李志 | Rationalnvest
李志 | Rationalnvest
🐶金融行业底层人士。 📈美股行情观察 & 数字游民分享 & 出海探索。 🏃♂️ 严肃的跑者 (全马350、半马150)。 👆关注我，摆脱焦虑，我们一起慢慢变富。
2Following
51followers
Feed
Feed
【US Bank Account Basics】What's the difference between Checking and Savings?
When I opened an account in the US, I never quite understood why, even though both can hold US dollars, there are two types of accounts. Many big banks offer low interest on Savings accounts and even charge monthly fees—so what’s the point?
This educational video starts with the history of US banks and, combined with my actual experience opening and using accounts at Chase and BOA, clearly explains the purposes, interest rates, and maintenance costs of these two account types.
It also discusses how many accounts "cloud residents" like us—who invest in US stocks but don’t live in the US—actually need to open.
According to Reuters on October 19, OpenAI is accelerating a new round of financing of about $30 billion, targeting a valuation exceeding $1.4 trillion.
Investment institutions such as SoftBank and Abu Dhabi's MGX are in talks to participate.
Notably, OpenAI hopes to lock in investor commitments ahead of Anthropic's expected IPO in mid-November; these two companies are competing for the same pool of investor funds.
Currently, OpenAI is showing investors an annualized revenue of about $50 billion, expected to reach $70 billion by the end of the year. Meanwhile, Anthropic's third-quarter revenue has already reached $16.8 billion, also aiming for a trillion-dollar valuation.
OpenAI not only wants to secure funding but also hopes to establish a new valuation benchmark for itself before Anthropic discloses more financial data upon going public.
Previously, OpenAI's valuation was $852 billion in March this year, and now it is seeking to surpass $1.4 trillion.
South Korean AI data center company Elice plummeted 32% on its first day of trading. This cloud computing company saw its market value evaporate by about 300 billion KRW in less than half an hour after listing.
The IPO issue price was 90,500 KRW, with the intraday low dropping to 61,400 KRW, a decline of up to 32%. Before listing, retail investors subscribed over 400 times, attracting about 10.3 trillion KRW in subscription funds.
Revenue in the first half of the year surged 200% year-over-year, with an operating profit margin close to 30%, already profitable.
Although Elice has revenue and profit, and its valuation is not particularly aggressive, it still faced market sell-off.
This is not an isolated case.
Previously, Firmus, supported by Nvidia, withdrew its IPO after lowering the issue price; AirTrunk postponed its listing; recently, CoreWeave’s stock price also sharply retreated from its highs.
A Goldman Sachs analyst’s comment may explain the biggest contradiction in the current AI infrastructure industry:
AI computing power demand can grow exponentially, but the capital to support this growth cannot expand synchronously.
Netflix's recent round of layoffs even saw the executive responsible for the layoffs being laid off themselves...
According to The Wall Street Journal, on October 14, Netflix announced layoffs of about 1,000 employees, accounting for 6% of its global workforce.
The most striking detail was on the day of the layoffs:
At 2 PM, laid-off employees lost access to their company email accounts.
Just a few hours later, Netflix co-CEO Ted Sarandos informed Chief Content Officer Bela Bajaria that she would also be leaving the company.
More dramatically, according to Hollywood media reports, Bajaria had previously participated in this round of layoffs but ended up being laid off herself.
She worked at Netflix for nearly 10 years, overseeing global content operations and contributing to the development of hit shows like Squid Game and Stranger Things.
However, amid slowing growth and pressure on user watch time, even former key contributors couldn't keep their positions.
Subsequently, Dan Lin, head of Netflix's film division, also announced his departure. In just a few days, about 1,000 employees were laid off, and two content executives exited one after another.
The Q3 earnings report will be released on October 20, so let's wait and see.
$NFLX
U.S. stocks rebounded tonight at the open, with the Nasdaq up 1%, and chip stocks also recovering...
