Deutsche Bank Hedges Against Data Centres

Deutsche Bank seeks hedges for data centre exposure as AI lending accelerates

Deutsche Bank is examining ways to hedge its growing exposure to data centres after lending billions of dollars to the sector in order to support the rapid expansion of artificial intelligence and cloud computing infrastructure.

Executives at the bank have discussed a range of risk-management options as the world’s largest technology groups commit vast sums to building the facilities needed to power their AI ambitions. Much of this expansion is being financed through debt, increasing pressure on lenders to protect themselves in the event of a slowdown.

People familiar with the discussions said Deutsche is considering shorting a basket of AI-related stocks to offset potential losses should valuations fall. The bank is also exploring the use of derivatives to buy default protection on parts of its loan book through a structure known as a synthetic risk transfer. Such deals allow banks to shift credit risk to outside investors in return for a premium.

Deutsche declined to comment. One senior executive said the investment bank had “bet big” on data centre financing, reflecting growing demand from the hyperscale cloud providers that drive the industry. The bank has lent predominantly to companies that build and service facilities used by Alphabet, Microsoft and Amazon. The loans are typically secured against long-term contracts that offer stable cash flows.

The extraordinary pace of spending on AI infrastructure has prompted warnings from some analysts that a bubble may be forming. Critics argue that lenders are deploying billions of dollars into an industry that remains untested at this scale and relies on assets that depreciate quickly as technology evolves. The comparison with the build-up to the dotcom crash has been raised in several investment notes.

Deutsche has been among the most active lenders in Europe. In recent months it provided financing to Sweden’s EcoDataCenter and to 5C in Canada, helping the companies raise more than one billion dollars combined for expansion. While the bank does not disclose its total exposure, industry estimates put the figure in the billions.

Finding effective hedges may prove challenging. Betting against AI stocks is expensive in a rising market, and a synthetic risk transfer requires a diversified pool of loans to obtain a credit rating. Investors are also likely to demand higher premiums to insure against the risk of default in such a fast-moving sector.

The surge in demand for computing power has pushed cost estimates for AI infrastructure as high as three trillion dollars between now and 2030. That has driven intense interest in data centres and the companies that support them, leading to expectations of stronger dealmaking and consolidation across Europe. Competition for suitable sites has increased, with developers racing to acquire land and secure power supply.

Deutsche is already adjusting its broader strategy around the sector. In September, the Financial Times reported that its asset management arm, DWS, was preparing to sell its own data centre business at a valuation of around two billion euros as part of a wider portfolio review.

Despite the concerns, Deutsche’s research team has pushed back against the idea that the industry is in a speculative bubble. In a note published in late September, analysts said their model, which tracked references to an AI bubble across English-language publications, indicated that fears had been exaggerated. As they put it, one bubble had already burst, which was the bubble in saying that there was a bubble.

The debate is unlikely to fade soon. As AI adoption accelerates, lenders must weigh the promise of long-term returns against the risk that technology cycles shorten and capital requirements rise. For Deutsche, the focus has shifted to ensuring that its early position in the market does not leave it overly exposed if conditions tighten.