The standard BBAI stock prediction treats BigBear.ai as a cheap defence-AI growth story that the market has overpunished, and at $2.82 the chart looks like capitulation: the shares closed on 10 September 2026 just 8.9% above their 52-week low of $2.59 and 47.8% below where they ended 2025. The filings tell a less comfortable story. BigBear.ai’s headline 13% second-quarter growth came entirely from an acquisition, a day after reporting it the company opened a 100-million-share at-the-market programme worth roughly a fifth of its share count, and it is holding about $0.82 a share in net cash it does not need for survival. Our 12-month levels are a $4.10 bull case, a $2.70 base case and a $1.75 bear case, and the stock’s direction depends less on AI sentiment than on what management does with that new share supply.
Here is the number nobody in the bull-case videos is quoting. Strip out Ask Sage, the generative-AI platform BigBear.ai bought for $272.1 million on 31 December 2025, and the rest of the business shrank. According to the Q2 2026 Form 10-Q, Ask Sage contributed $6.1 million of the quarter’s $36.7 million in revenue, which leaves $30.7 million from the legacy company against $32.5 million a year earlier: a 5.6% decline. For the half-year the legacy decline is 12.3%. On the company’s own pro forma basis, which counts Ask Sage in both years, first-half revenue grew 1.0%. That is the business the ATM is now funding, and it is why our base case sits slightly below spot rather than above it.
Key facts
- Q2 2026 revenue was $36.7 million, up 13% year on year, with gross margin at 32.8% — BigBear.ai Q2 results (Exhibit 99.1), 30 July 2026
- Ask Sage supplied $6.1 million of Q2 revenue; ex-Ask Sage revenue fell 5.6% — Form 10-Q, 30 July 2026
- Total available liquidity was $409.8 million against $17.7 million of 2026 convertible notes — Form 10-Q, 30 July 2026
- A new ATM with Jefferies allows up to 100,000,000 shares, about 20.9% of the 479.5 million outstanding — Form 8-K, 31 July 2026
- Backlog was $269.6 million, of which only $79.4 million (29%) was funded — Form 10-Q, 30 July 2026
- Full-year 2026 revenue guidance of $135 million to $165 million was affirmed — Q2 results, 30 July 2026
- The two analysts covering the stock both rate it Hold with $4.00 targets, cut from $5.00 on 31 July — stockanalysis.com forecast page, checked 11 September 2026
What is happening to BigBear.ai stock, and why
BigBear.ai closed at $2.82 on Thursday 10 September, down 0.35% on the day, with pre-market indications at $2.83 early on Friday. We anchor every level in this piece on the regular-session close, pulled from stockanalysis.com at 08:03 UTC on 11 September. Over the past year the shares have fallen 40.4%, from $4.73 to $2.82. The highest close of that window was $8.91 on 14 October 2025, which means the stock now trades 68% below its peak close.
The slide has been a steady de-rating as the market worked out what BigBear.ai actually earns. Full-year revenue has gone $155.2 million in 2023, $158.2 million in 2024 and $127.7 million in 2025, according to the company’s 2025 annual report. Even the midpoint of this year’s guidance, $150 million, would leave revenue below where it was two years ago, and it includes a full year of an acquired business.
The quarter itself was better than the tape suggests. Gross margin widened by 781 basis points to 32.8%, the net loss narrowed to $25.7 million from $228.6 million (most of the 2025 figure was non-cash derivative marks and a goodwill write-down), and management cited more than 20 new contract wins. The problems sit lower down. Selling, general and administrative expense rose to $31.8 million, or 87% of revenue, and adjusted EBITDA worsened to a loss of $11.6 million from $8.5 million. Operating cash outflow was $22.2 million in the quarter and $40.2 million for the half.
Having tracked BigBear.ai’s filings since the stock’s 2025 run, the pattern I watch is the gap between the press-release headline and the note disclosures. In Q2 that gap was the acquisition. The 10-Q explains the half-year revenue move as “the inclusion of Ask Sage offset by lower volume on the Army programs and significant one time contracts” that did not recur. The Army exposure matters: the company’s largest customer contributed 16% of revenue in both periods, and the four customers above the 10% threshold together accounted for 51% of Q2 revenue.
