Bear Market Garbage Time: Why Is This the Best Starting Point for Growth? The current market is clearly in a correction phase, with Bitcoin having dropped significantly from its all-time highs. Many feel confused, anxious, or even want to give up entirely. But I firmly believe: the low point of a cycle is never the end, but the best starting point for growth. Welcome to "Cycle Survival Philosophy," a light-learning series that doesn’t teach you to catch bottoms or tops, nor does it hype calls or bull markets. Instead, it accompanies you to systematically learn during bear markets and live wisely during bull markets. We don’t chase short-term fluctuations; we build a survival philosophy that can endure both bull and bear markets. Why is the bear market the best "garbage time"? In basketball, when the score difference is large and the outcome is basically decided, the last few minutes are called "garbage time." Players still compete, but the meaning of the game changes—it becomes a time to test willpower, hone fundamentals, and prepare for the next game. The crypto market is the same. During the bull market frenzy, everyone chases hot topics, narratives, and leverage; when the market enters a correction, noise significantly decreases, and price is no longer the sole focus. At this time: 1) Time becomes abundant: You no longer need to watch the market 24/7 and can systematically study. 2) Costs are very low: Small-scale practice in DeFi, on-chain analysis, and wallet management means the cost of trial and error is minimal. 3) Mindset is clearest: When the fear index is high, it’s easiest to see what is truly long-term value and what is just short-term emotion. 4) Cognitive compounding begins to accumulate: The truly skilled use the bear market—when others are sleeping—to read, think, and take action. The bear market is not a punishment but a low-cost window for cognitive upgrading. Miss it, and you might be swayed by emotions again in the next bull market. What is the core of cycle survival philosophy? Simply put, it’s about living better, thinking clearer, and acting more decisively in any cycle. Its three fundamental principles: Don’t waste garbage time — turn the low phase into a period of capability building. Prioritize compounding thinking — the compounding of knowledge, skills, and mindset is far more important than short-term price fluctuations. Balance body and mind — don’t be a "chain slave" who only watches candlesticks, nor a completely detached "outsider." We must understand blockchain and also know how to live.

"The Big Short" goes all in! Semiconductor surge triggers "short" trigger, doubling down on shorting AI computing power leaders "The Big Short" prototype Michael Burry @michaeljburry (Michael Burry) on July 30 (Thursday)—the day the global semiconductor sector surged—reversed to double down on shorting the AI computing power industry chain, while bottom-fishing consumer and gambling stocks. He clearly stated: "All are buys and additions, no sales," and these operations were completed on the same day, not old positions disclosed. 👀 Long positions taken on July 30 (bottom-fishing) • DraftKings (DKNG): increased holdings at about $23.40, considered a "large position" along with Flutter • Flutter Entertainment (FLUT): increased holdings, large position • Zoetis (ZTS): significantly increased holdings at about $76, "full position" • Lululemon Athletica (LULU): significantly increased holdings at about $118, "full position" Bottom-fishing logic: a major opportunity formed after a long-term decline with bottoming/consolidation; chips are transferring to stronger holders; current price risks have been largely released, especially suitable for long-term investors. 