A new departure in the artificial intelligence industry; Is the crisis spreading from the stock market to the banks? At the same time as the heavy decline in the shares of technology companies in the US and the decrease of about 2.3 trillion dollars in the market value of the companies of the Nasdaq 100 indexes in trading on November 5, another news increased the concerns about the future of the artificial intelligence industry.
The Dutch bank ING decided to sell a series of loans related to data center projects worth 500 million dollars. Part of this facility is related to the large Stargate project, which is being developed with the support of OpenAI, Oracle and SoftBank. The importance of this news has more to do with the financing conditions of the artificial intelligence industry than with the amount of loans. Companies such as Microsoft, Google, Meta and Oracle have made extensive investments in the development of AI infrastructure in recent years. However, a significant portion of data center projects depend on borrowed financial resources, and their return on investment also depends on the growth in demand for artificial intelligence services in the coming years.
In such a situation, increasing political pressure to limit the development of data centers in America could change the economic equations of these projects. Any delay in construction or operation means a delay in revenue, while financing costs and debt repayments will continue. On the other hand, if banks are forced to offer significant discounts for the sale of existing loans, financiers can be expected to be more cautious in granting new concessions and demand higher interest rates. As a result, the cost of capital increases and the economic justification of some projects is questioned.
Of course, the sale of ING loans alone is not a sign of a credit crisis and may have been done solely for the purpose of managing the bank's balance sheet and its liquidity. What matters is the final sale price of these loans and the willingness of investors to accept the risk of data center projects. The bottom line is that risk in the AI industry is no longer limited to high valuations of tech stocks.