Oli.
Oli.
🍓Web3投研 🍑人工智能 🚀《干翻狗庄》系列工具作者
5Following
259followers
Feed
Livestream
Feed
Can SanDisk's high growth continue? There is a very specific test in the October 29 earnings report: beyond price, how much can shipments still contribute?
The company's last earnings report disclosed that quarterly revenue grew about 51% quarter-over-quarter, with roughly one-third coming from increased sales volume and two-thirds from price increases. This breakdown is more useful than just looking at the growth rate alone.
The improvement brought by price hikes is certainly real revenue and indicates that the supply and demand conditions at the time favored the company. But if price increases slow down in the future, whether revenue can maintain rapid growth will depend more on actual shipments and product mix.
I am particularly looking forward to the continued expansion of the data center business. It has already become an important growth direction for the company, but customer deployment timing, purchasing rhythms, and product introductions may cause quarterly performance fluctuations.
So this time, I won't just compare performance against expectations; I will also look at where the growth is coming from and whether the guidance for the next quarter provides new basis.
Especially when the stock price already reflects many optimistic views, impressive historical numbers may not be enough. The market may be more concerned about whether the company can continue to exceed original expectations going forward.
I am optimistic about AI storage demand but do not want to budget future quarters based on last quarter's price increase pace. Growth that comes from selling more products to customers who truly need them while maintaining reasonable profits is the kind of growth that makes me willing to hold longer.
#闪迪存储业务高增长能否延续？
Regarding Meta's AI investment, what I want to track now is how much more it costs for each additional active user.
Muse has already attracted quite a few people to try it, and user growth is encouraging. But personal AI assistants require continuous model operation and task processing; the more frequently they are used, the higher the service costs may be. For this kind of product, there still needs to be some calculation between user activity and profit improvement.
Management mentioned in the earnings call that they hope to significantly improve Muse's cost efficiency over the next year. This is a goal worth tracking, but it cannot be prematurely considered as already achieved profitability.
I am interested in the product direction. A truly useful assistant can indeed bring users back repeatedly; the question is how these uses ultimately convert into revenue and whether they can cover service costs.
If efficiency improves fast enough, the company might be able to control losses while expanding usage. But if users love to use it more and more, yet costs decrease slowly, growth could actually increase short-term bills.
So I hope to see progress not only in registration numbers but also in activity, payment, and cost aspects in the future. At least let shareholders know whether the product is gradually approaching sustainable operation or still mainly relying on advertising business support.
I am not afraid of new businesses losing money temporarily; what I fear is that everyone only celebrates increased usage without continuing to ask how much the company actually has to pay for that usage.
#MetaQ3营收超预期，巨额AI投入引发担忧
If SpaceX and Tesla really merge, there is an issue even harder to clarify than technological synergy: the two companies face very different business environments.
SpaceX has sensitive U.S. government and defense contracts, while Tesla has significant production and sales activities in China. Recent analyses have begun discussing how to handle compliance, information, and operational arrangements if these businesses are placed under the same group.
This does not mean the deal will definitely be blocked, nor can any regulatory agency be conclusively judged at this point. But it reminds me that the difficulty of a merger goes far beyond the two boards agreeing on a price.
Some arrangements are relatively clear when operating separately, but once under the same group, they may need to be re-explained. Which businesses remain independent, which resources can be shared, and what approvals or contract conditions are involved—all must wait for the formal plan to answer.
I understand everyone's expectations for large tech groups, but I also worry that the market might take "the same founder" as a reason to overlook these issues.
For shareholders, synergy benefits are worth calculating, but the costs of maintaining complex business structures cannot be ignored. In the end, if extensive isolation arrangements are needed, the originally imagined shared advantages must be reassessed.
Now I prefer to wait for specific documents. Technologically cooperating and commercially complementing each other are very attractive; but for such a large deal, it is still necessary to clearly explain how the different businesses will coexist.
#SpaceX业绩大增，与特斯拉合并能否推进？
People buying gold to hedge against inflation might feel especially frustrated these days: energy prices are rising, yet gold prices once fell below $4000.
On Friday, spot gold touched around $3975 intraday. This trend reminds me that gold's reaction to inflation is far from the straightforward "prices go up, gold goes up" logic.