👀 Short positions taken on July 30 (adding positions) • Nvidia (NVDA): increased holdings of put options expiring December 18, 2026, with strike prices in the low range of $100–125 • Micron Technology (MU): expanded direct short positions at about $880 (not options, direct short selling) • SOXX (iShares Semiconductor ETF): added direct short positions at about $506 (not options, direct short selling) • QQQ (Nasdaq 100 ETF): increased holdings of put options expiring January 15, 2027, with strike prices in the high range of $500 • Tesla (TSLA): short position unchanged • Palantir Technologies (PLTR): short position unchanged Key distinction: Burry uses long-term put options on Nvidia and QQQ, while directly shorting Micron and SOXX, not all bets are through options. 👀 Burry's three core bearish logics on AI 1) Technical lifespan shorter than accounting lifespan: large cloud providers extend depreciation periods for servers and network equipment, possibly underestimating depreciation expenses by about $176 billion cumulatively from 2026 to 2028, thus overestimating AI investment-related profits 2) Capital expenditure growth outpaces verifiable AI monetization: huge investments but questionable monetization ability 3) Real technological revolution may also create supply bubbles: demand is real, but oversupply investment leads to bubbles 👀 Timeline of Burry's AI short positions evolution • Fall 2025: first public short on AI investment theme, Scion 13F shows holding 1 million Nvidia and 5 million Palantir put options; Palantir actually bought 50,000 put options at $1.84 each, premium cost about $9.2 million • November 2025: raised issue of cloud providers extending depreciation periods, pointing out possible cumulative underestimation of about $176 billion depreciation from 2026 to 2028 • April to May 2026: upgraded from single-point short to industry chain and index-level shorts, added Nvidia, QQQ, SOXX put options, established Oracle put options, and direct short Palantir • End of May 2026: rolled QQQ and SOXX options and increased risk capital • June 30, 2026: established direct Nvidia short at about $198.09, also shorted Applied Materials, SOXX, and included Tesla and Caterpillar in cyclical short portfolio • July 2, 2026: shorted Micron at about $1051.87 • July 17, 2026: increased Nvidia put options again, reduced Oracle position by half after profits became too large • July 30, 2026: added Nvidia, Micron, SOXX, and QQQ shorts amid strong semiconductor sector rebound 👀 Market background data on July 30 • South Korea KOSPI Composite Index: surged a record 18% on Friday (SK Hynix and Samsung Electronics, two major memory chip giants) • Taiwan stock market benchmark index: up 8% (TSMC holds huge market cap weight) • Japan Nikkei 225 Index: up 4% (Kioxia, Tokyo Electron, Advantest, leading AI computing power industry chain players) • US Philadelphia Semiconductor Index: largest gain since April 2025 👌 Core conclusions and risk reminders Burry's doubling down on shorts this time looks more like a mid-term tail risk alert rather than a short-term sell signal for Friday. 1/ His low strike price, long-dated options indicate a bet on nonlinear downward revisions of valuation and earnings expectations over the coming months, not requiring Nvidia, Micron, or SOXX to fall continuously every day 2/ He bets the market overestimates the full-cycle profits, free cash flow, and asset residual value that this round of capital expenditure can ultimately convert into 3/ But far out-of-the-money put options are extremely sensitive to time, volatility, and entry points; even if the final direction is correct, premature positioning may lose all premiums In summary: Burry "pours cold water" on the semiconductor rally day—an oversold rebound in the AI computing power chain does not equal a trend reversal; the market may be overestimating the long-term returns of AI capital expenditure.

Learn not to be attached to fighting. Some things don't need to be won, just escaped from. Solving problems is a skill; avoiding problems is even greater wisdom. Don't prove yourself on bad people or bad situations. Recognize the situation, admit mistakes, accept responsibility, and walk away. Don't complain, don't get entangled, don't waste energy internally. Protect your own energy. Your goal is to cross the swamp, not to deal with every crocodile.

