By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panelists generally agree that Cathie Wood's sale of $40.6 million of Taiwan Semiconductor (TSMC) stock was likely a tactical rebalancing move rather than a bearish verdict on the AI supply chain. However, they express concern about ARKK's repeated rotation into illiquid private names and potential liquidity traps, which could be forcing sales and deteriorating the fund's risk profile.
Risk: Liquidity traps and forced sales of high-quality assets to fund redemptions while pivoting into illiquid, speculative bets, which could rapidly deteriorate the fund's risk profile.
Opportunity: Rotation into higher-beta, specialized AI hardware for potential 'innovation' alpha.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Cathie Wood, chief of Ark Investment Management, is known for actively trading her holdings, sometimes selling stocks during sharp market pullbacks.
Semiconductor stocks experienced a major market pullback on May 15, with the iShares Semiconductor ETF (SOXX) dropping roughly 4%. Amid that sell-off, Wood sold millions of dollars worth of shares in one chipmaker company.
In 2025, Wood’s flagship Ark Innovation ETF gained 35.49%, far outpacing the S&P 500’s return of 17.88% in the same period. So far this year, Wood’s flagship Ark Innovation ETF (ARKK) was down 3.81%, trailing the S&P 500’s gain of more than 8%.
Wood gained a reputation after the Ark Innovation ETF delivered a 153% return in 2020. But her style also brings painful losses in bearish markets, as seen in 2022, when the Ark Innovation ETF tumbled more than 60%.
Those swings have weighed on Wood’s long-term gains. As of May 7, the Ark Innovation ETF has delivered a five-year annualized return of -6.25%, while the S&P 500 has an annualized return of 13.80% over the same period, according to data from Morningstar.
Cathie Wood expects “great acceleration” brought by tech developments
Wood focuses on high-tech companies across artificial intelligence, blockchain, biomedical technology, and robotics. She thinks these businesses have strong growth potential, though their volatility often causes fluctuations in the Ark’s funds.
According to Morningstar analyst Bella Albrecht, two of Wood’s Ark funds were among the worst-performing ETFs in the first quarter of 2026. The Ark Next Generation Internet ETF (ARKW) ranked second on the list, while the ARK Innovation ETF placed fifth.
From 2014 to 2024, the Ark Innovation ETF wiped out $7 billion in investor wealth, according to an analysis by Morningstar’s analyst Amy Arnott. That made it the third-biggest wealth destroyer among mutual funds and ETFs in Arnott’s ranking. The analyst hasn’t updated the 2025 ranking.
In a March Bloomberg podcast, Wood says the global economy is not heading into a downturn, but into what she calls a “great acceleration” driven by AI and other breakthrough technologies.
“We’re not going into the Great Depression, we’re going into the great acceleration,” Wood said, pointing to how past technological revolutions reshaped economic growth.
She noted that global real GDP growth averaged just 0.6% between 1500 and 1900, before the Industrial Revolution lifted it to around 3% for more than a century. Now, she argues, a new wave of innovation could push growth much higher.
“We think [technologies] are going to take growth into the 7 to 8% range,” Wood said, adding that the number may actually be conservative.
Wood also emphasized that AI is rapidly driving down costs across industries.
“These technologies are deflationary,” she said. “AI training costs are dropping 75% per year, and inference costs are falling as much as 85% to even 98% annually.”
In an earlier letter published in January, Wood rejects the “AI bubble” talk, saying that it "is years away" and that "the most powerful capital spending cycle in history" is coming.
Some investors seem to agree with Wood’s optimism. Over the past five days through May 14, the Ark Innovation ETF saw roughly $1.48 billion in net inflows, according to data from ETF research firm VettaFi.
Cathie Wood sells $40.6 million of Taiwan Semiconductor stock
On May 14 and 15, Wood’s Ark funds sold a total of 100,549 shares of Taiwan Semiconductor Manufacturing Company (TSM), or TSMC, according to Ark’s daily trade information. These shares are valued at about $40.6 million.
Shares of TSMC, the world's largest dedicated contract semiconductor foundry and a key supplier of Nvidia (NVDA), are up roughly 35% year to date as investors bet on a much bigger semiconductor market and an expanding role for the company in the AI supply chain.