If rising energy prices make the market worry that interest rates will stay high longer, non-interest-bearing gold faces opportunity cost pressure. Coupled with the dollar, real interest rates, and position adjustments, short-term performance can completely differ from the most intuitive safe-haven logic.
I do not completely dismiss gold's long-term allocation role because of this drop, but I will reassess the expectations when buying. Using it to diversify long-term risk and expecting it to rise immediately every time the situation tightens are two different needs.
The $4000 round number is very eye-catching, but after breaking below it, we still need to see whether funds continue to sell and how interest rate expectations change. Relying on a single round number makes it hard to judge whether the adjustment has ended or just begun.
What’s most uncomfortable is that the risks we worry about are still there, yet the asset bought for hedging makes the account suffer first.
At times like this, I prefer to clarify the holding period. Long-term allocation requires the ability to withstand short-term drawdowns, while short-term trading needs separate exit conditions and cannot change plans to long-term only after a drop.
#现货黄金跌破4000美元关口
Looking at BTC over the weekend, the easiest mistake is to take the spot market's immediate fluctuations as a basis for conclusions about ETF funds.
BTC trades 24/7, while US stock ETFs have their own trading and reporting hours. If the price rises on Sunday, you can't directly say institutions have repurchased through ETFs; if the price falls, you can't assume a new round of ETF redemptions is happening.
These two sets of timings must be distinguished first.
In recent days, funds have flowed in and out repeatedly, pulling sentiment back and forth several times. Watching the same data refresh over the weekend easily creates the illusion that "institutions are making new moves," but what you might actually be seeing is information from previous trading days.
I prefer to use this time to observe how the price behaves without the support of normal US stock trading hours. Spot market support and derivatives position changes are still worth monitoring, but each requires its own data support.
Wait until the US market reopens to see if ETF funds continue the previous trend; only then can the judgment be more complete.
I have expectations for the buying side to stabilize, but I don't want to impose my preferred explanations on the price every few hours. Especially with the interest rate decision next week, a little weekend volatility can easily be prematurely labeled as a policy-driven market.
What is most needed now might be less jumping to conclusions. Events happening in different markets at different times, forcibly connected into one story, may seem coherent but are not necessarily correct.
#BTC现货ETF资金反复，市场买盘能否企稳？
What storage manufacturers find most difficult now might be that even though customers are urging for deliveries, they still have to restrain the impulse to expand production.
SK Hynix previously discussed increasing capacity and maintaining capital expenditure discipline together. It doesn't sound very exciting, but I actually think this statement is worth remembering.
It's easy to get support for increased investment when there is a shortage, but the real difficulty is judging how much demand will be willing to take delivery at a good enough price after the new capacity is put into production.
Especially for advanced storage, the investment is not just a single equipment purchase order. After factories, production lines, and supporting facilities are put into use one after another, depreciation and other costs will also enter the financial statements. Good profits now do not mean every expansion later will yield the same returns.
I still am optimistic about the storage demand brought by AI, but I don't want to directly extend "customers are currently short of stock" to "any amount of capacity increase can be absorbed."
Going forward, I will pay attention to whether companies explain which products and customers the new investments correspond to, and whether the expansion pace leaves room for adjustment. Advancing with real demand is more reliable than fully planning long-term projects at once.
This round of the market has already provided storage companies with a good business opportunity, and I hope they use this opportunity for a longer time. Earning a little less now and expanding more steadily may be better than having all capacity built in a few years but then starting to desperately cut prices to compete for orders.
#AI需求支撑存储市场，芯片股行情反复
Robinhood's earnings report next week: besides trading volume, I also want to see interest income.
This company already has customer cash, margin financing and securities lending businesses, so revenue doesn't solely depend on how many trades happen today. When users put more assets in, even if they don't move them for now, it can still affect the platform's operating results.
But this also exposes it to another variable: interest rates.
If rates fall in the future, yields on some interest-bearing assets might be pressured; at the same time, lower financing costs could encourage customers to increase activity. How these two effects balance out depends on asset scale, product structure, and the terms the platform offers users.
So I don't like it when people say brokers will definitely benefit just because of rate cuts. For a platform like Robinhood with increasingly diverse revenue sources, you have to break it down.
I want to see whether, after customer funds and assets grow, the company can retain them at reasonable costs, rather than just using short-term promotions to attract moves.
