Author: Li XiaoyinSource: Hard AIThe end goal of AI is electricity.On May 10th, the Morgan Stanley research team released a research report on the topic of generative AI that global investors are concerned about. The report shows that currently, investors are mainly concerned about issues such as carbon emissions, power generation methods, and electricity prices in AI data centers.Morgan Stanley believes that with the development of AI technology, the construction of data centers will push up electricity prices, benefiting related clean energy stocks. If the issue of power shortage continues to escalate, it may lead to a pullback in high-priced chip stocks including NVIDIA.<h2>What areas are worth paying attention to in &#34;AI power supply&#34;?</h2>In the report, Morgan Stanley suggests focusing on areas that &#34;help data centers shorten power supply time&#34;.Specifically, these include: renewable energy storage developers with large-scale interconnection capabilities and diverse supply chains, large-scale cryptocurrency mining farms that can be converted into data centers, and some currently undervalued power developers.<h2>How will the carbon emissions of future data centers be?</h2>The report believes that the carbon emissions of data centers will be lower than expected.The reasons are as follows: 1) By 2030, supercomputers are expected to achieve decarbonization goals; 2) In the future, supercomputers will initiate new large-scale renewable energy and energy storage projects.The report also states that the potential for renewable energy supply may exceed market expectations:<blockquote>Our model shows that by 2027, if 70% of the electricity growth in US data centers is powered by renewable energy (and by using energy storage to supply power to the grid at times that better match data center power curves), then by that time, the total power of renewable energy/storage technologies for data centers will reach approximately 18 gigawatts, equivalent to about 33% of our baseline forecast for US data center power demand.</blockquote>According to the estimated data in the report, this means that by 2027, renewable energy will be able to meet one-third of the additional power demand of AI data centers.<h2>How will it affect electricity prices?</h2>The report suggests that electricity prices may rise, and the gross profit margin of power companies will significantly improve.Currently, in several US electricity markets with huge potential for data center development, Morgan Stanley has found that the forward electricity curve has shifted upwards:<blockquote>In two typical US electricity markets: the Pennsylvania-Maryland-Jersey (PJM) Interconnection and the Electric Reliability Council of Texas (ERCOT), based on peak electricity prices in early May this year, it is expected that most electricity prices for the same period in the next 1-2 years will mostly rise.<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/76cdbf16-0f00-42cc-96eb-1ea605d803cf.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/76cdbf16-0f00-42cc-96eb-1ea605d803cf.png"/></blockquote>In addition, considering the time for licensing, equipment procurement, and factory construction, it may take several years for gas-fired power plants to be fully operational. Therefore, the report predicts that the tight supply situation in the electricity market will not be eased in the short termThis is a great boon for power companies. Data released by the American NRG Energy company shows that electricity prices will rise by 25% by 2027. Based on forward prices, the company expects operating gross profit to increase by $420 million, a 27% increase from the current level.<h2>Which stocks will be most negatively affected if the &#34;power shortage&#34; is not alleviated?</h2>According to investor feedback, the report indicates that if the &#34;power shortage&#34; persists, investors are most concerned about chip stocks that are currently experiencing rapid growth.The report states that in a state of power shortage, the business of data centers will slow down, which means that the growth in sales of AI chips will also slow down. Large AI chip manufacturers like NVIDIA, whose stock prices are currently high, already incorporate investors' expectations of their future rapid growth. Therefore, the risk of a significant stock pullback will increase at that time.<h2>Is there a &#34;tail risk&#34; that could lead to a slowdown in the development of data centers?</h2>Capital expenditure could pose a significant risk.Wall Street News previously mentioned that in recent quarterly earnings conference calls, Microsoft reported a year-on-year increase in capital expenditure of over 50%, and stated plans to continue increasing spending on artificial intelligence by 2025. Meta and Alphabet also hinted at a 40% to 50% increase in capital expenditure, with expectations of a significant increase in capital expenditure this year.The report indicates that some investors believe this could bring potential risks to the companies

NVDA

NVDL

SOXL

Morgan Stanley stated that the tight supply in the electricity market will not be eased in the short term. If the "power shortage" continues and data center business slows down, it may lead to a pullback in high-priced chip stocks such as NVIDIA

Wallstreetcn

Gross profit margin to increase by 27%? Several key questions about the hot "AI power supply stock"