Gf Securities Co.,Ltd. has released a research report analyzing the likelihood of single tops in bull markets. The firm notes that the recent Ai-driven rally across markets in China, the US, Japan, South Korea, and Taiwan since the end of the first quarter is primarily fueled by a strengthening of the main industrial trend, not one-time factors. The exception may be South Korea's leverage, which could pose a sharp peak risk for storage stocks. For other sectors in other markets, the top is likely not a single, sharp peak but a complex, multi-layered structure. If new vertical models trigger further demand, these sectors may even hit new highs.
In July, major global equity markets experienced synchronized pullbacks, with tech assets seeing steep declines, forming clear sharp top patterns in the short term. The key question is whether such sharp or single tops are a common occurrence or an exception during bull market peaks. Gf Securities Co.,Ltd. explores this from the perspectives of top formation patterns and fund allocation ratios.
Market Peak Shapes: Single Tops vs. Multiple Tops
Single tops are a market rarity. In A-shares, they occurred only in 2015. The Dow Jones and S&P 500 have never exhibited them, and the Nasdaq only did so in 2000. Two scenarios can trigger a single top: an irreversible liquidity shock (deleveraging) or a rapid reversal of the profit cycle (one-off earnings or a strongly cyclical shift). Currently, if Ai forms a sharp top, every rebound should be used to reduce positions. If it forms a repetitive multi-top structure, positions can be maintained in anticipation of a second-half rebound. If Ai has the potential to reach new highs, now is an excellent time to add holdings.
Historically, only the 2015 internet-plus bubble and the 2000 tech bubble are classic examples of sharp tops. Most other cases have been repetitive multi-top structures. The formation of a sharp top always involves a one-time factor beyond the main industry trend. For the 2015 internet-plus bubble, the one-time factor was leveraged funds, with the bubble bursting due to regulatory deleveraging. For the 2000 tech bubble, it was the surge in capital expenditure for computer replacements spurred by the Y2K crisis; the bubble burst when the Y2K threat was disproven in early 2000, making replacements unnecessary. Returning to the present, the Ai rally since Q1 in markets like China, the US, Japan, South Korea, and Taiwan is driven by a strengthening of the main industrial trend, not one-off factors. Apart from South Korea's leverage, other sectors in other markets are unlikely to form sharp tops and will more likely develop repetitive multi-top structures. Future innovations in vertical models could spark new demand and push prices higher.
Fund Allocation Ratios: Single Tops vs. Multiple Tops
In rare cases, fund allocation shows a sharp top, quickly dropping from high levels. This typically happens when the fundamental outlook for a heavily allocated industry deteriorates and is quickly disproven. Few examples include fund allocations for baijiu in 2012 and non-bank financials in late 2014. The sharp decline in these allocations over the next two quarters was due to a rapid, hard-to-reverse deterioration in the fundamental logic, prompting a quick, consensus-driven rebalancing by institutions.
In most cases, institutional allocation ratios oscillate at high levels for a period, forming complex multi-top patterns. Key examples from industry cycles include: major financials from 2007-2009; mobile internet from 2013-2015; supply-side reform from 2016-2017; core assets from 2019-2021; and the new energy industry from 2021-2022. During these cycles, institutional holdings rose as industry logic was continuously validated. Fund allocation peaks in these cases rarely appeared as sharp tops. Institutional investors move from forming an industry consensus to allocating to the most certain profit directions, and finally to confirming the slowdown of the industry cycle. During this process, allocation ratios fluctuate at high levels, showing complex, multi-top characteristics.
Firstly, the development of a major industry cycle is not linear. Whether it's technological breakthroughs, policy benefits, or business model innovation, institutions need to gradually validate new logic. Even if holdings fall after hitting new highs, new catalysts can push them to new peaks. Secondly, if the industry chain is broad and deep enough, different segments benefit in different orders. During this process, institutions optimize their structures, and different peak allocation ratios correspond to different heavily held companies. Finally, when a high-growth industry slows down, the market is often divided. Two key empirical rules for performance near a turning point are [30%] and [-50%], requiring extensive research and verification. As fundamentals confirm a slowdown, allocation ratios are slowly digested from high levels.
Looking at the present, based on recent earnings guidance from core North American CSP companies, it's difficult to compare Ai to the baijiu sector in 2012 or brokerages in early 2015, where the fundamental logic deteriorated essentially. Therefore, referring to more industry cycle cases (mobile internet, core assets, new energy), the layout for new technological revolutions often progresses upward amidst twists and turns. Even if a single quarter's impulse value requires short-term correction, tracking subsequent catalysts and commercial progress is more critical.
Risk Warning
Risks include geopolitical conflicts exceeding expectations, leading to greater-than-expected upward pressure on global inflation; overseas inflation and US economic resilience causing a faster global liquidity tightening cycle; and weaker-than-expected domestic stabilization efforts in China, resulting in sluggish economic recovery and reduced market risk appetite.
Comments