When trading is hot, many operational issues get overshadowed by impressive transaction numbers. Once the market quiets down, the importance of interest income and fund retention becomes clearer.
I'm somewhat positive about Robinhood expanding its revenue streams, but the more business lines there are, the less you can be lazy in your analysis. On October 27, when reviewing the report, besides transaction volume, I'll also look at customer assets and net interest income together.
#Robinhood交易业务增长能否延续？
Looking at SK Hynix's latest earnings report, I want to highlight an issue that is easily overshadowed by the hype around new products: At what pace are customers upgrading to HBM4?
The company previously disclosed that HBM4 has already started mass shipments and will continue to ramp up in the second half of the year; the next generation HBM4E samples have also been sent out. But as technology advances, customers won’t automatically switch all purchases on the same day.
The launch timing, validation progress, and production schedules of different AI platforms will all affect which quarter the orders fall into. An increase in demand for the new generation doesn’t necessarily mean the previous generation products will immediately lose business.
This makes the revenue structure for the quarter more complex than just "HBM4’s share increases, so profits must be better." Older products may already have mature costs and customers, while new products need to gradually expand deliveries.
My expectation for SK Hynix is that it can smoothly connect the old and new products. It should keep up with the next-generation platforms without disrupting the arrangement of products that still have demand just to chase the latest label.
For the earnings report on October 28, I hope to hear more specific details about customer adoption and delivery pace, rather than just how much the next-generation technology is leading.
Investors tend to prematurely turn product roadmaps into profit statements. When the company says samples have been sent, the market starts thinking about mass production; when mass production begins, everyone immediately thinks about next year’s revenue. Technology is indeed progressing, but money is earned batch by batch, so the time lag must be allowed.
#SK海力士Q3财报在即，HBM4能否提振盈利？
The higher the oil price rises, the more I want to see what oil companies do with the money they earn.
The upcoming earnings reports of energy giants may show good profits, but the extra income from high oil prices—whether it is used to pay down debt, pay dividends, buy back shares, or continue expanding investments—makes a big difference to shareholders.
This is also why I think energy stocks and crude oil cannot be simply equated. When oil prices rise, companies have the opportunity to earn more; but if management treats the current high prices as a long-term state and immediately expands capacity based on the most optimistic demand, shareholders may not be comfortable when the cycle later declines.
Especially since new projects take time from investment to production. Just because supply is tight today doesn't mean the same prices will hold when production starts years later.
What I want to see is a company acknowledging that part of the current earnings comes from a special supply environment, then arranging long-term budgets based on more conservative prices. Exercising restraint when making money reassures me more than constantly announcing big projects.
For holders, it is also worth comparing which companies have truly improved their balance sheets and which only have temporarily attractive income statements.
High oil prices can certainly excite people, but how much shareholders ultimately get depends on how management allocates that money. I don't want to foot the bill for a potentially overexpanded plan at the peak of the market.
#中东原油供应风险再起，油价持续走强
Before this rate decision, what worries me more than the rise in oil prices is that consumers are beginning to feel that prices will be hard to bring down in the coming years.
In the University of Michigan's October survey, consumers' one-year inflation expectations rose from 4.6% to 4.7%, and long-term expectations also increased from 3.4% to 3.5%. These are respondents' expectations, not inflation that has already occurred, but for the Federal Reserve, this likely cannot be taken lightly.
A one-time energy price increase can wait for supply to recover; if everyone gradually accepts "everything will be more expensive in the future," business pricing and household consumption plans may also change accordingly, making it more difficult to handle later.
The same survey also shows that conditions for purchasing big-ticket items have clearly worsened. Consumers worry about price increases while feeling unable to afford purchases, a situation that is hard to ease with just a single pause in rate hikes.
I think the most difficult question for next week's meeting is this pressure from both ends. Continuing to tighten increases the burden on households; waiting too long risks price expectations becoming entrenched.
So if a pause really happens, I would first interpret it as a temporary observation. For BTC and tech stocks, one can expect policy not to immediately tighten further, but don't assume that subsequent rate cuts will follow right away.
What makes people anxious now is no longer just the possibility of rate moves this month, but how long this expensive living and financing environment will persist.
#10月FOMC临近，美联储暂停加息还有哪些变数？
