Goldman Sachs trader: The volatility at this time of year is a "normal phenomenon," nothing "abnormal".

Goldman Sachs believes that the recent correction of around 5% in the US stock market is a typical year-end seasonal fluctuation in the AI cycle and does not signal the end of the bullish trend. Goldman Sachs traders point out that despite the market experiencing a correction, there is still upside potential before the end of the year. With seasonal factors, the early stages of the AI investment cycle, and relatively light institutional positions playing a role, indices still have the potential to rise further. According to Shreeti Kapa, a trader at Goldman Sachs Fixed Income, Currency, and Commodities, a 5% decline at this time of year is a normal occurrence in the current cycle. She believes that despite a strong rebound since the April lows, the market overall is not overextended. Kapa's optimism is based on favorable seasonal factors at the end of the year, and she expects the market to have 5-10% upside before the end of the year, accompanied by broad market participation. She points out that many institutional investors are currently skeptical about the future market, believing that this year's market top has already occurred and are adjusting their positions accordingly. However, she believes that this widespread caution actually creates the possibility for the index to "substantially rise" in the remaining 35 trading days of the year. Regarding macro uncertainty, such as the risk of a federal government shutdown, Kapa sees this as a temporary issue. As for concerns about AI potentially replacing white-collar jobs, she acknowledges that it is a pillar of retail demand but believes that it "is not a problem today." Year-end upside potential remains The core logic supporting further market growth before the end of the year is based on the judgment that the AI revolution is still in its early stages. Kapa believes that institutional investors have not fully allocated their positions to the AI theme yet. Additionally, capital flows are expected to become favorable before the end of the year, and the market anticipates that the Fed's monetary policy next year may be even more dovish than last year. At the corporate fundamentals level, although large tech companies are investing significant capital in AI, they have strong balance sheets, reasonable P/E ratios, and compound annual earnings growth exceeding 20%. Kapa emphasizes that the main contradiction in the current market is not concerns about investment returns but rather "addressing spending issues and investing in this potentially historic technological revolution." Is it 1998 or 2000? The discussion about whether the current market is repeating the tech bubble is ongoing, but the key difference lies in today's valuation basis and profitability. Eric Sheridan, head of technology, media, and telecom research at Goldman Sachs, points out that compared to the internet bubble and the real estate bubble periods, there is much more discussion today about the "AI bubble." He acknowledges that private market valuations are much higher than public markets and focus more on revenue growth rather than profits, which is similar to historical risk signals. However, he emphasizes that the valuations of today's listed companies are still based on free cash flow, capital return rates, and profit margins, which is starkly different from the scenarios in 1999 when companies with no revenue received the highest valuations. He states that most of today's tech giants can generate significant free cash flow and conduct stock buybacks and dividends, which were "almost unheard of" in 1999. Furthermore, today's capital market activities are much lower compared to historical bubble periods, and the IPO market is "more discerning." Trillion-dollar investments remain manageable The AI competition has sparked huge imagination about the scale of future capital expenditures, but from a macro perspective, this investment boom may still be manageable. David Cahn, a partner at Sequoia Capital, raised a thought-provoking conversion: he converted the energy requirements for building AI data centers from "gigawatts" to "dollars." He estimates that constructing AI energy facilities ranging from 100 to 250 gigawatts implies data center expenditures of $4 trillion to $10 trillion. He believes that such massive investments must be based on breakthroughs in achieving general artificial intelligence (AGI). However, Joe Briggs, Goldman Sachs's chief economist, provides another perspective. He points out that while the nominal dollar value of AI investments today is huge, when measured as a share of GDP, its impact appears "more moderate." In the past 12 months, AI investments in the US still account for less than 1% of GDP, while historically, peak infrastructure investments have reached 2% to 5% of GDP. The Briggs team estimates that generative AI will ultimately create $20 trillion in GDP economic value for the US. Valuations and positions provide support Multiple data points show that current market valuations and investor positions are still below historical highs, providing potential support for future market movements. According to data from Goldman Sachs Global Investment Research (GIR), the Nasdaq 100 Index is currently trading at a discount of around 46% compared to the dot-com bubble period, and lower P/E ratios indicate that earnings are supporting valuations. Although current bond yields are lower than in the 1999-2000 period, market returns over the past 12 months have been relatively low, suggesting that there is still upside potential in the market in the next 6-12 months. In addition, investor position data also supports Shreeti Kapa's viewpoint. The data shows that after spending most of the third quarter in a neutral position, the current market position is actually in the "underweight" zone, which means that once market sentiment turns optimistic, there will be a significant amount of capital waiting to enter. This article is reprinted from the "Wall Street Seen News" app. Author: Ye Huiwen. GMTEight Editor: Song Zhiying.