GF Securities released a research report stating that while visibility for overseas AI supply chains, such as optical communication, has extended to 2027, making them a confirmed growth direction and a primary holding for institutional investors, these sectors remain relatively linked to Middle East conflicts. Consequently, short-term volatility remains difficult to control. Drawing from the experience of the internet technology bubble, industries whose growth trends are relatively insulated from geopolitics and high oil prices should hold allocation advantages regardless of how US-Iran tensions evolve. Therefore, from the perspective of managing portfolio volatility and hedging, the firm recommends allocating to two additional sectors with improving fundamentals that are less impacted by oil prices, alongside overseas computing power: the energy storage chain (inverters/lithium battery supply chain) and the domestic AI data center chain (particularly the ByteDance supply chain). GF Securities' key views are as follows:

1. 1999-2000: The Impact of the Kosovo War on the Internet Technology Sector In last week's report, the firm focused on the sequence of events: the Kosovo War leading to rising oil prices, which fueled US inflation and prompted the Federal Reserve to raise interest rates, and the subsequent effects on the US broad market and internet technology stocks. The core conclusion was that oil price increases driven by the war pushed inflation higher, temporarily suppressing the Dow Jones Industrial Average in Q3 1999, before it peaked and began declining in January 2000. In contrast, the NASDAQ internet technology sector demonstrated stronger performance, remaining unaffected by monetary tightening, with its bubble peak occurring nine months after the Fed's first rate hike.

2. 1999-2000: History shows that independent, high-growth industries can temporarily overcome high oil prices and interest rate hikes. (1) Denominator Side: The Kosovo War triggered liquidity tightening. OPEC production cuts in early 1999, combined with the Kosovo War (March-June 1999), drove oil prices from $10 per barrel to over $30. The Federal Reserve initiated a new rate-hike cycle in June 1999, implementing six increases by May 2000, leading to a shift towards tighter liquidity in the US stock market. (2) Numerator Side: High-growth expectations supported by the Y2K replacement cycle. The Y2K narrative developed from 1996-1998, amplified by mainstream media coverage, US Congressional records, and presidential addresses. Critical systems in finance, healthcare, defense, and government prioritized fixing hardware and software bugs, leading to a surge in orders for the internet technology supply chain during 1998-1999. Correspondingly, leading internet companies reported strong 1999 earnings: Dell's net profit grew 55%, Microsoft's 73%, IBM's 22%, and Intel's 21%, all maintaining high growth or accelerating significantly from 1998. Market expectations were that the replacement cycle driven by potential system failures would continue into the year 2000. (3) 1999 Market Performance: Strong fundamentals overcame liquidity tightening, leading to an internet valuation bubble. The NASDAQ 100 index (comprising leading companies) achieved 60% EPS growth in 1999, accompanied by a TTM P/E of 95x and a forward P/E of 65x. 1999 was a year of intensifying bubble conditions; anticipation that the millennium shift would further drive replacement cycles for small businesses and individuals meant strong internet sector earnings outweighed the impact of rising interest rates, pushing the NASDAQ to new highs. (4) 2000 Market Performance: When fundamentals deteriorated, the final straw broke the camel's back. The Y2K crisis did not materialize, with media outlets labeling it a "century-long scam" fabricated by tech companies. The internet sector peak in March 2000, seemingly punctured by the "Microsoft antitrust case," was fundamentally caused by a negative feedback loop. Companies' premature investments in system upgrades and replacements had透支 (overdrawn) industry demand for the next 2-3 years. Financial reports from leading internet companies in March-April 2000 widely missed expectations, and some .com companies began declaring bankruptcy. By the end of 2000, US computer and electronics inventory reached historically high levels, with excess stock further suppressing future performance. By 2001, NASDAQ 100 profit growth had plummeted from 60% in 1999 to -50%, marking the end of the internet boom.

3. Returning to the Present: Which industries might maintain independent high growth, insulated from geopolitics and high oil prices? Currently, overseas AI supply chains, such as optical communication, have deepened visibility through 2027 and remain a confirmed growth direction, constituting a major position for institutions. However, their performance is relatively linked to the evolving Middle East conflict (through the channel of oil prices → US interest rate environment → US AI sector → domestic supply chain), making short-term volatility hard to manage. Learning from the internet era experience, identifying sectors poised for sustained independent growth is crucial. If a sector's growth trend is relatively detached from geopolitics and oil prices, it should offer allocation advantages regardless of how US-Iran tensions develop next. Therefore, for managing portfolio volatility and hedging, GF Securities recommends allocating to two sectors with inherently improving fundamentals and lower sensitivity to oil prices, in addition to overseas computing power exposure: the energy storage chain (inverters/lithium battery supply chain) and the domestic AI data center chain (especially the ByteDance supply chain).

Risk warnings include geopolitical risks, overseas inflation risks, and potential underperformance of domestic economic stabilization policies.