"Overnight Finance" July 31 U.S. stocks surge across the board, Microsoft sets historical record On Thursday, the three major U.S. stock indexes closed sharply higher, with the Dow up 1.19%, the Nasdaq up 2.78%, and the S&P 500 up 1.66%. The Philadelphia Semiconductor Index soared over 8%, with Micron up 18%, SanDisk up 26%, and SK Hynix up 17%. Microsoft's stock price surged 15.5%, adding about $450 billion in market value in a single day, marking the largest single-day market value increase ever for an individual stock. For Chinese concept stocks, the Nasdaq Golden Dragon China Index rose 1.05%. Apple's earnings mixed, after-hours down 4% Apple's Q3 revenue was $109.4 billion and earnings per share $2.02, both exceeding expectations. iPhone sales jumped 21.7% to $54.25 billion, a record high for the third quarter. However, Greater China revenue of $18.8 billion and services revenue of $30.7 billion both missed expectations, leading to a 4% drop in after-hours trading. Amazon's cloud business strong, after-hours up 8% Amazon's Q2 revenue was $200.6 billion, beating expectations. AWS cloud revenue grew 37% year-over-year to $42.2 billion, far exceeding the market's expected 31% growth. However, Q3 revenue guidance of $197-202 billion was below the expected $204 billion, though after-hours trading briefly rose over 8%. Meta invests nearly $700 billion betting on AI infrastructure Meta disclosed in regulatory filings that it has committed nearly $700 billion through contracts for AI data centers and cloud computing. Of this, $349.3 billion are irrevocable contracts, with an additional $347 billion in leasing commitments not yet accounted for. In July alone, $68 billion in new commitments were added. OpenAI sharply cuts model prices OpenAI announced price cuts for two GPT-5.6 models: Luna down 80%, Terra down 20%, to compete fiercely with companies like Anthropic. However, the most powerful version, Sol, remains at the original price but will offer a faster API option. Hedge fund Situational Awareness sells off holdings after huge losses Situational Awareness, a $24 billion hedge fund founded by a former OpenAI researcher, has sold most of its $16 billion public market stock holdings to Citadel due to massive losses. The fund still holds $5 billion in Anthropic private equity and will continue to operate as a private investment company. Anthropic secures $15 billion data center loan, Google guarantees A Morgan Stanley-led syndicate plans to provide a $15 billion loan to Nexus Data Centers, a data center developer partnered with Anthropic, to build a large data center campus and a 1.6 GW natural gas power plant in Texas. Google provides financial guarantees for Anthropic's lease and power payment obligations. Other market dynamics • Most European stocks rose, with the DAX up 0.65%, CAC 40 up 0.92%, FTSE MIB up 1.24%, and the FTSE 100 down slightly 0.11% • The U.S. dollar index fell 1.01% to 99.868, with the euro and pound sterling strengthening against the dollar • Bitcoin rose over 1% to $64,700, Ethereum up 0.6% to $1,920 • Crude oil prices both closed lower, WTI down 1.03% to $83.59, Brent down 1.88% to $89.03 • The 30-year U.S. Treasury yield remains near the 2007 high #财报观察员：亚马逊指引不及预期，股价却反涨9%

Open ROOM now to start playing and mint your SBT! The higher your SBT level, the more $CTC your wallet can potentially earn on the EVM chain, and these $CTC will be directly used to participate in the $MINI launchpad. The snapshot is coming soon, so hurry up and level up. $MINI is the first project on the PenguinSwap launchpad and also the inaugural project on this brand-new launchpad on Creditcoin. This sale uses a Continuous Clearing Auction mechanism, truly achieving on-chain public price discovery. Creditcoin has always focused on connecting blockchain activities to real-world use cases, and Minicoin, as an IP real-world asset network, perfectly fits this direction—allowing creators to truly own and earn from their IP. The ROOM game uses the minini IP officially licensed by IPX to mini labs. IPX is the company behind LINE FRIENDS, owning globally recognized characters like BT21 and minini, with a strong audience and cultural influence in Asia. This provides Minicoin with an identifiable IP foundation, supporting the construction of an on-chain network for creators' own IP. Enter ROOM now through the official portal, mint your SBT, level up, and secure your launchpad qualification: Let's level up together and rush in as soon as the launchpad opens! #MINI #ROOM #PenguinSwap #Creditcoin

