对冲积鲸Reason

The gold-silver ratio has officially broken below 50, the last time being in 2011 when it even dropped to a low of 32.

The most common way for silver to die is not bad news, but risk control.

During the sharp correction of silver in 2011, CME continuously raised margin requirements in a short period, increasing trading costs by 84% over 8 days, which reinforced the sharp decline.

Ctrip this time, the fewer the words, the bigger the issue.

Regulation has escalated from mere talks and rectification to formal antitrust investigation.

The nature has upgraded.

Figure 2 is September 2025, Figure 3 is this time.

Earnings Whispers has once again organized the upcoming one-month US stock earnings schedule, still recommending saving and bookmarking to avoid forgetting.

Here are several key time points summarized for you—don't miss them.

Week 1 (January 13-16): Focus on financial stocks—JPMorgan Chase JPM, Citigroup C, Bank of America BAC. On Thursday (15th), TSMC TSM's earnings report will serve as the first major industry signal for the semiconductor sector.

Week 2 (January 20-23): Tuesday—Netflix; Thursday—Intel.

Week 3 (January 28-29, Super Week): The most critical week. Microsoft, Tesla, and ASML are all scheduled for Wednesday (28th); Apple and Amazon, unspecified, usually released concurrently or will appear on Thursday (29th).

Week 4 (February 3-4): AMD, Alphabet (Google), and ARM release earnings, continuing to validate demand for AI chips.

The earnings season typically starts with bank stocks, which reflect the overall economic condition; then attention shifts to streaming and consumer electronics earnings; finally, as chip, AI, and major internet company reports are released, market sentiment peaks.

Tonight's CPI report was another one with surprise but no major disruption. Although food prices were alarming, core inflation remains moderate, not undermining the narrative for rate cuts.

1. Overall CPI year-over-year rose 2.7%, in line with expectations and unchanged from the previous reading. Month-over-month, it increased by 0.3%, with housing being the main driver, alongside contributions from food and energy.

2. Core CPI, excluding food and energy, rose 2.6% year-over-year, matching the prior figure, with a month-over-month increase of 0.2%. The core month-over-month growth was lower than the overall CPI, indicating that underlying inflation pressures, after removing volatile items, remain relatively light.

3. This report shows a clear pattern of strong services, weak goods, and an unexpected surge in food prices.

Housing rose 0.4% month-over-month, remaining the single largest driver of inflation. Despite market expectations that rent growth would slow, the data shows its persistence remains strong, which is the primary reason core inflation cannot fall significantly further.

Food prices surged 0.7% month-over-month, a remarkably high increase. Both household food and dining-out prices rose noticeably. This directly pushed up the overall CPI figure and explains why average consumers are feeling higher inflation this month.

Energy prices increased 0.3% month-over-month, marking a rebound after a period of decline. This ends energy's previous drag on inflation, though the rise remains within manageable limits.

Prices for medical care, personal care, clothing, airfares, and education services all posted month-over-month gains.

However, communications, home furnishings, and operating costs saw month-over-month declines.

4. Although overall CPI appears slightly elevated at 0.3% due to the food price rebound, the Fed's key metric—core CPI—rose only 0.2% month-over-month, indicating that underlying inflation pressures are under control.

Following the data release, short-term interest rate futures rose, and traders increased bets on a Federal Reserve rate cut. As long as core inflation stays within the moderate 0.2% range, even with temporary fluctuations in food and energy, the Fed still has room to continue its accommodative policy into 2026 to support the labor market.

Some market changes and strategic thoughts from the past few days,

1. BNB Chain DEX trading volume has exploded, with BNB DEX 24h trading volume surging and at one point surpassing Solana.

2. Western media are reporting on the BNB meme season, and the hype is spreading beyond the Chinese-speaking community.

3. The BNB/SOL ratio is close to 6, reaching the highest point in two years (over the past two years, it has mostly fluctuated between 2.5 and 5).

4. The open interest in SOL derivatives is relatively high, and short selling needs to be cautious of short squeeze risks.

5. Combining points 1, 3, and 4, if you want to conduct pair trading on BNB/SOL in the short term, you must always pay attention to the divergence in trading volume between the two chain DEXs.

6. Meme funds cannot stay on one chain forever; it's time to pull back when needed.

$BNB $SOL $币安人生

The economic world does not have eternal prosperity, only the cycle of recurrence.

In 1929, people believed in perpetual prosperity, but the Great Depression taught them a lesson.
In 2008, people believed that housing prices would always rise, but the subprime mortgage crisis left them destitute.

Why do people always make the same mistakes? Because at the peak of the bubble, greed blinds the eyes and makes people forget the most basic common sense. The economy has its spring, summer, autumn, and winter; growth has its limits.

We should not try to predict when the wind will stop; if we feel the wind is getting too strong, we should turn around and seek shelter in the harbor.

We must have reverence for the cycle; the most important thing is not to leave the table.
$BTC $ETH $SOL

After talking for so long, Oracle is working hard as a landlord, while OpenAI is looking for cheap computing power everywhere. Everyone is calculating in this game to the point of bloody battles. So who is that person comfortably fishing and counting money?

The answer is only one: Nvidia, the one selling shovels. In this AI gold rush, whether you are Oracle, Google, or Amazon, whether you are OpenAI or any other large model company, you have to buy GPUs from Nvidia. It is the only indispensable arms dealer. The moment it sells chips at high prices, profits are locked in. As for whether you rent them out or leave them empty, whether you make money or lose money, it has nothing to do with it. So, don’t think that because Oracle signed a massive order, it means they have won.

In this industrial chain, the most comfortable and profitable is always the one who grasps the core technology and has pricing power upstream. Oracle is just contributing a thrilling footnote to Nvidia's story of hegemony.

$AI

The US stock AI sector just collectively plummeted, and Oracle has come with ghost stories.

The cause is a recent exclusive report from The Information about Oracle $ORCL, which I recommend everyone to check out the original article.

Below are some key points summarized from the report,

1. Last quarter, leasing NVIDIA Blackwell chips resulted in a loss of nearly $100 million, while leasing NVIDIA chip servers generated approximately $900 million in revenue, with a gross margin of only 14%.

2. The 14% gross margin already accounts for data center labor, electricity, and some equipment depreciation costs, while other unspecified depreciation expenses will further erode 7 percentage points of profit.

3. The average gross margin for this business over the past year was about 16%, and the overall gross margin for the company is expected to decline significantly from nearly 70% in recent years.

4. The top five AI cloud customers contribute about 80% of the business, with nearly all of the $317 billion cloud agreements in the last quarter coming from OpenAI.

5. The business is overly reliant on a few customers, with most data centers being leased, resulting in weak cost control capabilities.

The overall core contradiction is the imbalance between high revenue expectations and the reality of low profit margins, meaning increasing revenue does not lead to increasing profits.

$BTC $AI