Group 1 - The report emphasizes the importance of "crowding" as a key indicator reflecting market sentiment in popular sectors, constructed from four dimensions: volume, price, capital, and analyst forecasts [1][11][12] - The current trading crowding in the TMT sector has rebounded from the bottom to a high level, with many segments of the AI industry chain also showing high crowding, although some remain at moderate levels [2][12] - The report suggests that when crowding is low, it indicates a bottoming phase for stock prices, while high crowding suggests potential for significant price corrections [1][11] Group 2 - The transaction ratio has reached a historical high of 46%, raising concerns about whether the AI trading sentiment has peaked [3][17][20] - The report indicates that while a high transaction ratio may lead to increased volatility, it does not typically signal a systemic end to the market trend, as internal rotation and high-low switching can help digest short-term overheating [3][20] - Historical examples are provided, showing that significant changes in industry trends or fundamentals can lead to new trend formations despite high transaction ratios [3][20] Group 3 - The report introduces a "rotation intensity" indicator to measure the speed of internal rotation within the AI sector, noting that a convergence in rotation intensity often leads to a mainline market trend [4][28] - Following the Spring Festival, the main directions within AI have become clearer, with the computer and media sectors leading the gains, resulting in a decrease in rotation intensity [4][28][29] - The relationship between the AI index and rotation intensity suggests a pattern of "linked rises and rotating adjustments," indicating resilience in the sector rather than systemic declines [4][34] Group 4 - U.S. Treasury yields are highlighted as a significant factor affecting the pricing of high-valuation growth assets, with rising yields typically suppressing market risk appetite [5][37] - The report notes a strong correlation between TMT performance and U.S. Treasury yields, suggesting that changes in yields should be closely monitored from a trading perspective [5][37][39] Group 5 - The report discusses the importance of earnings performance in the AI sector, noting that the correlation between stock price movements and earnings growth is strongest during earnings disclosure periods [7][43] - It is observed that when the market focuses on fundamentals, the TMT sector may face adjustment pressures, while periods of trading on expectations can lead to better performance [7][43][45] - The example of optical modules is provided, illustrating how sustained earnings performance can lead to a strong positive correlation with stock price movements [7][51][52]