Report Industry Investment Rating No relevant content provided. Core Viewpoints of the Report - The essence of the US stock macro - strategy framework is the DDM model, with enterprise earnings, risk - free rate, and equity risk premium as the core elements [2][3]. - The position and direction of the profit cycle on the molecular end are the anchors for judging the US stock trend, and the continuous upward revision of EPS is the dominant factor driving the rise of US stocks [6][8]. - The US stock investor structure is mature, with institutional investors holding 60% of the positions, and passive investment is dominant [27]. - The current valuation of the overall US stock market is not cheap based on the next 12 - month consensus forecast PE [45]. - The performance of the US stock AI industry chain is realized from the upstream infrastructure layer to the downstream application software/media, and the upstream industry trend has strong certainty while the downstream is highly differentiated [104][106]. - The direction and degree of performance revision expectations are the core contradictions determining the three - stage market evolution of US stock AI [109][112]. Summary by Relevant Catalogs 1. US Stock Macro Strategy Framework: DDM Model 1.1 Essence of the Equity Investment Framework - The core of the market driving variables is the three elements of the DDM model: enterprise earnings (molecular end), risk - free rate, and equity risk premium (denominator end) [3][5]. 1.2 Molecular End - In the long run, the continuous upward revision of EPS on the molecular end is the dominant factor driving the rise of US stocks, and the position and direction of the profit cycle are the anchors for judging the US stock trend [6][8]. - Based on the Bloomberg style factor back - testing model, the factor representing profit growth ability has the highest winning rate in the US stock investment strategy framework [11]. - The ROE trend is strictly positively correlated with the US stock valuation in the long run [14]. - For cyclical assets, the core of the analysis framework is to grasp the position and direction of the macro - cycle, and the "soft data" of the US economy has a more significant guiding effect on US stock cyclical assets [15][18]. - For technology assets, the core of the analysis framework is to grasp the industrial prosperity/innovation cycle, and they may be desensitized from the macro - cycle under the guidance of the industrial trend [19]. - There are dynamic monitoring databases for the consensus forecast of EPS of US stock broad - based/key industries and the consensus forecast of Capex of the "Seven Sisters" in the US stock market, and the Capex of US technology giants is expected to continue to be revised upward in the next two years, with greater investment in the upstream AI segment [20][22]. 1.3 Denominator End - The US stock investor structure is mature, with institutional investors holding 60% of the positions, and passive investment is dominant, with some funds from actively managed mutual funds flowing into ETFs [27][29]. - In the long run, liquidity indicators are not strictly negatively correlated with the US stock valuation. The role of the 10Y US Treasury yield as a global asset pricing anchor has weakened, and the NFCI financial conditions index has a better effect in depicting the tightness of liquidity [32][37]. - To track US dollar liquidity from a quantitative and price perspective, one can focus on the Fed's asset purchase scale from the quantitative dimension and the repurchase market (SOFR - OIS spread), short - term/long - term financing markets (commercial paper spread/credit spread) from the price dimension [43]. 2. US Stock Special Valuation System 2.1 Future 12 - Month Consensus Forecast PE - Based on the future 12 - month consensus forecast PE, the current valuation of the overall US stock market is not cheap. This dynamic P/E ratio, which includes future profit expectations, is more suitable for evaluating high - growth industries [45][49]. - The PE TTM of the "Seven Sisters" in the US stock market is generally higher than the future 12 - month consensus forecast PE [57]. 2.2 ERP (Equity Risk Premium) - No detailed content provided other than the mention of the indicator [63]. 2.3 Global Stock Index Valuation vs ROE Dynamic Comparison - The strong cash - flow creation and profitability of US stock enterprises are the underlying logic supporting high valuations. Among US stock sectors, technology leaders represented by the "Seven Sisters" also have strong cash - flow creation and profitability to support high valuations [69][70]. 2.4 PEG - The PEG formula is P/E ratio divided by expected profit growth rate, which better measures the matching degree between stock valuation and growth. A PEG < 1 indicates a sector with low valuation and high growth. The PEG of the US stock technology sector is still lower than that of non - technology sectors [75][76]. 3. US Stock Trading Heat Tracking System 3.1 Trading - Level Indicators - For short - term timing, one can adopt reverse thinking when considering the trading volume concentration of popular US stock sectors, as buying in the downturn stage has higher cost - effectiveness than in the over - heated stage [82]. - The market breadth of US stocks has been deteriorating in the recent quarter, and the AAII bull - bear spread shows that the bullish sentiment of US stock investors has been weak during this round of the rise [92]. - Other indicators include the 14 - RSI and option sentiment of the S&P 500 and Nasdaq 100, as well as the net inflow of funds into different types of US asset ETFs [95][97][99]. 