Looking at the S&P 500 index, it rose about 0.5%, with the semiconductor sector leading today's rebound. AI-related stocks like Nvidia, AMD, Broadcom, and Marvell Technology generally climbed, and the semiconductor ETF (SOXX) surged as much as 3.6% in pre-market trading.
This Wednesday, Meta and Tesla will release their earnings reports, and the market will continue to focus on large tech companies' AI capital expenditures, profit growth, and future guidance.
Personally, I think today's rise should be seen more as an emotional repair; the direction of the upward trend still needs to wait for clearer signals.
US bank accounts, do ordinary people really need to open one?
A few days ago, after posting videos about opening accounts with Chase and BOA, some comments said that if your assets don't reach a certain scale, it's not worth the hassle.
My criterion is very simple: do you have an actual need for overseas funds?
For me, the most practical use of a US bank account is to connect US stock investments, USD deposits and withdrawals, and credit card repayments.
For example, during my recent trip to the US East Coast, I paid for hotels, dining, and shopping with a US credit card, then smoothly repaid it with USD from my overseas account.
This video combines my personal experience to talk about what problems a US bank account can solve.
Here's something fun for you:
Claude and GPT's interpretations of the little fox in my avatar. I asked them to generate four cartoon images to easily add when editing videos. The first one was generated by Claude, which is way too abstract...
The second one is from GPT, which feels much better and very cute.
Goldman Sachs data shows that hedge funds sold off all the AI stocks they had cumulatively net bought since early 2025 within just 5 trading days.
According to the latest data from Goldman Sachs Prime Brokerage, from October 8 to 14, hedge funds net sold global AI-related stocks for five consecutive trading days, with the cumulative net sell-off offsetting all net purchases since early 2025.
At the same time, global hedge funds experienced concentrated deleveraging:
Global fundamental long-short hedge funds: down 2.9% for the week ending October 15.
Asian equity long-short hedge funds: down 6.8% cumulatively in the first half of October.
Asian equity long-short hedge funds: down 2.6% on October 15 alone.
Semiconductor and AI supply chain stocks in South Korea and Taiwan were the main targets for reduction.
Goldman Sachs pointed out that since 2026, different hedge fund strategies have experienced three rare synchronized sharp declines, occurring on February 4, July 28, and October 15.
Behind this round of sell-offs is, on one hand, the market beginning to reassess the return cycle of AI capital expenditures, and on the other hand, rising financing costs putting greater pressure on highly leveraged AI investments.
The long-term demand for the AI industry and the short-term position adjustments by institutional funds are two different matters. However, when a large amount of capital concentrates in the same trading direction, deleveraging itself can further amplify market volatility.
Morgan Stanley released a capital expenditure flow chart for AI, predicting that by 2027, the capital expenditures of major global cloud providers and emerging cloud service providers will reach $1.58 trillion.
Where will this money ultimately flow?
According to Morgan Stanley's estimates, about $1.03 trillion will convert into related revenues of S&P 500 constituent stocks, among which:
1. NVIDIA alone accounts for about $690 billion, or 67%.
2. The remaining 499 S&P 500 constituents: a total of about $339 billion.
3. The rest of the funds flow to Korean memory manufacturers, Taiwanese server foundries, and other non-S&P 500 suppliers.
From the industry chain perspective, the 2027 capital expenditures are mainly distributed in the following areas:
1. Semiconductors and servers: $672 billion
2. International suppliers: $511 billion
3. Memory and storage: $217 billion
4. Networking and optical communications: $215 billion
5. Data center infrastructure: $119 billion
As of Q2 2026, major cloud providers have signed but not yet started lease commitments totaling $1.118 trillion.
Morgan Stanley expects 2027 to be the year with the largest free cash flow gap for these cloud providers.
Meanwhile, AI capital expenditure growth is expected to slow from about 60% in 2027 to about 14% in 2028.
