Robotics ETFs: AUM, Holdings, and Whether They Actually Beat the Index

01Why buy the theme

Picking the single robotics company that ends up winning is hard — this site's other 12 segment articles are full of disputed shipment counts, financially distressed public companies, and market-size estimates that vary by 100x depending on who's counting. A thematic ETF sidesteps that specific-company risk by holding dozens of names across the robotics and automation value chain at once, from industrial-arm makers to chip suppliers to pure-play robotics startups, betting on the theme playing out somewhere in the basket rather than in any one name.

02How the category formed

ROBO Global's Robotics & Automation Index ETF, launched October 22, 2013, was the first fund built specifically around this theme. ARK Invest's Autonomous Technology & Robotics ETF (ARKQ) followed in September 2014, taking a more concentrated, disruptive-innovation approach. Global X's Robotics & Artificial Intelligence ETF (BOTZ), launched September 2016, leaned into industrial-automation hardware names. iShares' Robotics and Artificial Intelligence ETF (IRBO), launched in 2018, took a different turn entirely in August 2024: its underlying index changed and the fund was rebranded the iShares Future AI & Tech ETF (ticker ARTY), shifting its holdings toward generative-AI, chip, and data-infrastructure names rather than robotics specifically — effectively exiting this category despite still holding more assets than every fund below except BOTZ. The most recent wave arrived in June 2025, when KraneShares launched its Global Humanoid Robotics and Physical AI ETF (KOID) and Roundhill launched its Humanoid Robotics ETF (HUMN) within weeks of each other — each marketing itself as the first US humanoid-robotics ETF, a direct conflict between the two issuers. WisdomTree's Physical AI, Humanoids and Drones Fund (WDRN) followed in May 2026.

03Fund flows

No research firm publishes a dedicated "robotics ETF market size" series, so the closest available proxy is Bloomberg's tracking of net inflows into AI-themed ETFs broadly: roughly $4.2 billion in 2024, surging to roughly $19 billion in 2025. That figure spans a much wider category than robotics-specific funds — most of the growth reflects the broader generative-AI investment wave — but it's the clearest available signal of how fast capital has been moving into thematic AI/robotics exposure as a category.

Data note: No research firm publishes a dedicated "robotics ETF market size" series, so the marketSize figures above use Bloomberg's net-inflow data for AI-themed ETFs broadly — a real, dated, sourced series, but a materially wider category than robotics-specific funds, presented here as the closest available proxy for capital-flow momentum into this investment theme rather than a precise measure of this segment alone. The marketShare entries deliberately exclude the iShares Future AI & Tech ETF (formerly IRBO, ticker ARTY), which held over $3.7 billion in assets as of mid-2026 — larger than every fund listed above except BOTZ — because the fund's underlying index changed in August 2024, shifting its holdings toward generative-AI, chip, and data-infrastructure names rather than robotics; including its current AUM in a "robotics fund" comparison would overstate the segment's real scale. KOID (KraneShares) and HUMN (Roundhill), both launched in June 2025, each separately market themselves as "the first" US humanoid-robotics ETF — a direct marketing conflict between the two issuers that this data doesn't attempt to resolve.

Market data as of 2026-07

04Six funds by size

Global X's BOTZ leads the group by assets under management at roughly $3.43 billion, followed by ARK's ARKQ at $2.88 billion and ROBO Global's namesake fund at roughly $1.97 billion. First Trust's ROBT trails further behind at $714.9 million, and the newest humanoid-focused entrants are smaller still: KraneShares' KOID at $271.9 million and Roundhill's HUMN at roughly $83.5 million — both still under two years old as of this writing. For scale, iShares' ARTY (the former IRBO) holds more than any fund on this list except BOTZ, but is excluded from this comparison for the reasons explained above.

05Holdings barely overlap

The three largest funds express genuinely different investment theses, and their top-10 holdings barely overlap. BOTZ is the most industrial-hardware-heavy, led by Keyence, ABB, and NVIDIA at roughly 9% each. ARKQ is the most concentrated in mega-cap "disruptive innovation" names — Tesla alone is roughly 11% of the fund, alongside AMD, SpaceX, and Palantir. ROBO takes the opposite approach: a broader, more equal-weighted basket where its top 10 holdings combined account for only around 17% of the fund, versus roughly 60% for BOTZ's top 10. Only a handful of names — Intuitive Surgical, SMC Corp, and Teradyne — appear in more than one fund's top holdings at all; no single stock appears in all three funds' top 10 simultaneously.

06Performance

The case for a thematic fund only holds up if it actually beats a simple, low-cost index fund over time — the results here are uneven. ARKQ's five-year annualized return trailed the S&P 500 by a meaningful margin (roughly 9.3% versus 15.2%, per Schwab/Morningstar data), even as its one-year return surged. BOTZ has tracked close to its own benchmark index rather than clearly beating the broad market. ROBO's own reported returns have been particularly volatile across different reporting periods and data sources, and trade press coverage describes the fund as having historically lagged the S&P 500 for much of the period since 2022 before a sharp 2025–2026 surge tied to the humanoid-robot investment wave. None of this means the theme itself is wrong — it means picking the right thematic fund, and the right entry point, has mattered as much as picking the theme.

07Challenges ahead

Two open questions sit side by side in this segment. The investment question is whether any of these funds, over a full market cycle, actually earns investors a premium for taking on thematic concentration risk rather than just buying a broad index — ARKQ's underperformance against the S&P 500 over five years is a real data point against the thesis, even as 2025–2026 returns have been strong. The product question is smaller but more immediate: KOID and HUMN both market themselves as the first humanoid-robotics ETF, launched within weeks of each other in June 2025 — a genuine dispute this site's data doesn't resolve, and a sign of just how fast issuers are racing to package the humanoid-robot narrative into a tradeable product.

08Timeline at a glance

09FAQ

10Sources