Brothers, after today's market close, the earnings reports from Microsoft and Meta have directly torn open the AI narrative. Microsoft's Q4 revenue was $90 billion, up 18% year-over-year; Microsoft Cloud brought in $5.93 billion, up 27%; Azure and other cloud services grew 43%. Azure's annual revenue officially surpassed $100 billion. Commercial Remaining Performance Obligations (RPO) soared to $678 billion, up 84% year-over-year. Copilot paid seats also exceeded 30 million. On the same day, Meta reported revenue of $60.8 billion, up 28%, which looks good in terms of growth. But free cash flow was only $784 million, plunging 91% year-over-year, nearly hitting a multi-year low. The full-year capital expenditure guidance was raised to $130–145 billion (up from a previous floor of $125 billion). The stock price clearly came under pressure after hours. Both are heavily investing in AI, but one has turned computing power into visible cloud revenue and backlog orders, while the other is still heavily building infrastructure, burning through cash. The market's attitude is "wait and see." Where does the difference lie? On the surface, it's a matter of timing, but fundamentally, their business models are vastly different. Microsoft's core logic is "selling shovels + selling applications." Azure directly turns GPUs, electricity, and networks into rentable cloud services, with customers paying based on usage, creating clear recurring revenue. The sharp rise in RPO indicates that not only large model providers are competing for computing power, but a broader range of enterprise customers are locking in future usage. AI here is incremental cloud business, and marginal returns are already visible. Meta's logic is more like "build the shovel first, then hope it digs up more advertising gold mines, while exploring new businesses." It doesn't have a comparable external cloud platform to immediately rent out excess computing power for monetization (although management has started exploring this direction). Money is mainly spent on training larger models, improving recommendations and ad efficiency, and potentially future Agents and enterprise tools. The short-term result is that the ad business is still growing, but capital expenditures almost consume all operating cash flow, and free cash flow has collapsed, naturally making the market nervous. Adding the macro background makes it clearer. By 2026, the entire Big Tech AI capital expenditure will reach historic levels—several mega players combined are guiding easily over $700 billion, with some estimates even higher. Where does the money come from? When operating cash flow is insufficient, companies start borrowing or even considering equity. The cost of capital is no longer the background noise of a zero-interest-rate era. Although the Fed hasn't immediately raised rates, the market is already repricing the possibility of "longer-term higher rates." The tolerance for burning cash has clearly shrunk. So the market's question now has changed: it's no longer "Are you still investing in AI?" but "How long will it take for the money you invest to turn into sustainable cash flow?" Microsoft's current answer is relatively clear—accelerating cloud revenue, soaring backlog orders, and AI applications beginning to reach paid scale. Meta's answer is still "We are confident, but need more time." The valuation tolerance for these two answers is naturally different. Tonight, we'll continue to see who is truly making money. Amazon and Apple are next up. Amazon's key growth and profit margins will directly test the logic of "whether cloud business can sustainably turn AI computing power into profit"; Apple is more conservative, focusing on whether service revenue and AI feature implementation can provide substantial lift, rather than aggressively piling up infrastructure. The AI story has moved from "telling a shared vision" to "starting to show report cards." Business models determine monetization pace, and monetization pace determines whether the market is willing to continue giving high valuations. What do you think? Has Microsoft already pulled ahead, or will Meta's more aggressive internal investment eventually realize a bigger imagination? #财报观察员：微软云收入破千亿，Meta却指引拉胯——AI故事分化了？

The core of Waller's speech this time is just one sentence: he is "depowering" the Federal Reserve and "centralizing power" for himself. In plain terms, he doesn't want everyone to keep focusing on what Fed officials say or what hints are dropped at press conferences. Don't ask anymore "Will there be a rate cut in September?" It's a pointless question—he will tell you to "look at the data," but this data is not the currently released CPI or PCE, but a new working group and new data system he plans to establish in the future. He is reshaping the rules of the game: previously the Fed guided the market, but from now on, the market will be guided by the data he defines. The speech actually has six key points, let me break them down for you: First, the 2% inflation target is constantly mentioned. This leaves room for future policies and