U.S. stock investors' feedback... beyond imagination pessimism

The US stock market trembles... What a day...

Deepseek knocked everything down, no exaggeration; I went out for drinks with a childhood friend last night, and while tipsy, I watched Nvidia drop like this from left to right...

OK, without saying too much else, let's directly look at some feedback from trading desks / investors;

1/ Institutional buyers are unexpectedly pessimistic about this drop; of course, this trend is more evident among Generalists, while specialists focused on technology are still showing signs of buying the dip;

2/ Bears believe this is the "Cisco moment"; although the capex for 2025 has already been booked, an oversupply of capex will emerge in 2026; Microsoft's retreat from the Stargate has exacerbated this "ghost story";

3/ Bulls believe that some other investors have misunderstandings about the training costs of Deepseek; the results of Deepseek, in the long run, equate to more AI usage + productivity improvements;

4/ Last night, many opinions discussed Jevons Paradox (refer to our public account from yesterday); from this perspective, semis are still relatively dangerous in the short term; however, internet + software will be much safer; MSFT, META, AMZN, DDOG, MDB, HUBS, CRM, GTLB, TEAM, SNOW; this line of thought is quite similar to what we discussed in yesterday's daytime article.

5/ From the perspective of semi rotation, the market is still relatively healthy; because fear has not spread out; the reason for yesterday's QQQ decline largely came from the weight of NVDA and AVGO; excluding semis, it was basically a flat trading day.

6/ Everyone is waiting for MSFT's earnings report + conference on Wednesday; currently, the bulls have very few weapons (whether in narrative or data points), and there is not much to counter the bears... From yesterday's feedback, even Jevons Paradox was heavily criticized by the bears; The reasons are: 1. The "compression" of inference computing power is already so terrifying. If DeepSeek or any other company continues to break through in architecture in the short term, what kind of even more exaggerated compression could we see? 2. Without an AI killer app, there can be no turning point for the Jevons Paradox; we have been waiting for an AI killer app for two years, so how many more years will it take? Right now, the pace of technological iteration is too fast, with the technology cycle >> expenditure cycle >>>> application cycle.

7/ The first trading day after DeepSeek; it's too early to say too much now. We mentioned in our public account a couple of days ago that many investors (especially generalists) only heard the news for the first time over the weekend and were still at a Wikipedia level of understanding about who Liang is / what DeepSeek is on Monday; refer to this article: Why are foreigners lagging behind on DeepSeek? (January 26)

8/ Continuing to wait for the next day's trading; sentiment / market narrative is created through trading, not imagined.

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment based on this is at one's own risk