TSMC just revealed that it now sees the semiconductor market reaching $1.5 trillion by 2030, topping its previous forecast of $1 trillion, according to its presentation materials ahead of a tech symposium on May 14, Reuters reported.
The company has become a key manufacturer for the AI industry because most advanced AI chips cannot be mass-produced without Taiwan Semiconductor Manufacturing Company’s chipmaking and advanced packaging technology.
Among the $1.5 trillion 2030 chip market, TSMC sees 55% of demand tied to AI and high-performance computing.
In April, TSMC reported rosy first-quarter 2026 results. Revenue increased 35.1%, net income and diluted EPS both increased 58.3%, and gross margin reached 66.2%.
Those figures are well above Wall Street's expectations. The company also raised its full-year 2026 outlook to more than 30% revenue growth in U.S. dollars.
“AI-related demand continues to be extremely robust,” TSMC CEO C.C. Wei said on the Q1 earnings call.
“Our business in the first quarter was supported by strong demand for our leading-edge process technologies,” TSMC’s CFO Wendell Huang said in a press release. “Moving into second quarter 2026, we expect our business to be supported by continued strong demand for our leading-edge process technologies.”
Barclays raised its price target on TSMC to $470 from $450 and maintained an overweight rating on the shares after the earnings, The Fly reported.
The analysts said TSMC reported another quarter of" impressive delivery" with an earlier than expected guidance upgrade, the analyst wrote in a research note.
Barclays believes TSMC's AI demand momentum is set to continue, making the stock "a core holding."
TSMC is not in the top 10 holdings of Wood’s Ark Innovation ETF.
Top 10 holdings of the Ark Innovation ETF as of May 15, 2026:
Other than selling shares of TSMC, Wood’s recent trading activity included purchases of Cerebras Systems (CBRS), while selling Advanced Micro Devices (AMD), Twist Bioscience (TWST), Teradyne (TER), and CareDx (CDNA).
Four leading AI models discuss this article
"Wood’s TSM sale is portfolio housekeeping, not a bearish signal on AI-driven semiconductor demand."
Cathie Wood’s $40.6 million TSM sale on May 14-15 coincides with a 4% SOXX drop but occurs outside ARKK’s top-10 holdings, suggesting tactical rebalancing amid inflows rather than a sector call. TSM’s Q1 2026 results—35% revenue growth, 66.2% gross margin, and raised >30% full-year guidance—plus its $1.5 trillion 2030 market forecast underscore resilient AI demand that Wood herself cites as deflationary. ARKK’s five-year -6.25% annualized return versus the S&P 500’s 13.8% highlights the risk that frequent trading amplifies volatility without capturing sustained semiconductor upside.
The sale could reflect Wood trimming ahead of expected valuation compression once AI capex peaks or supply constraints ease, especially since TSMC’s 35% YTD gain already prices in much of the raised guidance.
"Wood's TSMC sale is less important than her simultaneous CBRS purchase—the rotation matters more than the exit, but her five-year underperformance suggests even correct thematic calls don't guarantee fund outperformance."
Wood's $40.6M TSMC sale during a 4% semiconductor pullback reads as tactical rebalancing, not conviction loss—especially given TSMC's 35% YTD surge and Barclays' $470 price target. More telling: concurrent purchases of Cerebras (CBRS), a smaller AI chip designer, suggest Wood is rotating INTO edge-case AI plays rather than fleeing semiconductors. The real signal isn't the exit; it's the entry. However, ARKK's -6.25% five-year annualized return versus SPX's 13.80% exposes a brutal structural problem: Wood's conviction in 'great acceleration' hasn't translated to alpha. The $7B wealth destruction (2014–2024) suggests her thematic timing, not stock-picking, is the issue.
If Wood's thesis on AI-driven growth is correct, why is she trimming the most obvious beneficiary (TSMC with 55% of 2030 demand tied to AI) while the stock is up 35% YTD? Profit-taking before a correction could signal she sees valuation risk that Barclays' overweight rating misses.
"Wood is rotating from foundational foundry exposure into speculative AI hardware to capture higher growth potential, despite the structural stability of $TSM."
