We need to avoid "Grandpa and Grandma coins": the fate of these coins is often to cut retail investors like chives and go to zero In the cryptocurrency circle, there is a type of coin that is always the most lively and also the most dangerous—"Grandpa and Grandma coins." They have such a low threshold that almost no barrier is needed, and their narrative is so simple that even grandpas and grandmas at the market can understand it. Their communities often reach tens of millions, and when sentiment surges, the price can skyrocket. $CORE and $PI are typical examples. But remember one thing: what you really should do is to stay far away from them. Because the ultimate fate of these coins, in the vast majority of cases, is only one thing—to repeatedly cut waves of retail investors like chives, then the price goes to zero or approaches zero infinitely. History has repeatedly proven this. What are "Grandpa and Grandma coins"? "Grandpa and Grandma coins" are not an official classification but a precise joke within the community about a type of coin. They usually have these characteristics: Extremely simple participation: mining by clicking on a phone, Telegram mini-games, daily check-ins to "earn coins." Narrative extremely accessible: no complex technology, only talking about "everyone can participate," "community wealth sharing," "changing the world in the future." User structure highly retail: a large number of ordinary people and middle-aged and elderly investors who are new to cryptocurrency. Market driven entirely by sentiment: FOMO causes a surge, sentiment fades and it crashes, fundamentals can almost be ignored. $CORE (Core DAO) attracted a lot of attention with its simple packaging of "Bitcoin security + smart contracts"; PI (Pi Network) accumulated a huge "pioneer" community through years of phone-click mining. After launching on OKX, they instantly became the focus of retail discussion. But this is precisely the beginning of danger. Representatives of similar "Grandpa and Grandma coins" on OKX Currently, on OKX, coins with spot trading and obvious "Grandpa and Grandma" attributes include (dynamic and for warning reference only): Tap-to-Earn / Click mining type PI (Pi Network) $NOT (Notcoin) HMSTR (Hamster Kombat) CATI (Catizen) Classic Meme and sentiment-driven type DOGE, SHIB TRUMP (Official Trump) PEPE, BONK, FLOKI, MEW, BOME, TURBO, etc. Other accessible narrative types CORE (Core DAO) PEOPLE and other historical community coins These coins frequently appear on OKX's hot lists and Meme sections, with trading pairs mainly in USDT. The platform's low-threshold trading experience actually accelerates the influx of retail investors. Why do most of them eventually lead to "cutting retail investors like chives and going to zero"? No real value capture: the vast majority live on sentiment and expectations; once the mainnet launches, unlocking dumps, or the narrative fades, support instantly disappears. User structure determines fate: after many "Grandpa and Grandma" style retail investors buy at high prices, liquidity dries up, and prices can only go down. Project teams and early holders' harvesting logic: accumulate chips at low cost → create FOMO on exchanges → sell at high prices → retail investors take over. This is a script verified countless times. Historical data does not lie: from early "everyone mines" coins to recent Telegram mini-game coins, the vast majority have retraced 80%-99% from their highs within months to a year after listing, even going to zero. The sharp fluctuations after PI's launch and the rapid decline of various click-game coins are vivid examples. Although CORE has some technical narrative, its performance after retail sentiment fades is also not optimistic. Final words The crypto market never lacks stories, but it lacks sobriety. "Grandpa and Grandma coins" are best at using the simplest stories to deceive the most naive hopes. They can bring short-term carnival, but after the carnival, what remains is often a mess and zeroed K-lines. Those who truly survive long-term and make money rarely bet on these coins. Instead of chasing these "everyone can earn" illusions on OKX, it's better to spend time and funds on assets with real technical barriers and sustainable value. Remember: avoiding Grandpa and Grandma coins is not missing opportunities but avoiding traps. The fate of these coins has never been shared wealth but cutting retail investors like chives and going to zero.

