• Lender seeks to ramp up CLA mandates in data center deals
• Opportunities for enterprise data center deals in the USD 1bn-to-USD 5bn range

Norddeutsche Landesbank, or NordLB, is ramping up its ambitions in the data center project finance market as it looks to replicate its leadership position in renewable energy, said Maniesh Khatri, senior director at the German-domiciled lender.

Khatri, who rejoined NordLB in mid-2024 after a stint at DNB, said the bank is seeking to take on larger roles as coordinating lead arranger (CLA) in multi-billion-dollar financings for data center developers.

“We’re a structuring bank,” he said. “That’s what got us to the top in renewables, and that’s what we’re trying to do here.”

The push comes as deal sizes have rapidly expanded. According to Khatri, what began as USD 1bn transactions in 2022 have evolved into USD 20bn to USD 30bn projects today. “Your ticket sizes just need to go up if you want to be a CLA,” he said.

Khatri emphasized that as ticket sizes get larger, the lending markets explore potential partnerships to manage risk, such as credit risk insurance and credit default swaps. He also said that banks could seek to pick among enterprise data center deals in the USD 1bn – USD 5bn range.

“Putting multi-billion-dollar commitments doesn’t make that much sense from a bank’s perspective unless you have the confidence to syndicate the exposure,” he said.

Part of the calculus involves gauging how much exposure the bank can safely retain after underwriting and syndicating the debt. “The syndication team needs to have a pulse on the market before you even underwrite,” Khatri said. “They’re already talking to banks, asking: ‘If this comes out, would you like this financing? What size ticket could you take?’”

That pre-marketing work helps determine how much risk a lead bank should take up front. “If they’re not confident they can sell down, then you’re not going to underwrite as much,” Khatri noted.

In the meantime, the fee structure creates strong incentives for banks to lead and then downsize. “You might underwrite USD 2bn and end up holding USD 200m,” he said. “But along the way, you’re making fees from every tranche you syndicate down—12.5 basis points here, 25 there,” with the spread between what the sponsor pays and what downstream banks receive providing a margin.

At the largest institutions, internal asset-backed securities (ABS) teams play a key role in shaping deal strategy. The large banks often use ABS takeouts as the endgame for construction financing, meaning that even large underwritten positions can eventually be removed from the balance sheet altogether. “They can write their own destiny,” Khatri said. “If you know the ABS desk can take it out in two to three years, your effective exposure goes to zero—and you’re making money the whole way.”

Still, risk remains—especially around counterparties and evolving AI business models. For Khatri, the core underwriting question is who ultimately pays the lease payments. “Would you lose sleep over Amazon or Microsoft? Probably not, as you can count on their existing streams of revenues,” he said. “But with tenants where the majority of their revenues is dependent on the AI delivering on its promise, you have to scrutinize.”

He also noted the higher-level risk of not knowing how AI business models will ultimately work. “You’re taking the risk that the AI market doesn’t evolve the way people think it will,” he added. “It’s a big question right now—how durable are these business models really?”