The only way is鈥own?
05 August 2025
With AI hype lifting tech valuations to new heights, Karl Loomes explores whether short sellers are lining up to burst the bubble
Image: stock.adobe.com/Balerinastock
So there I was, talking in an AOL chatroom, the chirp of the dial-up modem ringing in my ears, grateful that Y2K hadn鈥檛 destroyed the world and about to Ask Jeeves what exactly a 鈥榞oogle鈥� is. This, then, is the backdrop to the last true technology bubble at the turn of the millennium. Then it was the internet 鈥� a 鈥榥ew鈥� technology that everyone had heard of, most only had a vague comprehension of, and everybody wanted to use. Sound familiar?
Over the past two years or so artificial intelligence has become the buzzword on most peoples鈥� lips. Businesses across industries and sectors are rushing to use it (or at least seen to be using it) and while most people agree that it will change the world in some way, nobody can agree if this is a good or bad thing.
Despite some similarities with the tech bubble of 2000 however 鈥� most notably the hype in the press 鈥� market consensus would suggest that while AI-related shares have definitely seen growing demand on this wave, valuations are nowhere near the point where the 2000 bubble burst.
Arguably some of the shine is wearing off. This, I believe, is not because the technology is lacking, but rather the opposite 鈥� it is becoming normal to use and see AI in our day-to-day lives. The hype is fading because it is actually useful enough for us to integrate into our lives.
The level that share prices are, or have been, driven by hype compared to an actual fundamental shift in the underlying potential of a company is hard to discern. Add to this the US-China trade war and 鈥楾rump鈥檚 tariffs鈥�, and the water becomes even murkier. Tech-stocks are still up though, and short sellers, it seems, are betting they will be heading down.
Where are short sellers looking?
Data from across the sector shows a growing appetite for short exposure in tech, particularly in the 鈥楽oftware & Services鈥� sector. According to Sam Pierson, director of research at S3 Partners, 鈥渢he group includes some of the most highly shorted US equities by percentage of float, which includes some current and former high flyers with AI and crypto exposure, including Applied Blockchain, Coreweave, and Core Scientific鈥�.
Pierson notes that the notional short value across this group is currently at a year-to-date (YTD) peak of US$151 billion. Notably, this rise reflects both an increase in the number of shares shorted and in the total value of those positions.
Compared with other segments of the tech sector, the activity is telling. For 鈥楾ech Hardware & Equipment鈥�, Pierson says the 鈥減ercentage of float [is increasing], but the notional value is flat鈥�, suggesting positions are being scaled up in share terms but not necessarily in dollars. In contrast, the 鈥楽emiconductor鈥� industry group is seeing flat percentage of float but volatile notional values 鈥� behaviour consistent with fluctuating prices rather than strategic positioning.
These trends are mirrored in the securities lending revenue data. According to Serena Nayak, product specialist at S&P Global Market Intelligence: 鈥淒uring the second half of the year, six out of the top-10 revenue generating stocks were technology related, notably, Coreweave, which reported US$311 million in revenues during the second quarter.鈥�
She says: 鈥淪pecials activity actually rose during Q2 due to heightened market volatility, an uptick in IPO activity, and a growing number of tech stocks whose valuations soared post-IPO, particularly those involved in AI development. These stocks became prime targets for short sellers while also presenting potential short squeeze risks as their prices climbed.
鈥淭hese stocks became prime targets for short sellers while also presenting potential short squeeze risks as their prices climbed.鈥�
On the hardware front, she adds: 鈥淭he quantum computing sector attracted significant interest from securities lending market participants in Q2, with Quantum Computing and Arqit Quantum Inc standing out, collectively generating over US$30 million in revenues.鈥�
Nor is this interest limited to one region. Nayak notes: 鈥淎 similar trend was observed across Asia, where Wistron Corp., a Taiwanese tech firm, Horizon Robotics, a Hong Kong-listed software company, and Posco DX Co., a South Korean software firm, emerged as popular borrowing options.