How BigBear.ai management is responding
Management’s answer to slow organic growth is deal-making, and it has said so plainly. “The second half of 2026 is all about execution discipline and positioning ourselves for accretive, catalytic M&A and building momentum for even stronger topline growth in 2027,” said Kevin McAleenan, CEO of BigBear.ai, in the second-quarter release. He added that the company is “on track for our target of 17% revenue growth, and we intend to accelerate.”
The finance chief framed the balance sheet the same way. “We have steadily been maturing the underlying financial discipline of the company and have significant cash in reserve so that when the right opportunity presents itself, we can move fast,” said Sean Ricker, CFO of BigBear.ai, in the same document.
The next day, 31 July, the company signed an Open Market Sale Agreement with Jefferies covering up to 100 million new shares, with the agent paid up to 3% of gross proceeds. The accompanying prospectus supplement says the money is for “working capital and general corporate purposes” and that a portion may go to “strategic investments in complementary products or technologies.” It also discloses net tangible book value of $0.84 a share at 30 June.
Read those three documents together and the sequence is clear. BigBear.ai does not need equity to stay solvent: at the first-half operating burn rate of roughly $80 million a year, its $392 million of net cash covers close to five years. The ATM is acquisition currency. That can be good for shareholders if the targets are bought cheaply and grow, as the CargoSeer customs-AI purchase for $5.0 million in January might. It is bad if the company pays another Ask Sage-sized price with stock issued near a 52-week low.
Since the ATM went live, the news flow has been thin. The company’s investor-relations press page lists one release between the 30 July results and 11 September: the appointment of retired Lt. Gen. Sean Gainey to the board. His background, confirmed in the 17 August Form 8-K, includes serving as a three-star commander at US Army Space and Missile Defense Command and four years as director of the Defense Department’s Joint Counter Unmanned Aerial System Office. It is a credible hire for a defence-AI business. It is not a contract.
Insiders have not been adding. We reviewed every Form 4 BigBear.ai insiders filed in 2026 and found no open-market purchases. The filings show share withholding for taxes, equity grants and a handful of sales, including 10,000 shares sold by CFO Sean Ricker at $4.33 on 22 May and 80,000 shares sold by director Pamela Braden at $4.00 in March.
BBAI market data: the valuation the market is paying
At $2.82 and 479.5 million shares, BigBear.ai’s market value is about $1.35 billion. Net of the $17.7 million of convertible notes that mature on 15 December 2026, the company holds $392.1 million in cash and investments, which puts enterprise value near $960 million. On the $150 million guidance midpoint that is 6.4 times sales; on the $135 million low end it is 7.1 times. That is a premium multiple for a business with 32.8% gross margins and a shrinking legacy book, and a steep discount to the roughly 30 times 2025 sales the market paid at the October 2025 peak (on the year-end share count).
The chart shows why the bear level is not far-fetched. The stock has made lower highs since October, and the late-May rally to a $5.34 close on 1 June unwound within about a week. The $2.59 low on 29 July came the day before results; the shares then bounced to $3.27 on 7 August before drifting back.