also shows that the "Fed is very independent," sounding hawkish but actually laying the groundwork. Second, weakening forward guidance. Not only are other officials speaking less, even his own press conferences will "reduce influence." Future press conferences may become increasingly boring, and non-critical decisions might be ignored altogether. Third, Treasury yields rising? That means the market is tightening on behalf of the Fed. He shifts the blame to the market, implying financial conditions are already tight, so there's no rush. This further weakens the Fed's own guiding role. Fourth, the economy is resilient. The usual reason for no rate cuts. Fifth, no expectations given for September. Reinforcing the stance of "I won't guide you, you guess yourselves." Sixth, three colleagues want to raise rates? That's their opinion; I only look at the data. No decisions before the data comes out—but this "data" is his future new tool. Characterization: Procedural hawkishness. Sounds hawkish (2% target, questioning June CPI, high rates are beneficial), but softer compared to true rate-hike advocates. Consider it a "relatively neutral hawk" or "a hawk with a strategy." The market reaction is also interesting: • During the speech, bonds, gold, and the dollar fluctuated wildly, then repriced after the speech • Short-term bonds (1-year) fell then bounced, showing the market is still afraid of high rates and hikes • Long-term bonds (10-year, 30-year) yields rebounded because he questioned CPI, emphasized the 2% target, and combined with high oil prices, inflation concerns resurfaced • Gold and the dollar rose, then fell, then rebounded, showing market confusion about future rates • Stocks fell after the speech because the press conference gave no sweeteners; the market had to return to fundamentals in earnings reports What to watch in the next two days? • US earnings reports (META, Microsoft, etc.) • Tomorrow's June PCE data Although Waller said "June CPI has little impact on policy," if tomorrow's PCE and CPI are similar, the market will still move in the short term—words say it doesn't matter, but the market (body) is honest. Summary in plain language: Waller is playing a big game. He doesn't want to be a "Powell" type spokesperson whose words are dissected by the market; he wants to be a "data architect"—stop guessing what I say, start guessing how I define the data. Press conferences are becoming "official nonsense time," and real power is shifting to his future working group. In the short term, this is hawkish; in the long term, this is an internal recentralization of the Fed's discourse power. #美联储三票主张加息，今晚PCE成新看点

"Overnight Financial News" 1. Policy Drive: "15th Five-Year Plan" for the Sports Industry Released The General Administration of Sport of China issued the "15th Five-Year Plan for Building a Strong Sports Nation," setting clear goals for 2030: 4 square meters of sports venue area per capita, 40% of the population exercising regularly, and a sports industry scale exceeding 7 trillion yuan. Favorable policies are expected to boost the entire industry chain's prosperity, coupled with Nike's channel adjustments, creating good development opportunities for local brands such as Li Ning, Anta, and Xtep. 2. Global Market: US Stocks Plunge, Crude Oil and Gold Surge Overnight, the three major US stock indices all fell more than 1.5%, with the Dow Jones down 2.19%. Large tech stocks broadly declined, with the optical communication and storage sectors sharply down (Micron Technology fell nearly 10%). Chinese concept stocks bucked the trend, with the Nasdaq Golden Dragon China Index rising 1.73%, and New Oriental soaring over 15%. WTI crude oil surged 6.74% to $84.6 per barrel, and gold rose 0.66% to $4065.5 per ounce. 3. Macroeconomic and Industry Highlights • Federal Reserve divisions widen: For the fifth consecutive time, interest rates were held steady at 3.5%-3.75%, but three officials voted for a rate hike, signaling a rise in hawkish sentiment. • Moonlight (Kimi) closed its Series F financing early due to oversubscription by several times, raising over $3.5 billion and reaching a valuation of $35 billion. • Nine departments jointly issued a document to strengthen data development and utilization in the tech-finance sector, releasing Data Catalog 1.0. • BYD passed the new national standard certification for L2 assisted driving; Lead Intelligent's foldable screen 3D printing equipment achieved mass delivery. • Raybio Biotech expects revenue growth of 200%-320% in the first half of the year; Siwei Technology issued a profit alert due to booming AI data center demand; Fosun International expects net profit of 1.5-1.8 billion yuan in the first half; New Oriental's annual net profit increased 27.8% to $475 million. 4. Stock Highlight: Anker Innovations Demand for energy storage business (portable, balcony solar storage, residential energy storage) continues to grow. Starting July 27, it officially entered the Hong Kong Stock Connect, expected to expand its investor base and improve H-share liquidity. Industrial Securities believes its multi-category synergistic development and platform value will continue to be unleashed. #美联储三票主张加息，今晚PCE成新看点

