BigBear.ai heads into its third-quarter report with a split personality. The top line finally started growing again last quarter, gross margins are sitting far above where they were a year ago, and management sounded more confident than it has in some time. Yet losses widened, cash kept shrinking, and the stock has lost nearly a quarter of its value since that report. The question for Tuesday's release, due after the close on October 27, is whether the operating momentum is real enough to justify the spending that is funding it.
Wall Street expects a loss of 5 cents per share on revenue of about $38.9 million. That would be a solid 17.5% revenue gain over the year-ago $33.1 million, and the per-share loss would narrow from 7 cents. But the estimate is a penny worse than last quarter's 4-cent loss, which tells you analysts expect the investment spending to keep running ahead of revenue for now. Sequentially, the forecast calls for roughly $2 million more revenue than the $36.7 million posted in the second quarter, a reasonable continuation of the reacceleration rather than a heroic leap.
Management has reaffirmed its 2026 revenue range of $135 million to $165 million on four straight calls without raising it. The Street's $146.6 million full-year estimate sits comfortably inside that band, slightly below the $150 million midpoint. With about $71 million booked in the first half, consensus implies roughly $75.5 million in the back half, and the third-quarter figure carries a lot of that load. A print near or above $39 million would make the midpoint look achievable and could finally give management room to tighten the range upward. A miss would revive the concern that the year depends on a fourth-quarter surge.
The previous call supplied plenty of catalysts that now need to show up in the numbers. More than 20 new contracts in the second quarter, a five-year CargoSeer customs deal in El Salvador, a NAVAIR generative AI win, and a new air-gapped hardware device for classified environments all point toward growth, but backlog actually slipped to $270 million from about $282 million in the first quarter. Renewed backlog growth would be the clearest sign that wins are outpacing revenue burn-off. Gross margin is the other key tell. It climbed to 32.8% on a richer mix of generative AI platform revenue, and holding that level or better would confirm the improvement is structural rather than a quarter of favorable mix.
The weak spots are equally clear. Adjusted EBITDA deteriorated to a loss of $11.6 million from $9.9 million a quarter earlier as sales and research spending rose, and there is no profitability guidance to anchor expectations. Cash and investments fell to $410 million, down from $715 million in the third quarter of 2025, and management signaled it intends to be aggressive on acquisitions. Any deal announcement will be judged on whether it adds revenue and margin rather than simply absorbing cash.
The market has delivered its verdict so far. Shares have fallen 22.9% since the last report while the S&P 500 gained 2.5%, and at $2.16 the stock sits barely above its post-earnings low of $2.12 and far below its 200-day moving average near $3.80. Sentiment is modestly bearish and essentially unchanged from last quarter. That sets a low bar emotionally, but not operationally. The narrative holds if revenue lands near $39 million, gross margin stays above 30%, backlog resumes growing, and losses stop widening. If revenue is merely in line while EBITDA slides again and cash drops further, investors will have little reason to believe the spending is buying durable growth.