Wood’s exit from $TSM while simultaneously buying $CBRS suggests a pivot toward higher-beta, specialized AI hardware rather than a bearish view on the sector. While $TSM’s 66% gross margin and 30%+ growth are stellar, Wood likely views it as a 'utility' play that has already re-rated. By rotating into $CBRS, she is chasing the 'innovation' alpha she requires to justify her fund's high volatility. However, this ignores the geopolitical risk premium inherent in $TSM; her sale might be less about sector sentiment and more about risk-adjusting her portfolio against potential Taiwan-related supply chain shocks that could derail her 'great acceleration' thesis.
Wood’s sale may simply be a liquidity management necessity to fund redemptions or rebalance from a 'safe' semiconductor bellwether into speculative names that fit her narrative-driven investment style.
"ARK's TSMC disposition appears to be routine liquidity/rebalancing rather than a fundamental change to the AI growth thesis."
ARK's sale of $40.6 million of Taiwan Semiconductor (TSMC) stock reads more like portfolio housekeeping than a bearish verdict on the AI supply chain. Ark funds have a diversified set of AI bets, and TSMC isn’t even among its top 10 holdings, suggesting the trade is a liquidity or rebalancing move rather than a structural tilt. The article reasonably highlights AI optimism and TSMC's solid Q1 2026 results, but it glosses over why Ark would trim a Taiwan-linked name amid a broader semiconductors rally. Missing context: Ark's cash needs, timing of redemptions, and what replaces the cash in the portfolio.
The sell could be signaling hidden concerns about geopolitical risk or a shift in valuation, and if other funds follow, it could weigh on semis even as AI demand remains robust.
"ARKK's pattern of swapping TSMC for speculative privates like Cerebras amplifies volatility instead of managing geopolitical or valuation risk."
Gemini flags geopolitical risk in TSMC but misses how ARKK's repeated rotation into illiquid private names like Cerebras has compounded its -6.25% five-year lag versus the S&P. Those entries often coincide with redemption pressure, turning tactical sales into forced liquidity events that widen tracking error rather than hedge Taiwan shocks. The 35% YTD gain in TSMC already embeds much of the AI upside Wood cites, so the move looks more like performance chasing than risk adjustment.
"ARKK's underperformance and tactical trading are separate problems; the real question is whether redemptions are forcing illiquid rotations that worsen both."
Grok conflates two separate issues: ARKK's five-year underperformance versus tactical rebalancing efficacy. The -6.25% lag reflects thematic timing misses (2020–2022), not necessarily this TSMC trade. More pressing: nobody's quantified whether ARKK's redemption pressure is *forcing* sales or enabling them. If redemptions are accelerating, Wood's rotation into illiquid names becomes a liquidity trap, not alpha pursuit. That's the structural risk worth isolating.
"ARKK is replacing high-quality, liquid AI infrastructure exposure with illiquid, speculative assets, creating a structural liquidity trap that exacerbates fund volatility."
Claude, you hit the nail on the head regarding the liquidity trap. If ARKK is forced to sell liquid, high-quality assets like TSM to fund redemptions while pivoting into illiquid, speculative bets like Cerebras, the fund’s risk profile is deteriorating rapidly. This isn't just tactical rebalancing; it’s a structural degradation of the portfolio’s quality. Wood is effectively swapping institutional-grade semiconductor exposure for high-beta, unproven paper, which only compounds the volatility that has already decimated long-term shareholder value.
"Redemption-driven liquidity risk exists, but ARKK’s cash runway/redemption cadence create liquidity- and valuation-mismatch, not just rotation into private bets."
Claude, framing ARKK’s move as a pure redemption trap presumes redemptions are the sole driver. In reality, the implied risk is a liquidity- and valuation-mismatch: ARKK can fund redemptions only if liquid assets cover it, otherwise forced sales of large names (TSMC) can coincide with illiquid private bets (CBRS) and widen tracking error. The missing link is ARKK’s actual cash runway and redemption cadence, not just rotation talk.
The panelists generally agree that Cathie Wood's sale of $40.6 million of Taiwan Semiconductor (TSMC) stock was likely a tactical rebalancing move rather than a bearish verdict on the AI supply chain. However, they express concern about ARKK's repeated rotation into illiquid private names and potential liquidity traps, which could be forcing sales and deteriorating the fund's risk profile.
Rotation into higher-beta, specialized AI hardware for potential 'innovation' alpha.
Liquidity traps and forced sales of high-quality assets to fund redemptions while pivoting into illiquid, speculative bets, which could rapidly deteriorate the fund's risk profile.