On the morning of February 27, GanNeng Corporation, YuNeng Holdings, and Huayin Electric Power all hit the daily limit-up, with Fuling Power and Gansu Energy soaring significantly. The power sector overall performed strongly, with the Power ETF increasing by over 1.6%.
In terms of news, besides the recent announcement by the National Energy Administration of the first batch of pilot projects for enhancing new power system construction capabilities, a very solid logic has emerged. Reports indicate that China’s AI call volume has surpassed the US for the first time, with four large models dominating the top five globally. Behind token call volume is power and computing capacity. China’s low electricity prices allow for indirect overseas expansion through token call volume.
The Power Logic Behind Token Call Volume
On February 27, power stocks collectively surged, with GanNeng rising three consecutive days, Huayin Power hitting the daily limit, and YuNeng Holdings also hitting the limit in seconds. This is partly related to a change in style, and also possibly linked to the evolution of a certain logic.
According to data from OpenRouter, the world’s largest AI model API aggregation platform, from February 9 to 15, Chinese models recorded a call volume of 4.12 trillion tokens, surpassing the US models’ 2.94 trillion tokens for the first time during the same period. The following week, February 16 to 22, China’s weekly call volume further surged to 5.16 trillion tokens, a 127% increase over three weeks, while US model call volume fell to 2.7 trillion tokens.
So, what exactly is the underlying power logic? Data shows that electricity is the “fuel” and main cost for producing AI tokens. In the operation costs of large AI models, electricity costs can account for 60%-70%, far exceeding hardware depreciation and network costs.
China also has a significant cost advantage in electricity. Western China (such as Inner Mongolia and Gansu) offers very low-cost green power (wind and solar), around 0.2-0.3 yuan per kWh, compared to an average of about 1-1.5 yuan per kWh in Europe and America. This means producing the same amount of tokens in China could cost only one-fifth to one-quarter of what it costs in the West.
By utilizing China’s cheap electricity to produce tokens and selling them globally via APIs, a profitable model emerges: “Power stays within the country, high-value digital services go abroad.” This is the investment logic behind “Token Going Overseas.” Some analysts believe that tokens are derivatives of electricity, and that token exports essentially mean electricity exports.
HALO Trading Also Supports the Sector
It’s worth noting that the recently popular HALO trading strategy (Heavy Assets, Low Obsolescence) is also boosting the power sector. HALO suggests that AI technology is disrupting light-asset industries like software and IT services, while creating huge demand for physical assets such as power, energy, and hardware. Investors are therefore seeking companies with high barriers to entry and tangible assets that are less susceptible to technological obsolescence as “safe havens.”
In recent research reports, CITIC Securities classifies industries based on physical dependence and regulatory/emotional barriers into four categories: Damaged (low physical dependence, low barriers), Reshaped (low dependence, high barriers), Fortress (high dependence, high barriers), and Beneficiaries (high physical dependence, low barriers). In the future, the gap in returns between physical asset beneficiaries and those harmed by code expansion may continue to widen, and this divergence trend is expected to persist.
Zhongyuan Securities believes that power demand is shifting from traditional industries to residential use and emerging sectors like artificial intelligence and data centers. The power supply structure is also being optimized, with over 430 million kW of new wind and solar capacity added nationwide by 2025, a 22% year-on-year increase.
Dongwu Securities states that to support the high proportion of new energy consumption, China needs to further promote market-oriented electricity pricing reforms to underpin the increasingly complex new power system construction. In the future, the electricity market will gradually assign full pricing to various attributes of power commodities (energy value, regulation value, capacity value, environmental value, etc.).
Dongcai Visual Guide · Key Insights
(Source: Securities Times)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
The three major hot stocks "秒板"!Hardcore good news is coming!
Power Stocks Surge!
On the morning of February 27, GanNeng Corporation, YuNeng Holdings, and Huayin Electric Power all hit the daily limit-up, with Fuling Power and Gansu Energy soaring significantly. The power sector overall performed strongly, with the Power ETF increasing by over 1.6%.
In terms of news, besides the recent announcement by the National Energy Administration of the first batch of pilot projects for enhancing new power system construction capabilities, a very solid logic has emerged. Reports indicate that China’s AI call volume has surpassed the US for the first time, with four large models dominating the top five globally. Behind token call volume is power and computing capacity. China’s low electricity prices allow for indirect overseas expansion through token call volume.
The Power Logic Behind Token Call Volume
On February 27, power stocks collectively surged, with GanNeng rising three consecutive days, Huayin Power hitting the daily limit, and YuNeng Holdings also hitting the limit in seconds. This is partly related to a change in style, and also possibly linked to the evolution of a certain logic.
According to data from OpenRouter, the world’s largest AI model API aggregation platform, from February 9 to 15, Chinese models recorded a call volume of 4.12 trillion tokens, surpassing the US models’ 2.94 trillion tokens for the first time during the same period. The following week, February 16 to 22, China’s weekly call volume further surged to 5.16 trillion tokens, a 127% increase over three weeks, while US model call volume fell to 2.7 trillion tokens.
So, what exactly is the underlying power logic? Data shows that electricity is the “fuel” and main cost for producing AI tokens. In the operation costs of large AI models, electricity costs can account for 60%-70%, far exceeding hardware depreciation and network costs.
China also has a significant cost advantage in electricity. Western China (such as Inner Mongolia and Gansu) offers very low-cost green power (wind and solar), around 0.2-0.3 yuan per kWh, compared to an average of about 1-1.5 yuan per kWh in Europe and America. This means producing the same amount of tokens in China could cost only one-fifth to one-quarter of what it costs in the West.
By utilizing China’s cheap electricity to produce tokens and selling them globally via APIs, a profitable model emerges: “Power stays within the country, high-value digital services go abroad.” This is the investment logic behind “Token Going Overseas.” Some analysts believe that tokens are derivatives of electricity, and that token exports essentially mean electricity exports.
HALO Trading Also Supports the Sector
It’s worth noting that the recently popular HALO trading strategy (Heavy Assets, Low Obsolescence) is also boosting the power sector. HALO suggests that AI technology is disrupting light-asset industries like software and IT services, while creating huge demand for physical assets such as power, energy, and hardware. Investors are therefore seeking companies with high barriers to entry and tangible assets that are less susceptible to technological obsolescence as “safe havens.”
In recent research reports, CITIC Securities classifies industries based on physical dependence and regulatory/emotional barriers into four categories: Damaged (low physical dependence, low barriers), Reshaped (low dependence, high barriers), Fortress (high dependence, high barriers), and Beneficiaries (high physical dependence, low barriers). In the future, the gap in returns between physical asset beneficiaries and those harmed by code expansion may continue to widen, and this divergence trend is expected to persist.
Zhongyuan Securities believes that power demand is shifting from traditional industries to residential use and emerging sectors like artificial intelligence and data centers. The power supply structure is also being optimized, with over 430 million kW of new wind and solar capacity added nationwide by 2025, a 22% year-on-year increase.
Dongwu Securities states that to support the high proportion of new energy consumption, China needs to further promote market-oriented electricity pricing reforms to underpin the increasingly complex new power system construction. In the future, the electricity market will gradually assign full pricing to various attributes of power commodities (energy value, regulation value, capacity value, environmental value, etc.).
Dongcai Visual Guide · Key Insights
(Source: Securities Times)