4. US Stock AI Three - Stage Investment Framework 4.1 Key Targets and Performance Realization in the US Stock AI Industry Chain - The report lists key targets in the US stock AI industry chain, including upstream infrastructure layer, mid - stream, and downstream application/software/end - side companies [103]. - The performance realization path of the US stock AI industry chain generally follows from the upstream infrastructure layer to the downstream application software/media. The upstream industry trend has strong certainty, while the downstream is highly differentiated, and most application - layer business models are still being verified [104][106]. 4.2 Three - Stage Investment Framework for US Stock AI - The direction and degree of performance revision expectations are the core contradictions determining the three - stage market evolution of US stock AI. Since 2023, the increase in different stages of the US stock AI industry chain has basically matched the degree of upward revision of performance expectations. The stock price performance of the three - stage US stock AI also follows the same pattern driven by the upward revision of EPS profit expectations [109][112]. - The commercialization of the advertising and marketing and audio - visual segments in the US stock AI application sector is the fastest, and the differentiation in EPS, revenue growth, and stock price increases among individual stocks in the application sector is well - matched [116][119]. 4.3 Concerns about the "AI Bubble" behind the Recent US Stock Risk - off - The recent borrowing boom of US technology companies has raised market concerns, which are initially reflected in the pricing of the bond market, but currently, it is more of a local and accidental risk pricing rather than a systematic risk pricing [123]. - Overall, US technology stocks have not significantly "over - invested" cash flow in Capex [124].

Summary of Key Points from Conference Call Records Industry or Company Involved - The focus is on the **AI industry** and its related sectors, particularly the **TMT (Technology, Media, and Telecommunications) sector**. Core Insights and Arguments 1. **Timing for Investment in Technology Sector**: On June 8, the technology sector was deemed ready for investment based on analysis of congestion, rolling yield differences, trading volume ratios, and calendar effects, with a specific recommendation to focus on the upstream computing power sector, which has been validated subsequently [4][1][2]. 2. **Trends in AI Market**: The AI market is currently experiencing three major trends: movement from upstream to downstream, diversification within the sector, and the transition from AI to AI-enhanced applications [3][1]. 3. **Internal Differentiation in AI**: There are two main differentiations within AI: between upstream and downstream sectors, and between North American and domestic computing power, primarily driven by performance [5][1]. 4. **Early Stage of AI Market Expansion**: The AI market is still in its early expansion phase, suggesting a strategic focus on the diffusion of domestic computing power into upstream semiconductor equipment and materials, as well as downstream applications in sectors like internet, gaming, and consumer electronics [6][1]. 5. **AI Investment Framework**: The AI investment framework includes key indicators such as timing indicators, calendar effects, internal rotation factors, sector comparisons, and historical references from the 2013-2015 internet boom [7][1]. 6. **Trading Volume Analysis**: Adjusted trading volume ratios show that in 2023, the ratio reached approximately 50%, compared to a maximum of 40% in 2019, indicating a more accurate market analysis [7][1]. 7. **AI Rotation Intensity Indicator**: This indicator tracks the performance of 50 major news directions in the AI industry, showing a significant inverse relationship with the AI index, suggesting that when internal rotation converges, the sector typically experiences an uptrend [9][1]. 8. **Calendar Effects in TMT Sector**: The TMT sector exhibits performance-related calendar effects, with high win-rate periods in February-March, May-June, and October-November, influenced by risk appetite, earnings releases, and consumption peaks [2][10][11]. 9. **Factors Influencing AI Sector Rotation**: The AI sector is divided into three main chains: upstream computing power, midstream algorithm technology, and downstream applications. The North American computing power chain has consistently outperformed the domestic chain since May 2025 [12][1]. 10. **Historical Insights from 2013-2015**: The historical performance of the TMT sector from 2013 to 2015 provides insights into current market dynamics, emphasizing the importance of earnings in driving stock performance [19][20]. Other Important but Potentially Overlooked Content 1. **Potential for Downstream Sectors**: There is significant potential for growth in downstream sectors such as cybersecurity, operating systems, and cloud computing, which are currently underrepresented in institutional portfolios [15][1]. 2. **Market Sentiment and Trading Volume**: Concerns about high trading volume ratios do not necessarily indicate an end to the market rally, as historical trends show that significant market shifts can occur despite high trading volumes [17][1][18]. 3. **AI's Long-Term Impact**: The transition from AI to AI-enhanced applications is expected to mirror past trends seen in the internet sector, with a broader impact across traditional industries [16][1]. 4. **Investment Opportunities in Software Applications**: Areas such as SaaS, online education, and digital marketing are highlighted as having substantial potential for performance improvement and investment opportunities [15][1].