My Pyramid Boosting Plan - Buy as Much as Falls, but Always Assuming It Will Fall More Than You Think At the end of this article, it was mentioned how to increase long-term positions. Now let's talk about it. First, let me express my judgment This pullback is macro driven, not a fundamental collapse. The earnings season is basically over, and the remaining fluctuations mainly come from interest rate expectations, geopolitical risks, and liquidity siphoning from SpaceX's IPO. This macro sentiment driven pullback has always been an opportunity to add positions in history, as the next earnings season will re validate the logic of AI fundamentals. So I have already figured out how to buy in batches. Pyramid positioning plan to buy individual stocks, using the overall market as a reference The first batch is around S&P 7350, with a 10% position. But it's not about buying when the price arrives, but waiting for the buying signal to appear after Monday's inertia stabilizes. One or two hours before the opening, there is a high probability of mechanical selling with margin calls. Don't rush in first, wait for the market to tell you it has stabilized. The second batch is around S&P 7260, with a 20% position. There is a gap left in the early stage, and technically there is a gravitational pull to fill the gap. If it falls here, it will be a better entry price. The third batch is located near S&P 7150, with a 30% position. This is a chip intensive platform formed during the consolidation in April, with strong support. To be honest, this is the endpoint position that I think this adjustment is more reasonable in my heart. Starting from 7630, there has been a decline of approximately 6%, overall within the range of 5 to 10% correction, which is a normal and healthy correction in a bull market. If I can fall to this point, I will be more confident that this is the bottom range, which will increase my position a bit. The fourth batch will reserve 40% of the positions in S&P 6800 to 7000. This is a strong support range that has been sideways for a full four months from January to April this year. To be honest, I'm not sure if I'll make it to this position, but I have to keep it. On the one hand, it is the safety margin, and on the other hand, if it really falls here, it will be the best opportunity to buy at the bottom in the second half of this year. When it arrives, it is time to heavily invest. If it hasn't fallen, use this 40% as defensive cash and wait for the signal of stabilization on the right side before choosing an opportunity to increase positions. Why do I think the US stock market will continue to rise in the second half of the year Firstly, the commercialization of AI is still accelerating. The budget for enterprise AI continues to grow, and the demand for inference is taking over the training needs. Agents are moving from demos to production environments. This is the real change that can be seen in the financial report data. Secondly, Hyperscale's capital expenditures are still expanding. Microsoft, Google Meta、 Amazon has not yet signaled any reduction in AI investment, and is expected to exceed $1.1 trillion next year. The AI infrastructure cycle is still ongoing, and the demand for upstream hardware will not disappear out of thin air. Thirdly, the valuation expansion in the early stages of the technological revolution lasted much longer than market expectations. You can refer to 1999, when the macro environment was not entirely the same, but history has repeatedly proven one thing: as long as the fundamentals continue to cash in, liquidity tightening may not necessarily be the end of the tech bull market, it only slows down its pace. Fourthly, potential IPOs such as OpenAI and Anthropic indicate that the demand for AI assets in the capital market remains strong, and high growth AI companies can still achieve high valuations. Before these most important unicorns went public, Wall Street had no incentive to let the market collapse. The fundamental logic has not changed, the more the price drops, the better the odds. Under what circumstances would I change my judgment Don't watch interest rates rise, short-term crashes, or market sentiment collapse. These are all noises. There are only a few signals that can truly change my judgment: Firstly, Hyperscale has started to lower its expectations for future capital expenditures If Microsoft, Google Meta、 Amazon has begun proactively lowering its data center investment plans for the next 12 to 24 months, indicating a change in management's expectations for the growth rate of AI demand and return on investment. This is more alarming than quarterly performance fluctuations, as the capital expenditure decisions of these companies themselves are the most direct vote on future demand. Secondly, the revenue growth rate of the AI application layer has significantly slowed down OpenAI, Anthropic, and other leading AI platforms are the most direct manifestations of the demand for AI commercialization. What needs to be paid attention to is the expansion speed of enterprise customers, the growth trend of API call volume, and whether the net revenue retention rate continues to weaken. If these data deteriorate simultaneously, it indicates that there is a real problem on the demand side. Thirdly, there is a systematic contraction in enterprise AI budgets Optimizing token costs for enterprises is not a problem in itself, it is a logical progression from continuously optimizing cloud computing costs in the past. We need to pay attention to whether the overall AI budget is starting to decline and whether AI projects are shifting from the expansion phase to the reduction phase. If this phenomenon spreads from individual cases to the industry level, it means that there is a substantial change in the demand side. Fourth, the return on investment of AI continues to deteriorate If the overall AI related revenue growth in the industry is significantly lower than the capital expenditure growth, and this situation persists for multiple quarters, the market will eventually begin to question the rationality of AI infrastructure investment. The change in ROI is ahead of market sentiment, which is a core indicator that needs to be continuously tracked. Fifth, the improvement in modeling capabilities cannot continue to be translated into commercial value If the performance of the model continues to improve, but the willingness of enterprises to pay and the creation of commercial value begin to stagnate, the entire AI investment logic needs to be re examined. summary The core logic of this position increase plan is only one sentence: the fundamental logic has not changed, and the more the price drops, the bolder it is to increase positions. Now that the short-term earnings season has ended, the volatility in the coming weeks will mainly be dominated by macro sentiment. The next financial reporting season is the real validation window. If Nvidia, Microsoft, Google, and Meta had once again submitted strong AI related data at that time, all those who sold in panic this time would regret it. Keep cash, wait for prices, and execute according to plan.