Meme ETF soars 35% in the year but still suffers losses, sounding the alarm on the "valuation trap" in the AI speculation frenzy.

Investors chasing market themes are learning a familiar lesson: popular investments do not necessarily bring profits. The Roundhill Meme Stock ETF (MEME), which tracks retail favorite stocks, has risen by about 35% since 2026, with a 3.5% increase in a single day on Thursday, but as of the closing price on July 9th at $8.41, it is still down about 15% from its reissue price in October 2025. The fund manages assets of about $20 million, far below its initial level. This performance gap highlights a principle often overlooked during market frenzies: short-term returns can be driven by popularity and hype, but the longer you hold, factors such as profitability, competitive position, and purchase price take precedence. MEME ETF: Up 35% for the year but still in the red Since 2026, the Roundhill Meme Stock ETF (MEME) has risen by about 35%, with a 3.5% increase on Thursday, driven mainly by strong performances from component stocks such as AST Spacemobile, Terawulf, and Lumentum Holdings. The AI boom has not only fueled the surge of high-profit chip stocks, but also brought back speculative trading - leveraged ETFs are popular, small-cap stocks are volatile, and the "meme stock season" is back. The fund began trading on October 8, 2025, with an issue price of around $9.90. It was launched at the peak of the previous retail trading cycle, followed by market adjustments, causing MEME ETF to drop to a 52-week low of $5.33 at one point. Even with a 35% rebound in 2026, it is still about 15% below the issue price. The fund's structure magnifies this volatility risk. Its portfolio turns over nearly five times a year - one of the highest turnover rates on Wall Street - with close to 60% of assets concentrated in its top ten holdings. Holdings include AST Spacemobile, Terawulf, Lumentum Holdings, Bloom Energy, Applied Optoelectronics, and IREN Limited. Despite the eye-catching performance in 2026, investors who bought in at the beginning of issuance are still down about 15%, while the S&P 500 and Nasdaq indices have risen by about 12% during the same period. CEO of Roundhill Investments Dave Mazza said, "The fund's establishment coincided with the peak of the last retail cycle, which more reflects timing rather than whether these stocks can achieve strong growth as we have seen this year." Olga Bitel, Chief Investment Strategist at William Blair, emphasized, "Just because retail investors are heavily involved in these exciting companies and IPOs does not mean you shouldn't do your homework, find out what the company does, where it stands in the ecosystem, and whether it can deliver on its promises." From Cisco to AI giants: Lessons from overvaluation history, good companies good investments The performance gap of the MEME ETF reveals a principle that is often overlooked during market frenzies: short-term returns can be driven by popularity and hype, but the longer you hold, factors such as profitability, competitive position, and purchase price take precedence. Cisco Systems became the world's most valuable company at the peak of the internet bubble in 2000, but investors who bought near the stock's peak waited more than 25 years until May 2026 for the stock price to return to the highs of the internet bubble era. Cisco's stock price plummeted nearly 90% after the bubble burst. This lesson applies to the most hyped companies in the current market - including potential future listings such as SpaceX, and AI startups OpenAI and Anthropic. Now, AI startups Anthropic and OpenAI are testing investors' patience in a similar fashion. In a May Series H funding round, Anthropic was valued at $965 billion, with a secondary market valuation soaring to $1.2 trillion. Javier Avalos, co-founder of secondary trading platform Caplight, described Anthropic as "the most sought-after company in the history of venture capital secondary markets." Glen Anderson, CEO of Rainmaker Securities, stated, "The demand for Anthropic far exceeds the supply, making it difficult to make trades as there are no sellers." However, both companies have not achieved stable profitability. OpenAI had revenues of around $13 billion in 2025, but total expenses were as high as $34 billion, with a net loss of around $39 billion. Operating losses excluding one-time projects amounted to about $21 billion. While Anthropic is about to achieve its first quarter of operating profit, it continues to make huge investments in developing and deploying AI models. OpenAI's expenses in the first quarter of 2026 still exceeded revenue. Polymarket data shows that traders believe there is a 76% probability of an Anthropic IPO by the end of 2026. However, there have been reports that OpenAI's IPO plans may be postponed until 2027. Speculative frenzy in the AI boom The AI boom has not only driven a sharp rise in stocks of high-profit chip companies, but has also brought back speculative trading. Its features include the prevalence of leveraged ETFs, the dramatic volatility of small-cap stocks, and the resurgence of "meme stock" trading. Regulatory authorities have issued warnings. The Bank for International Settlements (BIS) warned in its latest annual report that the five largest hyperscale cloud services providers are expected to invest over $1 trillion in AI-related capital expenditure between 2025 and 2026, raising concerns of overheating in the sector. BIS General Manager Pablo Hernndez de Cos warned, "If the development of artificial intelligence fails to meet expectations, the entire industry could become more fragile, and the current investment frenzy could come to a halt." Morgan Stanley technical analyst Jason Hunter pointed out in a client report that the internal differentiation of AI trades is beginning to resemble the situation before the 1999 internet bubble - semiconductor, storage, and AI hardware suppliers' stocks continue to rise, while the cloud giants that bear large capital expenditures are lagging behind. TS Lombard predicts that global AI capital expenditure will reach around $800 billion by 2026, with economists warning that "Nvidia may be the new Cisco." Outlook: When the frenzy subsides, fundamentals will eventually dominate The market has repeatedly shown that high expectations can become a burden. Investors who buy into the market when sentiment is high often find that even strong company performance cannot support high stock prices. Will McGough, Chief Investment Officer of Prime Capital Financial, summed it up, "The army of retail traders certainly contributes to trend formation, but investment clearly isn't free." The current AI boom shares unsettling similarities with the internet bubble of 2000: both are driven by a narrative of "this time is different", both have companies with sky-high valuations that have yet to turn a profit, and both have retail investors rushing in the frenzy. The lesson from Cisco reminds investors - even if a company eventually becomes an industry leader, an overly high purchase price may take decades to recover. As Bitel said, the essence of long-term investing has not changed: understand the fundamentals of the company, find out where it stands in the ecosystem, and whether it can deliver on its promises. When the tide of AI speculation recedes, the naked will be exposed - and history has shown repeatedly that this time will be no different.