Group 1 - The report highlights that since October 2025, price increases have gained attention, particularly in the fields of new energy, AI, and certain black, non-ferrous, and chemical products, with many areas experiencing transaction congestion close to historically high levels [4][10][12] - Historical cases indicate that price increases driven solely by sentiment or expectations typically reach a peak when congestion levels hit 90-100%, leading to a phase of adjustment, and subsequent rebounds are unlikely to surpass previous highs without additional supporting logic [14][16] - For price increases to break previous highs after an adjustment, they usually require new incremental logic support, such as a shift from expectation to reality in economic verification or the emergence of new catalysts [14][16] Group 2 - The report identifies two main areas where price increases are likely to continue: industrial metals, driven by global economic recovery expectations and supply constraints, and the AI chain, which remains a direction with confirmed economic prospects and potential incremental catalysts [51][52] - Strong sectors often reach a stage of congestion bottom when sentiment (transaction share) declines to 50-70% of previous highs, presenting a good buying opportunity [52] - The report suggests monitoring the TMT sector's transaction share to determine when it returns to the 20-25% range, indicating a potential buying point, while the Hang Seng Technology sector should be observed for a return to the 30-35% range [60][66] Group 3 - The report discusses the historical adjustment patterns of major technology sectors, noting that the average adjustment period is around 40 trading days, with an average absolute decline of approximately 15% [71][72] - The current adjustment in major technology sectors has seen declines of 15-20%, nearing historical averages, but the adjustment duration has been shorter than the historical average, suggesting a need for patience [71][72] - The report recommends continuing to allocate resources according to calendar effects, particularly in banking and white goods, while observing potential shifts in growth styles as economic data is released [71][72]