"Interest Rates Alone Are Not Enough"... AI Investment Seeks Breakthrough via GPU and Data Center Collateralized Finance Platforms

An analysis has emerged suggesting that the growth trend of the artificial intelligence (AI) industry and related investments will not shrink even in an environment where high interest rates are maintained. While there is a burden of financing costs due to rising interest rates, the view is that the low debt ratios of Big Tech companies and the high profitability of data centers can offset this. In particular, new financial models using GPU (Graphics Processing Unit) or data centers themselves as collateral are emerging, providing a breakthrough for securing funds.

Emergence of 'AI Finance Platforms' using GPU and Data Center Collateral... Diversification of Financing Methods

Recently, new financial structures utilizing data centers and GPUs, which are core infrastructure for the AI industry, are being discussed. According to a video released on the YouTube channel 'Understanding: All the Knowledge in the World', while Big Tech companies such as Amazon are emphasizing profitability through the construction of data centers, NVIDIA has suggested the possibility of building financial platforms that use GPUs or data centers as collateral.

In the video, Lee Seon-yeop, CEO of AFW Partners, analyzed, "Even if it is impossible to completely avoid the influence of interest rates, there may be no problems in terms of financing for the time being." The explanation is that if data centers can generate sufficient profit, companies will gain the momentum to continue investing even while bearing somewhat high interest rates. In fact, a trend is appearing where NVIDIA proposes collateralized finance platforms in a situation where Amazon claims it can break even in three years through data centers, or Neocloud companies mention the possibility of rapid return on investment.

To support this trend, six global financial giants, including BlackRock and Goldman Sachs, are showing interest and participating in the development of financial products based on AI infrastructure. Although it is still at the MOU stage, if these financial products become full-scale, they could contribute to reducing the volatility of financing.

In the case of Big Tech companies, unlike general companies, they are characterized by having relatively low debt ratios, so the direct impact of interest rate hikes is relatively small. Lee emphasized, "Even if Google raises its financing at a rate 2 percentage points higher than current levels, it will not be a big problem if profitability supports it," adding, "The key is the profitability and investment efficiency of the data center rather than the interest rate level itself." However, he added that if growth rates are lower than interest rates, general companies could face problems as they may not be able to withstand the interest burden when interest rates rise.

Must Watch 'Credit Spreads' and 'Government Fiscal Spending' Besides Interest Rates

A warning was also raised that when judging the overall flow of the market, one should not be preoccupied with a single benchmark interest rate indicator alone, but should comprehensively examine credit spreads and liquidity supply channels. The analysis is that while high interest rates also signify market instability, a crisis may actually arrive when the market ignores the rise in interest rates and continues its upward trend.

Lee recommended checking credit spreads to identify precursors of a crisis. In particular, he mentioned the need to pay attention to the movements in the PEF market. The video explained, "While the scale of the PEF market is not enough to shake the entire debt market, if problems occur in a specific sector, it can act as a canary that sends a warning to the market." This means that one must be wary of problems in a specific sector spreading into a credit crisis for the entire market, as in the past case of Countrywide Financial in 2008.

Furthermore, he cited 'government fiscal spending' as a liquidity supply channel that is difficult to capture with traditional interest rate indicators alone. He pointed out the situation where governments of major countries, including the United States and South Korea, are supplying liquidity by spending massive amounts of finances through bond issuances and other means. Lee said, "In the past, market liquidity was mainly determined by interest rates, but recently, the influence of each country's government releasing money through policy has increased," and said that the change in liquidity energy must be watched. In particular, he analyzed that market energy is determined not only by interest rates but also by the fiscal policies of governments, mentioning the tendency of governments to spend finances to maintain political power.

Whether AI Demand Continues and the Task for Korean Semiconductor Companies

It is difficult to predict when the investment cycle of the AI industry will end. This is because the speed of technological development is so fast that existing demand forecasting models may be rendered powerless. There is a constant possibility that every time a new AI model emerges, it will create a much larger amount of GPU demand than before.

Meanwhile, regarding the status of Korean semiconductor companies in the global market, the need to improve governance and shareholder return policies was mentioned. The suggestion is that for large companies like Samsung Electronics and SK hynix to be continuously chosen by global investors, they must build a shareholder-friendly management environment that meets global standards, in addition to technical superiority.