| Metric (30 June 2026 unless stated) | Value | Why it matters |
|---|---|---|
| Shares outstanding | 479.5 million | Up from 437.0 million at end-2025, mostly from note conversions |
| New ATM capacity | 100 million shares (20.9%) | About $274 million net at $2.82 after 3% commission |
| Net cash | $392.1 million ($0.82/share) | Roughly 29% of the market value |
| Enterprise value / 2026 guidance midpoint | 6.4x | 7.1x on the $135 million low end |
| Legacy revenue growth, Q2 (ex-Ask Sage) | -5.6% | Headline +13% is acquired |
| Funded backlog | $79.4 million of $269.6 million | 55% of backlog is priced options the government may never exercise |
| Operating cash burn, H1 | $40.2 million | Close to five years of runway without new equity |
| Analyst targets (2) | $4.00, both Hold | Northland and Cantor Fitzgerald, cut from $5.00 on 31 July |
Backlog quality deserves a closer look. Funded backlog rose to $79.4 million from $54.9 million at the end of 2025, which is genuine progress. But $149.0 million of the $269.6 million total is “priced, unexercised options,” and the 10-Q itself notes that options “do not create enforceable rights and obligations until exercised.” Guidance needs $78.8 million of second-half revenue at the midpoint, or about $39.4 million a quarter, 7% above the Q2 run rate. The low end needs only $63.8 million, which the company would clear even if revenue slipped. The range is wide enough that “affirmed guidance” says little.
For context on how the market prices defence and government AI at scale, our Palantir (PLTR) bull and bear model covers the category leader, whose commercial growth is what BigBear.ai’s multiple is implicitly borrowing from. Other small-cap names that trade on backlog and dilution rather than earnings include IonQ and AST SpaceMobile.
The structural tension: dilution versus deal-making
The core problem for a BBAI price target is that per-share value is being pulled in two directions at once. Every share sold through the ATM at $2.82 brings in about $2.74 of cash, which is more than three times the $0.82 of net cash each existing share carries. On paper that is accretive to cash per share: the prospectus supplement calculates tangible book rising to $1.17 a share if all 100 million shares were sold at $2.83. What it cannot show is what happens next. If the cash is spent on acquisitions at six or seven times sales, it turns into goodwill, and the existing holder now owns roughly 17% less of a company whose value depends on whether that goodwill grows.
BigBear.ai’s history makes the market cautious. Share count has nearly doubled from 250.6 million on the November 2024 10-Q cover to 479.5 million, through at-the-market programmes, warrant exercises and note conversions. The 10-Q’s own account of the 2025 programmes is inconsistent: its management discussion says 142.3 million shares were sold for $300.0 million in the first half of 2025, while Note 16 gives 77.3 million shares for the same $300.0 million. That is the kind of disclosure detail that keeps institutional investors at arm’s length, and the coverage list has shrunk to two analysts. One of them, Cantor Fitzgerald, was the sales agent on BigBear.ai’s earlier ATM programmes, according to the 10-Q.
The $17.7 million of 2026 notes convert at $10.61 a share, far above spot, so they will be repaid in cash on 15 December. That removes the last debt and leaves equity as the only funding tool. If you run a similar exercise on Super Micro, the difference is stark: there the balance sheet question is working capital, not share supply.
Retail attention is thin. Our /last30days scan found five YouTube videos with about 12,000 combined views, mostly framing a recovery from “the high $2 range” toward $3.75 to $4.00; Reddit returned nothing qualifying and X was not searchable, so we give this little weight.
The call: BBAI stock prediction to September 2027
Our model values BigBear.ai as enterprise value (a multiple of 2027 revenue) plus projected net cash, divided by projected shares. Every scenario assumes about $85 million of cash burn over 12 months and roughly 10 million shares of equity compensation. The difference between the cases is how much of the ATM is used, at what price, and what the acquired and organic revenue is worth.
| Scenario | 12-month level | vs $2.82 spot | Probability | Key assumptions |
|---|---|---|---|---|
| Bull | $4.10 | +45% | 20% | ATM fully used at ~$3.60; ~$250m spent on acquisitions; 2027 revenue $190m; 10.6x EV/sales |
| Base | $2.70 | -4% | 45% | Half the ATM used at ~$2.80; 2027 revenue $170m; 6.0x EV/sales |
| Bear | $1.75 | -38% | 35% | ATM fully used at ~$2.10; 2027 revenue $140m; 3.7x EV/sales |
The probability-weighted value is about $2.65, roughly 6% below spot. That is not a dramatic call. The point is that at $2.82 the market is already paying 6.4 times sales for a company whose organic revenue is shrinking, so the burden of proof sits with the growth story.