鈥淎nother noteworthy mention is PONY AI ADR, which commanded an average volume-weighted fee exceeding 3,843 basis points during the quarter, generating US$30.4 million in revenues.鈥�
Geopolitics
Not all of this short interest, securities lending revenue, or arguable expectations of tech stock share prices falling, are simply on the back of an overbought market driven by hype. The same period of growing interest and coverage of AI has also coincided with the second term of Donald Trump.
President Trump has been very vocal in his economic intentions from the start of his term, as well as enacting numerous policies that impact the sector. The most notable of these, of course, has been his tariff policy, where semiconductors in particular 鈥� the computer chips that enable AI to function and train 鈥� have been at the forefront.
Nayak explains: 鈥淚n the context of the technology and software sectors, [the data shows] a notable rise in short interest 鈥� expressed as a percentage of the sector's market capitalisation on loan 鈥� following the introduction of the new US tariff policy. This policy shift has ignited geopolitical and geoeconomic tensions, as evidenced by the increases in short interest.
鈥淚nvestors anticipated a surge in protectionist measures from the US, which could adversely affect major technology firms through higher import and export tariffs, supply chain disruptions, or escalated input costs.鈥�
In this context, short sellers may be not just reacting to valuations but hedging against a broader macroeconomic risk picture 鈥� one with echoes of 2000鈥檚 dot-com bust, which was partly catalysed by global volatility and rapidly shifting investor sentiment.
The next bubble?
At present, the growing short positions in AI-related stocks do not appear to signal a full-scale bubble burst. Unlike 2000, where companies with no revenue or real business models were trading at irrational multiples, today鈥檚 AI names often have revenue, products and real-world adoption.
But the level of short interest and the size of these bets suggests not everyone is convinced that these firms are worth their lofty valuations 鈥� especially as AI shifts from a narrative driver to a functional tool. As Sam puts it, 鈥渟hort percentage of float is also at a YTD peak, where the other two tech industry groups are not鈥�.
In the meantime, the bears are watching. And short sellers 鈥� often the first to sense weakness 鈥� are lining up their positions. Whether this is the beginning of a correction or just a healthy market rebalance remains to be seen.
Over the past two years or so artificial intelligence has become the buzzword on most peoples鈥� lips. Businesses across industries and sectors are rushing to use it (or at least seen to be using it) and while most people agree that it will change the world in some way, nobody can agree if this is a good or bad thing.
Despite some similarities with the tech bubble of 2000 however 鈥� most notably the hype in the press 鈥� market consensus would suggest that while AI-related shares have definitely seen growing demand on this wave, valuations are nowhere near the point where the 2000 bubble burst.
Arguably some of the shine is wearing off. This, I believe, is not because the technology is lacking, but rather the opposite 鈥� it is becoming normal to use and see AI in our day-to-day lives. The hype is fading because it is actually useful enough for us to integrate into our lives.
The level that share prices are, or have been, driven by hype compared to an actual fundamental shift in the underlying potential of a company is hard to discern. Add to this the US-China trade war and 鈥楾rump鈥檚 tariffs鈥�, and the water becomes even murkier. Tech-stocks are still up though, and short sellers, it seems, are betting they will be heading down.
Where are short sellers looking?
Data from across the sector shows a growing appetite for short exposure in tech, particularly in the 鈥楽oftware & Services鈥� sector. According to Sam Pierson, director of research at S3 Partners, 鈥渢he group includes some of the most highly shorted US equities by percentage of float, which includes some current and former high flyers with AI and crypto exposure, including Applied Blockchain, Coreweave, and Core Scientific鈥�.
Pierson notes that the notional short value across this group is currently at a year-to-date (YTD) peak of US$151 billion. Notably, this rise reflects both an increase in the number of shares shorted and in the total value of those positions.
