025 and at its Defense Tech conference on June 25, 2026. Banks covering companies whose shares they have placed is normal, disclosed practice and none of it makes the research inaccurate; it does mean the target is one interested party’s model rather than independent consensus.

The more informative signal is again the absence. A company with $52.9 million of revenue, three prime defense contracts and a clean balance sheet has attracted no second analyst and no reporting institutional holder. That reflects liquidity, float and price thresholds rather than the business, and it means there is no natural institutional buyer base to reprice the equity if results improve.

19 Catalyst table after the Q2 release

Webcast: August 7 conference call.

20 The two cases after Q2, stated as fairly as possible

Castellum has removed the balance-sheet risk that defined the old story. Cash reached $16.865 million, there is no debt, H1 revenue grew 9.6%, backlog remains near $272 million, the pipeline reached $953.5 million and the share count was essentially unchanged. If fixed-price closeout costs fade, higher-margin direct labor replaces subcontractor work and business-development investment produces funded task orders, the current revenue base could generate far better earnings without a new financing cycle.

Q2 may be the more representative economics of the platform: nearly flat revenue, 34.0% gross margin, a $1.094 million operating loss and adjusted EBITDA of only $25,792. Most of the last detailed backlog was unfunded or priced options, customer concentration is high and management is spending ahead of uncertain contract and acquisition returns. A low revenue multiple is not necessarily cheap when recurring margins remain close to zero and acquisition funding could reopen dilution risk.

21 Scenario framework after Q2 2026

22 Merlintrader bottom line

Castellum’s Q2 is mixed, and describing it as simply strong or weak would miss the structure. The balance sheet improved again: cash rose to $16.865 million, debt stayed at zero and the common-share count was effectively unchanged. First-half revenue grew 9.6%, backlog stayed near $272 million and the qualified pipeline reached $953.5 million. Those facts support the view that the cleanup is real and the company has enough financial flexibility to invest.

The operating evidence moved the other way. Q2 revenue was down 1.1% year over year, gross margin fell to 34.0%, the operating loss widened to $1.094 million and adjusted EBITDA fell to $25,792. Management’s explanation is plausible but not yet sufficient: fixed-price completion costs may be temporary, while a subcontractor-heavy mix and structurally higher operating investment may persist.

The next proof points are unusually specific. The August 7 call needs to separate temporary costs from normalized economics. The 10-Q needs to show the June 30 funded-backlog split, cash-flow detail, contract-type mix and customer concentration. Q3 then needs to demonstrate that a debt-free platform with a large backlog can produce not only record revenue, but better gross profit and recurring earnings per share. Until those lines improve together, CTM remains an execution story rather than a completed turnaround.

Related Research On Merlintrader

• Castellum ($CTM): FY2025 results and 2026 execution — the annual numbers when they were first filed.
• Castellum ($CTM) is now debt free — the February 2026 repayment in context.
• Castellum ($CTM): 2025 repaired the story — how the cleanup year was executed.
• Space, Defense & AI Stock Hubs 2026: the new infrastructure race — the full index of company hubs in this sector.
• Dilution, ATMs and PIPEs: how equity funding actually works — background for the capital structure section above.
• BigBear.ai ($BBAI) Stock Hub — another small-cap federal technology contractor with a very different capital structure.
• Weekly Market Pulse — the week ahead across catalysts and earnings.