On August 23, a research report released by CITIC Securities indicated that the recent downturn in tech stocks cannot be attributed solely to the increase in U.S. long-term bond yields. Instead, the crux of the matter lies in the forward-looking pricing dynamics of AI-related stocks. This scenario is influenced by three pivotal variables: First, whether the pace and magnitude of commercialization can align with market expectations; second, whether computational advantages can be successfully converted into market share and pricing power; and third, whether the existing computational gap will exert a substantial influence on the evolution of future AI models. At present, the market is primarily focused on the pace and scale of commercialization, while the most significant potential variable is whether "anti-distillation" technology could potentially widen the model gap once again in the future.

In terms of macroeconomic factors, the U.S. Treasury's repurchase of long-term bonds has a relatively limited impact on the market. Conversely, the short-term weakening of the U.S. dollar and diminished expectations of interest rate hikes are expected to help narrow the K-shaped divergence observed in global markets. Nevertheless, the underlying factors propelling the sustained rise in U.S. long-term interest rates remain intact and could continue to unsettle the market in the days ahead. Against this backdrop, the short-term capital structure of A-shares is intricate, heightening the complexity of market speculation. Consequently, investors are advised to temper their expectations and refrain from over-interpreting broad market narratives.