Compared with other segments of the tech sector, the activity is telling. For 鈥楾ech Hardware & Equipment鈥�, Pierson says the 鈥減ercentage of float [is increasing], but the notional value is flat鈥�, suggesting positions are being scaled up in share terms but not necessarily in dollars. In contrast, the 鈥楽emiconductor鈥� industry group is seeing flat percentage of float but volatile notional values 鈥� behaviour consistent with fluctuating prices rather than strategic positioning.
These trends are mirrored in the securities lending revenue data. According to Serena Nayak, product specialist at S&P Global Market Intelligence: 鈥淒uring the second half of the year, six out of the top-10 revenue generating stocks were technology related, notably, Coreweave, which reported US$311 million in revenues during the second quarter.鈥�
She says: 鈥淪pecials activity actually rose during Q2 due to heightened market volatility, an uptick in IPO activity, and a growing number of tech stocks whose valuations soared post-IPO, particularly those involved in AI development. These stocks became prime targets for short sellers while also presenting potential short squeeze risks as their prices climbed.
鈥淭hese stocks became prime targets for short sellers while also presenting potential short squeeze risks as their prices climbed.鈥�
On the hardware front, she adds: 鈥淭he quantum computing sector attracted significant interest from securities lending market participants in Q2, with Quantum Computing and Arqit Quantum Inc standing out, collectively generating over US$30 million in revenues.鈥�
Nor is this interest limited to one region. Nayak notes: 鈥淎 similar trend was observed across Asia, where Wistron Corp., a Taiwanese tech firm, Horizon Robotics, a Hong Kong-listed software company, and Posco DX Co., a South Korean software firm, emerged as popular borrowing options.
鈥淎nother noteworthy mention is PONY AI ADR, which commanded an average volume-weighted fee exceeding 3,843 basis points during the quarter, generating US$30.4 million in revenues.鈥�
Geopolitics
Not all of this short interest, securities lending revenue, or arguable expectations of tech stock share prices falling, are simply on the back of an overbought market driven by hype. The same period of growing interest and coverage of AI has also coincided with the second term of Donald Trump.
President Trump has been very vocal in his economic intentions from the start of his term, as well as enacting numerous policies that impact the sector. The most notable of these, of course, has been his tariff policy, where semiconductors in particular 鈥� the computer chips that enable AI to function and train 鈥� have been at the forefront.
Nayak explains: 鈥淚n the context of the technology and software sectors, [the data shows] a notable rise in short interest 鈥� expressed as a percentage of the sector's market capitalisation on loan 鈥� following the introduction of the new US tariff policy. This policy shift has ignited geopolitical and geoeconomic tensions, as evidenced by the increases in short interest.
鈥淚nvestors anticipated a surge in protectionist measures from the US, which could adversely affect major technology firms through higher import and export tariffs, supply chain disruptions, or escalated input costs.鈥�
In this context, short sellers may be not just reacting to valuations but hedging against a broader macroeconomic risk picture 鈥� one with echoes of 2000鈥檚 dot-com bust, which was partly catalysed by global volatility and rapidly shifting investor sentiment.
The next bubble?
At present, the growing short positions in AI-related stocks do not appear to signal a full-scale bubble burst. Unlike 2000, where companies with no revenue or real business models were trading at irrational multiples, today鈥檚 AI names often have revenue, products and real-world adoption.
But the level of short interest and the size of these bets suggests not everyone is convinced that these firms are worth their lofty valuations 鈥� especially as AI shifts from a narrative driver to a functional tool. As Sam puts it, 鈥渟hort percentage of float is also at a YTD peak, where the other two tech industry groups are not鈥�.
In the meantime, the bears are watching. And short sellers 鈥� often the first to sense weakness 鈥� are lining up their positions. Whether this is the beginning of a correction or just a healthy market rebalance remains to be seen.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to 网爆吃瓜 Finance Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to 网爆吃瓜 Finance Times
