iShares Future AI & Tech ETF (formerly IRBO, iShares Robotics & AI Multisector ETF)
iShares (BlackRock)
Formerly IRBO tracking a broad, roughly equal-weight robotics/AI index; renamed and re-indexed to ARTY (iShares Future AI & Tech ETF) tracking the Morningstar Global Artificial Intelligence Select Index (~Aug 2024) — now a more concentrated, chip/AI-infrastructure-tilted basket (~68 holdings).
Top holdings
Memory/HBM, top holding
AI accelerators
AI cloud/GPU infrastructure
Foundry
Custom AI silicon
Equal-weighted, not cap-dominant
Performance
July 2026: 12.49% (Jun 30 close $76.16 -> Jul 31 close $66.65; no distribution in the window) — Peak-to-trough -19.5% into the Jul 29 low ($61.33), then a 2-session bounce of +8.7% that did not recover the month. Still -17.5% below the Jun 2 YTD high ($80.75). ARTY (-12.5%) beat SMH (-17.6%) and SOX (-20.6%): its non-semi sleeve (MSFT +24.6%, AI software, NAVER) cushioned, while MU/MRVL/CRWV/SMCI took full damage at full weight.
Recent moves
The defining event is the IRBO→ARTY rebrand and index switch (broad robotics → concentrated AI infrastructure); AUM grew from a few-hundred-million robotics fund to ~$3.75B.
Our take
Note the identity change: the fund the ticker IRBO once named now trades as ARTY with a different, AI-infrastructure-tilted index — anyone screening on the old 'broad equal-weight robotics' description is looking at a fund that no longer exists. As re-indexed it is a near-equal-weight AI-chip-and-cloud basket competing directly with the semi ETFs.
The teaching point is the identity switch: on Aug 12 2024 BlackRock renamed IRBO (a broad equal-weight robotics/AI fund on the NYSE FactSet Robotics & AI Index) to ARTY and swapped it onto Morningstar's Global AI Select Index.
Anyone screening on the old 'robotics' description — or reading the 14.49% since-inception average annual (inception June 2018) or a 3-year number — is looking at a track record that blends two different strategies. Treat only post-Aug-2024 performance as this fund's own. Second point: ARTY is engineered flat. Top holdings sit AMD 5.2 / NVDA 5.1 / TSM 5.0 / AVGO 4.9 / MU 4.6, top-10 just ~41% (as of Jul 19 2026). That dampens the single-name torque that drives SMH (~21% NVDA) — you own the AI-infra theme broadly but no one winner carries you. Third: it is genuinely global — TSM (Taiwan) and NAVER (Korea) sit in the top 10, adding FX and Asia-policy exposure most US buyers of an 'iShares AI' fund don't expect. Net: a diversified AI-infrastructure basket that competes head-on with the semi ETFs, at a fee edge over thematic-AI peers.
- 2026-07Top weights at 31 Jul 2026 read Broadcom 5.16, TSMC 5.15, Nvidia 5.13, AMD 4.79, Micron 4.27. Micron left the top five after -28.7% in July; the ~41% top-10 and 5.2% cap still hold.
Thesis
ARTY is a near-equal-weight, globally-diversified basket of ~65 AI-infrastructure names (chips, cloud, AI software) tracking Morningstar's Global AI Select Index — deliberately spreading risk so no single name tops ~5.2%, the structural opposite of NVIDIA-concentrated semi funds. The catch buyers miss: this is a different fund than the ticker's history implies.
Passive index basket, re-indexed Aug 12 2024 from the old IRBO robotics index to Morningstar's Global AI Select Index. ~65 holdings, near-equal-weight (top 10 ~41%), spanning US + non-US AI compute, data/infrastructure, software and services. 0.47% fee. Rebalanced/reconstituted on the index schedule; not leveraged, no daily reset.
