橙子Joyce

Large-scale deleveraging and the selloff in tech stocks in July have just wrapped up, and market sentiment flipped rapidly at the beginning of August.
Goldman Sachs’ latest report shows that investors are quickly rebuilding risk exposure, with demand for call options surging to a historical high. The market has started to enter a “buy the rally” positive feedback loop driven by position unwinds—“the more it rises, the more you buy.” Goldman Sachs liquidity strategist Lee Coppersmith said plainly: “July completed the position reset, and now investors are scrambling to catch up with the action throughout August.”
Data shows that on Tuesday, trading volume for S&P 500 index (SPX) call options surpassed 4 million contracts, setting a record for the highest-ever single-day figure. At the same time, the SPX put/call options skew recorded its largest two-day decline in nearly a decade, reflecting a sharp surge in investors’ demand for upside risk exposure.

Kashkari believes the U.S. economy and job market remain strong. He says current interest rates are not restrictive enough, and inflation is still far from the 2% target; therefore, the Fed should begin rate hikes as soon as possible in a gradual manner rather than waiting until inflation has become entrenched and then being forced to tighten aggressively. He emphasized support for gradual rate hikes but did not commit to taking action as early as September. When asked whether the Fed would raise rates three times this year, Kashkari responded that it is not impossible. Differences within the Fed regarding the rate path are beginning to come into view.
On Wednesday, the Minneapolis Fed president Kashkari publicly stated that now is the time to start slow rate hikes to curb inflation and avoid being forced to tighten sharply later. According to CNBC, Kashkari, who has a voting role on the Fed’s monetary policy committee (FOMC) this year, said in an on-site interview that he leans toward beginning gradual rate hikes as early as September, but he made no clear commitment on a timetable. He stressed that he does not advocate large rate hikes; instead, he wants to act sooner by moving forward in small steps.
This statement sharply contrasts with the position of most FOMC voters last week, and it has made market expectations for policy direction in September and October even more complicated.
Kashkari is one of three dissenting members at last week’s FOMC meeting, along with him. The three regional Fed presidents at the time—including Kashkari—each argued for a 25-basis-point hike, while the other nine members voted to keep the policy rate unchanged. This was the first time since Fed Chair Waller took office in May that the FOMC meeting saw a dissenting vote.
On Wednesday, Kashkari said he is not yet sure what policy action the Fed should take at its next FOMC meeting in September, adding that he wants to observe how subsequent economic data unfold. Asked whether the Fed could raise rates three times before year-end, he replied, “That is not impossible.”
“If inflation continues to level off, or even worsens further, then I think we will have to start gradually adjusting interest rates in order to bring inflation down,” he added.
Economic resilience leads Kashkari to question the basis for the current policy being restrictive enough to warrant hikes. He noted that corporate profits are strong, and both consumers and the labor market remain solid. Against that backdrop, he sees no evidence that monetary policy currently has a clearly restrictive effect.
“Corporate earnings are very impressive, consumers can hold up, and the labor market can hold up. Looking at these developments, I can’t help but ask: what evidence is there that monetary policy is now especially restrictive?” he said. He also said the U.S. economy faces a series of supply shocks that continue to weigh on consumers, and inflation is still some distance away from the Fed’s 2% goal. In his view, rather than waiting until inflation is deeply entrenched and then being forced to hike aggressively, the better approach is to respond earlier with small steps.
Internal divisions in the committee are clear. On the day before Kashkari made the remarks above, Anna Paulson, president of the Philadelphia Fed and another FOMC voter this year, expressed a markedly different view. According to CNBC, Paulson believes that existing evidence shows that the current level of interest rates has already created “moderate restriction” on economic conditions, and she supports keeping rates unchanged while continuing to assess incoming data. She also said that voting to keep rates unchanged at last week’s meeting was “not a difficult decision” for her.
The public disagreement between the two officials reflects the Fed’s internal tension regarding inflation prospects and the pace of policy.
Kashkari said he is still not sure what decision the committee will make at the meeting scheduled for September 15 to 16, and he believes the data at that time will be key. The market currently prices in a slight tilt toward a September hike, while the probability of an October hike is higher.
With no pressure applied, communication strategy still needs clarification
Although Powell previously expressed a preference for lower rates, Kashkari said this Fed chair did not put any pressure on him.
“He told me, ‘Do what you think is right for the economy.’ I said, ‘I really appreciate that,’” Kashkari recounted.
The three dissenting votes were the first instances of opposition during Waller’s tenure, drawing significant attention from the public. Kashkari also pointed out that the FOMC ultimately must decide on an appropriate communication strategy, but he did not disclose specific details. This suggests that internal discussion at the Fed is still ongoing about how to convey policy signals to the market.