"Overnight Financial Report" 1. Overall Market Overview On July 27, the US AI hardware sector experienced a significant decline. The Philadelphia Semiconductor Index closed down 2.2%, with SanDisk plummeting 11% and SK Hynix falling 7.5%. The sell-off sentiment spread to Asian markets, with the South Korean Composite Index triggering a circuit breaker after dropping more than 8% intraday and closing down 10.8% for the day. The Nikkei 225 Index also closed down 4%. The essence of this round of decline is a global reassessment of the credit and supply landscape in the technology sector. The real demand for AI has not substantially weakened; funds are shifting toward Hong Kong stocks for risk aversion and portfolio reallocation. 2. Underlying Logic Behind the Decline in US AI Hardware and Rising CDS of Overseas Cloud Providers 1. Significant upward revision of capital expenditure expectations fuels credit anxiety Google raised its 2026 capital expenditure guidance and increased AI infrastructure investment for 2027, prompting the market to raise overall spending forecasts for the four major cloud service providers to $731.9 billion and $950.2 billion for 2026 and 2027, respectively. The market realizes that the AI investment cycle and scale exceed earlier estimates, raising concerns about increased leverage among tech companies, cash flow pressure, and monetization efficiency lagging behind spending growth. CDS spreads for major cloud providers continue to widen. 2. Nvidia's closed-loop financing model amplifies market risk aversion Nvidia offers financing and guarantees to downstream customers, encouraging them to use funds to purchase its chips, creating a tightly bound credit closed loop. Coupled with Nvidia's $500 billion cooperation with SK Hynix and negotiations with OpenAI for up to $250 billion in financing guarantees, the market fears leverage accumulation within the chain, where single-point risks could trigger a chain reaction of credit contagion. 3. Fundamentals have not materially deteriorated Mainstream AI chip leasing prices remain stable, and real demand for computing power is resilient. Overseas cloud providers can raise funds through equity, various bonds, and government support, ensuring sufficient financing supply. Currently, credit risk remains at the market expectation level and will not constrain AI industry development in the short term. 3. Dual Core Reasons for the South Korean Stock Market Crash First, two key breakthroughs in the domestic memory industry: ChangXin Memory Technologies listed on the A-share market, using fundraising to open large-scale expansion channels, unlocking import substitution space for domestic memory; domestic immersion DUV lithography machines have been commercialized, breaking equipment constraints for domestic memory companies. The original logic of tight memory supply is shaken, and funds have pre-priced the impact of increased domestic capacity on South Korean memory giants. Second, the valuation premium formed by overseas tech companies relying on technology blockades and oligopolistic monopolies is entering a correction phase as the domestic supply chain continues autonomous breakthroughs. The valuation system of the Korean stock market's memory-weighted sector is being systematically repriced, triggering a sharp market sell-off. 4. Hong Kong Stocks Become the Main Global Capital Sink in This Round Global AI sector trading was crowded earlier, with concentrated profit-taking and capital needing to shift to undervalued assets. Hong Kong stocks have multiple advantages: the index has deeply declined previously, with valuations long below historical averages, offering outstanding cost performance; listed companies have improved shareholder returns and digested unlocking selling pressure, combined with RMB stabilization boosting southbound capital allocation willingness, creating a positive capital flow cycle; short positions and short-selling volumes remain high, and short covering will continue to bring incremental buying. Overall, the trend of global capital returning to Hong Kong stocks is sustainable. 5. Risks to Watch in This Market Escalation of global geopolitical conflicts, deterioration of China-US relations, central banks tightening monetary policies beyond expectations, slower-than-expected AI commercialization progress, and further abnormal widening of credit spreads for US tech giants. #停火48小时告吹，美伊边打边谈