A scenario eerily reminiscent of the 2008 financial crisis appears to be unfolding around the AI investment frenzy.

During the 2008 crisis, Michael Burry, the real-life inspiration behind *The Big Short*, gained fame by shorting the housing bubble, while Deutsche Bank traders quietly marketed credit default swaps (CDS) to investors seeking to hedge against real estate market risks.

Now, history seems to be repeating itself. Burry has publicly turned bearish, with 80% of his portfolio aggressively shorting NVIDIA and Palantir. Shortly after, Deutsche Bank began exploring short positions in AI stocks to hedge against massive loan exposures in data centers, even considering similar hedging instruments—only this time, the underlying assets have shifted from mortgage-backed securities to AI data center loans.

These striking parallels have prompted global regulators to sound alarms over AI asset bubbles. The Monetary Authority of Singapore explicitly warned that tech and AI sectors show "relatively stretched valuations," cautioning that a reversal in market optimism could trigger a "sharp correction." South Korea’s exchange also issued a rare "investment caution" notice for chipmaker SK Hynix. Meanwhile, Goldman Sachs and Morgan Stanley CEOs have echoed concerns, warning that U.S. stock valuations are too high and could face at least a 10% pullback.

**"The Big Short" Returns: 80% Bet on AI Collapse** Michael Burry, who rose to fame by predicting and shorting the U.S. subprime mortgage crisis, is now translating his warnings about an AI bubble into aggressive bearish bets. Recent regulatory filings reveal that his Scion Asset Management has allocated roughly 80% of its portfolio to shorting Palantir and NVIDIA, with notional values exceeding $1 billion.

The filings show Burry holds $912 million in notional put options against Palantir (equivalent to 5 million shares) and $186 million in puts against NVIDIA. However, key details like premium costs, strike prices, and expiration dates remain undisclosed.

Before disclosing these positions, Burry posted cryptic social media messages, quoting classic films: "Sometimes we see bubbles. Sometimes we can act. Sometimes the only winning move is not to play." He hinted that AI investments offer meager returns, likening the current hype to the excessive capital expenditures in fiber optics during the dot-com bubble—and predicting many leading AI firms will eventually collapse.

His strategy mirrors his pre-2008 playbook, where he famously used CDS to bet against the housing market amid widespread optimism. Now, he’s targeting the AI sector, the current epicenter of market euphoria.

However, it’s worth noting that as both stocks have rallied since the filing date, Burry’s short positions have likely incurred significant paper losses. Markets are watching closely to see if the legendary investor can replicate his past success.

**Deutsche Bank’s Two-Pronged Strategy: Lending While Hedging** Driven by AI demand, data center financing has become a core focus for Deutsche Bank’s investment arm. A senior executive revealed the bank has made "big bets" in this space, primarily lending to firms servicing hyperscalers like Alphabet, Microsoft, and Amazon. These loans are often backed by long-term service contracts, ensuring stable returns.

In recent months, Deutsche has provided debt financing to Sweden’s EcoDataCenter and Canada’s 5C, helping raise over $1 billion for expansion. While the bank hasn’t disclosed total sector exposure, estimates suggest tens of billions.

Yet, amid this lending spree, internal discussions about risk hedging have emerged. According to reports, Deutsche is evaluating two approaches:

1. Directly shorting a basket of AI-related stocks to profit from a market downturn. 2. Using synthetic risk transfer (SRT) derivatives to offload portions of loan default risk to external investors.

The move has drawn attention not just for its prudence but for its historical echoes. During the 2008 crisis, Deutsche trader Greg Lippmann spearheaded the CDS market that enabled investors to short housing.

Some observers note that the SRT structure—pooling and tranching risk assets—bears an uncanny resemblance to the collateralized debt obligations (CDOs) of 2008, raising concerns about hidden risks.

But hedging AI exposure is no easy feat. Shorting AI stocks in a bull market is costly, while SRT deals face hurdles like achieving diversified loan pools for ratings and investor demands for higher risk premiums.

Notably, Deutsche analysts argued in September that AI bubble fears were overblown, declaring "the bubble about the bubble has burst." This internal contradiction highlights the complex balancing act major financial institutions face in today’s market.