Beware of the bubble! Deutsche Bank considers shorting AI stocks for risk hedging

As the investment frenzy triggered by artificial intelligence (AI) pushes the data center industry to peak valuations, some key financial players involved are beginning to examine potential risks.

On November 5th, the Financial Times reported, citing informed sources, that Deutsche Bank executives have been discussing internally how to manage the bank's risks in the data center industry. The bank has provided billions of dollars in loans to this sector to meet the demands of AI and cloud computing. Now, the hedging strategies being considered by the bank include shorting a basket of AI-related stocks to guard against market downturn risks.

Additionally, the bank is evaluating a complex derivative transaction known as "synthetic risk transfer" (SRT) to purchase default protection on some of its loans. The emergence of these internal discussions comes as global regulators and some investors issue increasingly clear warnings about the AI-driven asset bubble, suggesting that financial institutions' attitudes toward this tech feast may be undergoing a subtle shift.

The "Gamble" of AI Financing and Hedging Considerations

Driven by AI demand, financing for data centers has become a "gamble" for Deutsche Bank's investment banking business. According to a senior executive at the bank, the bank has "made a big bet" on data center financing. The bank primarily provides loans to companies serving tech giants like Alphabet, Microsoft, and Amazon, which are typically secured by long-term service contracts.

Reports indicate that Deutsche Bank has provided debt financing in recent months to Swedish group EcoDataCenter and Canadian company 5C, helping them raise over $1 billion for expansion. Although the bank has not disclosed the specific total amount of loans to the industry, it is estimated to be in the billions of dollars.

It is against this backdrop that discussions about hedging risks have surfaced. In addition to shorting AI stocks, the synthetic risk transfer (SRT) transactions being considered by the bank essentially involve packaging and selling the default risk of loans to external investors through derivatives. Both methods aim to provide a layer of protection for the bank's balance sheet against potential market corrections.

Rising Bubble Concerns, Warnings from Regulators and Markets

Deutsche Bank's cautious stance is not an isolated case; it reflects growing general concerns in the market about an AI bubble. Some skeptics point out that a large amount of capital is flowing into an inadequately tested industry, where assets face rapid depreciation risks due to swift technological iterations, reminiscent of the internet bubble at the beginning of this century.

The warnings from regulators are also becoming louder. Wallstreetcn reported that the Monetary Authority of Singapore (MAS) explicitly pointed out in its recent Financial Stability Assessment report that the technology and AI sectors are showing "relatively tight valuations," warning that a reversal of market optimism could trigger a "sharp correction." Coincidentally, the Korea Exchange also issued a rare "investment caution notice" regarding chipmaker SK Hynix, due to its stock price soaring significantly under the influence of AI concepts These official warnings, combined with the recent sell-off of semiconductor stocks triggered by the disappointing earnings outlooks of companies like Palantir and AMD, together form the macro backdrop for Deutsche Bank's risk assessment.

Actions of the "Big Short" and the Challenges of Hedging

As institutional investors begin to position defensively, some well-known investors have already expressed their bearish stance through concrete actions. According to Wallstreetcn, investor Michael Burry, famous for the movie "The Big Short," is one of them. Regulatory filings show that his managed Scion Asset Management fund has allocated about 80% of its positions to shorting representative companies of the AI boom, NVIDIA and Palantir, with a notional value exceeding $1 billion.

However, hedging against AI risks is not an easy task. As the report analyzes, shorting a basket of AI stocks in a continuously booming market can be very costly. Meanwhile, synthetic risk transfer transactions also face challenges, as they require a sufficiently diversified loan pool to obtain ratings, and in the current environment, investors may demand higher returns for taking on these risks.

Contradictory Signals Within Deutsche Bank

It is noteworthy that the views on AI risks within Deutsche Bank do not seem to be unanimous. Just this September, analysts from the bank released a report stating that concerns about an AI bubble have been overstated, asserting that "the bubble about the bubble has already burst."

This internal contradiction in viewpoints highlights the complex situation faced by large financial institutions in the current market environment: on one hand, there is a desire to seize the historic opportunities brought by AI, while on the other hand, there is a need to remain vigilant against potential significant risks.

This weighing of opportunities against risks may become one of the themes in the global financial markets for the foreseeable future