Bull case: $4.10
The bull case needs management to deliver what McAleenan promised. Ask Sage keeps growing inside government accounts, the Gainey appointment helps win counter-drone and missile-defence software work, and a sizeable acquisition closes at a sensible price. If 2027 revenue reaches $190 million and the market pays 10.6 times sales, still a third of the October 2025 peak multiple, the stock reaches $4.10 even after 100 million new shares. That level sits just above both analysts’ $4.00 targets. What would make it more likely: a Form 8-K announcing a contract above $50 million, or a third-quarter report showing legacy revenue growing again.
Base case: $2.70
The base case assumes BigBear.ai hits the middle of guidance, grows about 13% in 2027 to $170 million, and uses half the ATM near today’s price. The multiple drifts to 6.0 times, close to where it is now. Per-share value barely moves because dilution and cash burn absorb the growth. This is the “dead money” outcome, and it is our most likely one.
Bear case: $1.75
The bear case is legacy erosion continuing while the ATM runs at lower prices. Revenue lands at $140 million, the Army programmes keep shrinking, and the market re-rates the stock to 3.7 times sales, closer to a government IT contractor than an AI platform. The floor is higher than it looks because the company would be sitting on about $0.87 a share of cash, but $1.75 is 38% below spot.
Invalidation and what would change my mind
The bull case is invalidated if third-quarter revenue comes in below $36 million or funded backlog falls back under $70 million. The bear case is invalidated if Q3 revenue reaches $40 million with the legacy business growing, or if the 10-Q cover page shows fewer than 20 million ATM shares sold. The single most useful data point will be that cover-page share count: on last year’s calendar the Q3 10-Q arrived on 10 November. If it shows heavy issuance below $3 with no acquisition announced, I would move the probability weight towards the bear case. A large, fairly priced acquisition would move me the other way.
FAQ
What is the BBAI stock prediction for the next 12 months?
Our model sets a $4.10 bull case, a $2.70 base case and a $1.75 bear case for BigBear.ai to September 2027, against a $2.82 close on 10 September 2026. We assign probabilities of 20%, 45% and 35%, which gives a weighted value of about $2.65. The outcome depends mainly on how many of the 100 million ATM shares are issued, at what price, and what the proceeds fund.
Why did BigBear.ai stock fall after second-quarter results?
The results themselves were mixed rather than poor: revenue rose 13% and gross margin widened to 32.8%. The pressure came from what followed. On 31 July the company opened an at-the-market programme for up to 100 million new shares, about 20.9% of its share count, and both covering analysts cut their targets to $4.00. Excluding the Ask Sage acquisition, second-quarter revenue fell 5.6%.
How much cash does BigBear.ai have?
BigBear.ai reported $409.8 million of cash and available-for-sale investments at 30 June 2026. Its only remaining debt is $17.7 million of convertible notes due 15 December 2026, which convert at $10.61 a share and will likely be repaid in cash. That leaves about $392 million of net cash, or roughly $0.82 a share, against a first-half operating cash burn of $40.2 million.
What is the BBAI price target from analysts?
According to stockanalysis.com, which aggregates S&P Global and TipRanks data, two analysts publish 12-month targets. Michael Latimore of Northland Securities and Jonathan Ruykhaver of Cantor Fitzgerald both rate the shares Hold, and both lowered their targets from $5.00 to $4.00 on 31 July 2026. That implies about 42% upside from $2.82, slightly below our $4.10 bull case.
Is BigBear.ai stock diluting shareholders?
Yes, and the capacity to do more is in place. Shares outstanding rose from about 437 million at the end of 2025 to 479.5 million at 30 June 2026, mainly from note conversions. The new Jefferies agreement allows up to 100 million more. Sales at current prices raise cash per share, but existing holders own a smaller slice, and value depends on how the cash is spent.
Disclaimer: This article is analysis for information purposes only and is not investment advice or a recommendation regarding any security. Scenario levels and probabilities are the author’s estimates based on public filings and may prove wrong. Investing in equities involves risk, and your capital is at risk.