--- title: "Nasdaq robot fund RoboStrategy sells Figure AI's $39 billion mark on numbers no shareholder can audit" description: "A Nasdaq fund sells retail access to Figure AI, the $39 billion humanoid maker, on numbers its own executives supplied. The checkable record, a confirmed week of 55 robots against a 12,000-unit 2026 target, does not match the pitch." publisher: "The Inference" section: "Robotics & Defense" published: 2026-10-24T19:10:04.685Z modified: 2026-10-24T19:10:04.685Z canonical: https://theinference.org/article/nasdaq-robot-fund-robostrategy-sells-figure-ai-s-39-billion-mark-on-numbers-no-shareholder-can-audit language: en keywords: "Figure AI, RoboStrategy, Humanoid Robots, Valuations, Closed End Funds, Disclosures" --- # Nasdaq robot fund RoboStrategy sells Figure AI's $39 billion mark on numbers no shareholder can audit > A Nasdaq fund sells retail access to Figure AI, the $39 billion humanoid maker, on numbers its own executives supplied. The checkable record, a confirmed week of 55 robots against a 12,000-unit 2026 target, does not match the pitch. [Andrew Kang](https://theinference.org/markets/people/andrew-kang), chief executive of RoboStrategy (Nasdaq: BOT), a closed-end fund that holds private robot makers so retail investors can buy the humanoid boom through its shares, has a simple pitch: robots are at the "GPT-3 stage," and [Figure AI](https://theinference.org/markets/companies/figure-ai) is the field's [Apple](https://theinference.org/markets/companies/apple) [1]. He [told Maeil Business on October 24](https://www.mk.co.kr/en/stock/12184851) that Figure is "already making hundreds of robots a month" after a visit to its Sunnyvale headquarters [1]. The fund prices a stake in Figure, a private company valued at $39 billion in a Series C last September, on confidence like that [2]. The checkable record tells a narrower story. ## The confirmed rate falls far short of the target What can be verified from company materials: BotQ, Figure's manufacturing platform, [lifted its Figure 03 assembly rate 24-fold in 120 days and confirmed 55 units built in a single week](https://www.bytepreview.com/en/articulos/figure-ai-from-a-70-million-series-a-to-a-39-billion-valuation) [2]. Figure targets 12,000 units of annual capacity by 2026 [2]. Kang's "hundreds of robots a month" claim, the kind of figure that anchors the fund's narrative, sits between those two numbers and matches no published figure [1]. His estimate that all unlisted physical-AI companies together are worth about $200 billion is likewise a founder's arithmetic, not a filing [1]. *Chart: **Figure's confirmed output trails its 12,000-unit 2026 target** Bars comparing Figure AI's confirmed production of 55 units in one week, equal to an annualized run rate of about 2,860 units, its stated 12,000-unit 2026 annual capacity target, and Tesla's Fremont Optimus line designed for 1 million robots a year.* *Units per year. Figure confirmed 55 units in one week (55 x 52 = about 2,860 annualized, shown as an estimate derived here) and targets 12,000 units of annual capacity by 2026, both from BytePreview. Tesla's Fremont line is designed for 1 million robots a year, per Musk on the Q3 call reported by Yahoo Tech/Cult of Mac, a plan figure, not output; no verified Optimus unit count exists. Kang's 'hundreds of robots a month' claim matches no published figure. [2][5]* ## The proof points are self-reported too The treadmill milestone the press relayed as progress carries the same stamp. Figure's CEO posted that [the Figure 03 walked 348 miles and 1.2 million steps in nine days](https://www.techrepublic.com/article/news-figure-03-humanoid-robot-348-mile-endurance-test/); the [footage shows a cable attached](https://theinference.org/article/figure-s-founder-says-its-robot-walked-348-miles-on-a-cord-the-39-billion-mark-rests-on-rented-robots), its purpose undisclosed, and no independent confirmation of uninterrupted operation [3]. The household trial Figure cites, 237 of 420 tasks completed in 30 unseen homes, is the company's own account [3]. Buyers of BOT are paying for pre-IPO exposure they cannot audit. The fund's [net assets were $277.7 million as of September 30, $11.35 per share](https://www.nasdaq.com/press-release/robostrategy-inc-updates-net-asset-value-september-30-2026-2026-10-05), up 0.36% in a month, with no per-holding marks disclosed [4]. If those private marks are overstated, early holders who marked the book up benefit and shareholders of record absorb the loss; if they are right, the same shareholders collect. The disclosure asymmetry is the point. ## The counterpoint is real, and so is the clock The bull case survives. Figure raised more than $1 billion from [Nvidia](https://theinference.org/markets/companies/nvidia), [Intel](https://theinference.org/markets/companies/intel) Capital and [Qualcomm](https://theinference.org/markets/companies/qualcomm) Ventures at that $39 billion mark, and [its warehouse demo ran a rotating fleet](https://theinference.org/article/bmw-counted-figure-robots-one-at-a-time-the-valuation-is-priced-on-100-000) for some 200 hours [2][3]. Kang also stresses that private robot companies stay private longer, so funds like his are the only listed door [1]. But the door now has a competitor. [Tesla](https://theinference.org/markets/companies/tesla) [began "early production" of Optimus at Fremont on October 21](https://tech.yahoo.com/general/articles/tesla-starts-optimus-production-elon-215423739.html), with external deliveries targeted for late 2027, a line designed for 1 million robots a year, and no verified output number yet [5]. When a rival with a real production line starts counting, a $39 billion mark built on "hundreds of robots a month" has to be [earned in audited units](https://theinference.org/article/figure-committed-3-5-billion-to-compute-for-a-robot-that-fails-44-of-its-trials) [1][2]. Anyone holding BOT should not buy or sell on the pitch; the question that decides it is whose counts the fund is holding, and who signs them. ## Takeaway RoboStrategy sells retail investors exposure to Figure AI at a $39 billion mark, but the only checkable production number is 55 robots in one week against a 12,000-unit 2026 target; the marks behind the fund's NAV come from the company's own executives, not any auditor. ## Sources 1. [Maeil Business (mk.co.kr), Andrew Kang interview, 24 October 2026](https://www.mk.co.kr/en/stock/12184851) 2. [BytePreview, Figure AI: from a $70 million Series A to a $39 billion valuation, 17 October 2026](https://www.bytepreview.com/en/articulos/figure-ai-from-a-70-million-series-a-to-a-39-billion-valuation) 3. [TechRepublic, Figure AI's Humanoid Robot Walks 348 Miles in Just 9 Days, 22 October 2026](https://www.techrepublic.com/article/news-figure-03-humanoid-robot-348-mile-endurance-test/) 4. [GlobeNewswire via Nasdaq.com, RoboStrategy, Inc. Updates Net Asset Value as of September 30, 2026, 5 October 2026](https://www.nasdaq.com/press-release/robostrategy-inc-updates-net-asset-value-september-30-2026-2026-10-05) 5. [Yahoo Tech (Cult of Mac), Tesla starts Optimus production, but Elon Musk has a long wait in store for buyers, 23 October 2026](https://tech.yahoo.com/general/articles/tesla-starts-optimus-production-elon-215423739.html) --- Nasdaq robot fund RoboStrategy sells Figure AI's $39 billion mark on numbers no shareholder can audit — The Inference. Canonical: https://theinference.org/article/nasdaq-robot-fund-robostrategy-sells-figure-ai-s-39-billion-mark-on-numbers-no-shareholder-can-audit