Revenue of $7.8 billion, up 92% year over year; diluted loss per share of $0.09.
Net loss of $541 million, an improvement of $467 million compared with a net loss of $1.0 billion in the same period last year.
Adjusted EBITDA was $3.5 billion, up 191% year over year, increasing from $1.2 billion to $3.5 billion.
Business highlights
Backed by an extreme vertical integration model to achieve strong growth; overall revenue from its aerospace business, Starlink communications business, and artificial intelligence business surged 92% year over year.
Within the past 90 days, completed two successful test flights of Starship V3 prototype units, continuously advancing the rapid implementation of rocket full-arc fast reuse technology.

The key focuses of SpaceX’s performance this time are centered on three areas: the subscriber growth rate and revenue scale of the Starlink business, the specific commercialization timeline for Starship, and the direction of the full-year capital expenditure guidance. Before the earnings announcement, the target prices set by various institutions showed clear divergence, reflecting significant differences among institutions in how they assess SpaceX’s valuation.
At the same time, the release of shares subject to lock-up expiry on August 6 is on a massive scale. Since the stock’s listing peak, SpaceX’s share price has fallen by more than 50% in total, and the market’s sensitivity to potential selling pressure before the unlock appears to have been on the high side. If the earnings fail to simultaneously boost market confidence, it is not to be ignored that a high-volume downward pressure trend may emerge around the unlock date. It is worth noting that Musk himself recently responded on a social platform, saying that looking back at the current SpaceX share price would be a "great buying opportunity"; this statement may provide some emotional buffer against the selling willingness of some early shareholders.

A Bank of America Merrill Lynch research report shows that Samsung Electronics will include 60%-70% of storage sales in long-term agreements, with terms clearly favoring the supply side: the scope of price cuts is limited (no more than 5% per quarter), while price increases have essentially no upper limit. Against the backdrop of a surge in AI compute demand and constrained capacity expansion, Samsung uses long-term agreements to lock in major customers while retaining flexibility to raise prices. Spot DRAM and NAND prices continue to rebound before the peak season in the fourth quarter; storage prices are supported upward by both AI capital expenditures and the replenishment cycle.
Demand for AI servers continues to grow, continuously supporting upward pressure on storage prices, and the pricing power of leading storage manufacturers has increased significantly.

Amazon (AMZN.US)
Google-C (GOOG.US)
Microsoft (MSFT.US)
① Wall Street’s views on AI winners and losers are becoming increasingly polarized; ② the three largest hyperscale cloud service providers—Amazon, Microsoft, and Alphabet—saw their combined market value increase by nearly $1.5 trillion this week (about RMB 1 trillion). after reporting strong growth in cloud business.
As global tech giants continue to release earnings and confirm or raise their capital expenditure outlooks, it shows that the AI investment boom has not cooled down. Wall Street’s views on AI winners and losers are becoming increasingly polarized.
So far in this earnings season, six of the “Big Tech Seven” in the U.S. stock market have already reported earnings. This week, their combined market value saw capital inflows of nearly $2 trillion.

Bernstein maintains its “outperform the market” rating on SpaceX, with a target price of $239. It believes the key to long-term valuation hinges on whether the orbital AI data centers can be realized, rather than short-term performance. The emphasis is on Starship achieving fully reusable operations, enabling roughly 3,600 launches per year by 2031. The technical challenges—thermal management, power supply, and latency—are feasible, but semiconductor production capacity, regulatory approvals, and the legal framework remain the primary risks.
After SpaceX’s IPO, the stock price has fallen back by about one-third from its post-IPO peak. However, Bernstein believes the market should not focus on short-term performance; instead, it should focus on the core logic that determines the company’s long-term valuation—whether its orbital AI data centers can ultimately be brought to fruition.