Assessment
- Fee edge: 0.47% undercuts most thematic-AI peers (AIQ 0.68%, CHAT/ARKQ 0.75%, ARKW 0.88%)
- Near-equal-weight (top-10 ~41%, no name >5.2%) minimizes single-stock blow-up risk vs SMH
- Genuine global reach — TSM, NAVER, SK-complex names diversify beyond US megacaps
- BlackRock scale/liquidity; $3.61B AUM, passive rules-based, no manager-discretion risk
- Track-record discontinuity: since-inception 14.49% and any 3yr figure span the pre-Aug-2024 robotics strategy — not this fund
- 'Diversified' only within one sector — effectively 100% AI/semiconductor-infrastructure, a single cyclical theme
- Equal-weight dilutes the NVIDIA torque: structurally lags SMH when one megacap leads the tape
- High-beta AI-capex names (CRWV, SMCI, ORCL) leave it exposed to any datacenter-spend slowdown
- Global holdings add Taiwan/Korea geopolitical + currency risk under an 'iShares' label many read as US
- 2026-07July 2026 ran against that mechanism: Nvidia +0.3% and Broadcom +3.1% led the tape, yet ARTY -12.5% beat SMH -17.6% and SOX -20.6%, cushioned by its non-semi sleeve (Microsoft +24.6%, software, NAVER).
Record
1-year total return +55.50% (NAV incl. dividends, as of Jul 20 2026, stockanalysis) — driven by the semiconductor/AI-infrastructure book (AMD, NVDA, TSM, AVGO, MU, Marvell, CoreWeave) riding 2026's AI-chip rally. Because the fund is near-equal-weight, it captured the broad rally but without the concentrated NVIDIA lift that pushed SMH higher on the way up. Interpret longer windows with care: the 3-year and since-inception (14.49% avg annual since June 2018) numbers blend the pre-Aug-2024 robotics strategy, so they describe a different portfolio, not ARTY's own record. A clean July YTD figure was not confirmable at check time (May 2026 readings ranged ~28-34%). Not a leveraged or daily-reset vehicle, so no volatility-decay mechanics apply — path dependence here is ordinary equity beta, not compounding drag.
- 2026-07A point YTD figure is now available in place of the range: +38.4% to 31 Jul 2026, reproduced independently as +38.3% from the 31 Dec 2025 close of $48.18 to the $66.65 close.
Risks & fit
- Index-switch discontinuity makes 3yr/since-inception stats misleading as a forward guide
- Single-theme cyclicality: an AI-capex or semi-cycle downturn hits nearly the whole book at once
- Crowded AI-infrastructure trade; high-multiple names (CRWV, SMCI) swing hard on sentiment
- Taiwan/Korea exposure adds geopolitical + FX risk beneath a US-brand wrapper
- Equal-weight structurally trails if mega-cap leadership (NVDA) persists
The 'broad, equal-weight AI basket beats concentrated bets' logic breaks if AI-infrastructure returns stay driven by one or two megacaps (NVIDIA/Broadcom): then ARTY's ~5% caps dilute the winners and it durably trails SMH/SMH-like concentrated funds. Conversely, the thesis is confirmed if AI leadership broadens across chips, cloud and software — where ARTY's breadth captures more of the move than a top-heavy peer. Watch relative performance vs SMH and AIQ across a full up-and-down cycle, not a single trending leg.
Someone wanting one-ticker, diversified AI-infrastructure exposure who specifically does NOT want single-name NVIDIA concentration, will accept global (Taiwan/Korea) names and full AI-sector cyclicality, and values a below-average thematic-AI fee. Not a fit for anyone seeking the maximum-torque AI-GPU bet (that's SMH/single-stock leveraged names) or a truly diversified core holding — it remains 100% one theme.
0.47% net expense ratio — below most thematic-AI ETFs (AIQ 0.68%, CHAT 0.75%, ARKW 0.88%) but well above broad-tech index funds (XLK ~0.09%, QQQ 0.20%). A real edge within the AI-theme category, not versus plain tech beta.