Ares CEO Says It Avoids AI Chip Financing Due to Business Complexity
Ares Management CEO Michael Arougheti said the company is avoiding financing the high-performance processors that run advanced AI models, and is instead investing in the data centers that house those chips. He said due diligence on data center projects is simpler, relative returns are attractive, and it is impossible to conduct due diligence to assess the value decay and depreciation curves of GPU technology. The company estimates AI capital expenditure demand at about $5 trillion, of which about $2 trillion is for physical AI data centers and supporting infrastructure, and $3 trillion is for purchasing GPU chips.
Why read this
AI score 53
Ares avoids chip financing because GPU depreciation and value decay make due diligence difficult, and instead turns to data centers, illustrating the risk trade-offs in AI infrastructure financing.
Topics
Body · AI translation
Related topics top start Related topics top end Futures promotion begin Futures promotion end Miaopai begin Miaopai end Video player start Video player end Market chart begin Market chart end
Ares Management Corp. is avoiding financing the high-performance processors that run advanced AI models, and is instead investing in the data centers that house those chips.
A person familiar with the matter said the asset manager is concerned that investing in GPUs could leave it stuck with assets facing obsolescence shortly after they come to market. The person requested anonymity because the information is not public.
CEO Michael Arougheti said on an analyst call on Thursday, referring to data center financing and GPU chip financing: "Frankly, due diligence on data center projects is simpler, and the relative returns are also attractive."
He added that this is not a rejection of the GPU financing business, but that investment opportunities in data centers and the supporting power infrastructure are more attractive to them.
Arougheti said the data center financing market is enormous, while the capital available to deploy is relatively limited. The company estimates that AI-related capital expenditure demand is about $5 trillion, of which about $2 trillion is for physical AI facilities and supporting infrastructure, and $3 trillion is for purchasing GPU chips.
Almost all major U.S. banks and alternative asset managers have participated in this round of AI financing boom, with related deals including Broadcom's roughly $60 billion chip debt financing (the chips will be leased to Anthropic), and SpaceX's planned fundraising of about $40 billion to purchase Nvidia chips. Chips cost far more than the data centers that house them, but are typically depreciated over only 5 to 6 years.
Arougheti said on the earnings call that he cannot conduct due diligence and evaluation of the value decay and depreciation curves of this type of technology.
He said: "At least at this stage, we choose to focus on the half of the market we know best and where we have the deepest accumulated capabilities."
Well-known compute cloud provider CoreWeave Inc. rents GPUs to AI developers, and its bonds fell sharply this month, with prices already entering distressed-debt territory; the cost of debt insurance has also soared, reflecting market concerns that the scale of debt supporting the AI boom continues to rise. news_keyword_pub,usstock, iframe below body iframe below body General statement General statement QR code at end start QR code at end start
QR code at end start QR code at end start Editor name and work code
Editor name and work code