SK Hynix to Launch Buyback Soon SK Hynix officially announced on August 19, 2026, that its board has approved a share repurchase and cancellation plan totaling approximately 40 trillion KRW (about $28.6 billion), setting a record for the largest scale in the history of Korean listed companies. Based on the previous day's closing price, about 24.07 million shares will be repurchased, accounting for approximately 3.3% of the total shares outstanding. The buyback period will last about three months starting August 20, with all repurchased shares to be canceled upon completion. The company also raised its cumulative free cash flow shareholder return target for 2025-2027 to "over 50%" and stated it will simultaneously advance dividend increases. This move is based on the company's view that the current stock price significantly undervalues its AI memory business competitiveness and strong cash generation capability, with net cash reaching about 69 trillion KRW by the end of Q2. This accelerates the execution of existing shareholder return policies and sends a clear signal of value revaluation to the market. SanDisk and MU Follow Suit Following the news, US-based memory stocks like Micron Technology (MU) and SanDisk (SNDK) quickly showed correlated reactions. The storage sector is highly homogeneous, and the structural shortage of HBM, DRAM, and NAND driven by AI data centers is a common factor. Hynix, as the global leader in HBM, is seen as confirming the industry's entry into a "high profit + high return" phase with its buyback. Recently, Micron and SanDisk have strengthened continuously due to AI demand, long-term contract locking, and their own return plans. Hynix's buyback further reinforces sector sentiment resonance, with funds short-term inclined to map the Korean giant's shareholder-friendly policies onto their US peers, driving both to rise. This linkage is both emotional contagion and a collective bet that the storage supercycle is not over. Will Good News Turn Bad After Being Fully Priced In? The concern that "good news fully priced in equals bad news" is not unfounded, especially given the stock's significant cumulative rise and increased macro interest rate and geopolitical volatility, where some short-term funds may choose to take profits. But judging solely by "fully priced in" is inaccurate. The buyback is not a one-time event but a sustained buying support over the next three months, plus cancellation directly reduces share capital and improves per-share metrics, so there are still positive mid-term effects. More importantly, the tight AI memory supply-demand pattern remains unchanged, and the performance and cash flow of Hynix, Micron, and SanDisk still heavily depend on this fundamental. If subsequent quarterly guidance remains strong and prices stay high, the buyback is more like "icing on the cake" rather than the "final blow." The real risk lies in demand slowing or capacity expansion exceeding expectations, not the buyback itself. Watch Tomorrow's Opening for Results Tomorrow (August 20), the opening of US and Korean markets will be a key window to test the market's true attitude. Investors need to closely watch whether Hynix's stock price can hold steady after the buyback officially starts, whether Micron and SanDisk's follow-up gains continue, and the overall semiconductor sector's trading volume and capital flow. A spike followed by a pullback after the open is normal emotional digestion; if it maintains relative strength supported by buying, it indicates capital's recognition of the "accelerated shareholder returns + AI demand" dual drivers. Short-term volatility is inevitable, but mid-term still needs to return to supply-demand and performance verification. The storage sector holds both opportunities and risks; rationally assessing the opening results is more important than blindly chasing gains or panicking sell-offs. #闪迪回落逾9%，存储估值分歧加剧 $SNDK $MU $SKHYNIX

The current memory chip industry is in an AI-driven super cycle, not on the brink of death. SK hynix, as the absolute leader in HBM, has indeed been the first to realize the most extreme profit explosion, but its uniqueness does not mean the entire memory sector has reached its end. On the contrary, the supply-demand gap for general DRAM, enterprise SSDs, and related NAND products continues to widen, with a clear upward price trend and a long expansion cycle, making a bubble burst unlikely in the short term. More importantly, related U.S. memory technology companies have not yet launched large-scale "bloodsucking" IPOs, and capital exit pressure has not yet been concentratedly released, further delaying the arrival of the cycle peak. Looking at the fundamentals first. In the first half of 2026, the global memory market remains tight, driven by AI server demand. DRAM bit demand growth is expected to reach about 25%, NAND close to 20%, while new capacity is constrained by the complexity of advanced processes, EUV equipment, and new factory construction cycles, making significant short-term release difficult. SK hynix, leveraging its leading position in HBM3E/HBM4, achieved revenue exceeding 130 trillion KRW in the first half, with record net profits and a gross margin above 80%, with significant contributions from customers like Nvidia. But its core logic is "high value-added products first," and prices for ordinary server DRAM and enterprise SSDs are also rising sharply, with peers like Samsung and Micron also showing high growth. Inventory levels are generally at historic lows (2-4 weeks), far below levels before the downturn cycle. Long-term supply agreements (LTA) coverage has increased to 50%-70%, locking in demand visibility for the coming years. All these indicate that this is not a traditional cyclical short-term speculation but a structural shortage. SK hynix is an "exception" because it was the earliest and most deeply tied to the AI computing power chain. Its HBM market share has long been maintained above 55%, with deep collaboration with Nvidia, strong pricing power, and profit margins far exceeding peers. However, it is notably specialized—its NAND share is relatively weak and highly dependent on a few large customers. Once HBM supply and demand gradually balance, or Chinese manufacturers (ChangXin Memory, Yangtze Memory) accelerate substitution in general DRAM and mature NAND, SK hynix's excess profits may converge. However, this does not drag down the entire memory industry: demand for high-capacity DRAM for servers and QLC enterprise SSDs is still exploding, and storage content in automotive and edge AI devices is also increasing. 2027 has been warned by many as the "most severe memory shortage in history," with real capacity ramp-up expected in the second half of 2027 to 2028. The supply-demand mismatch will continue for at least 1-2 years. The "bloodsucking" from the capital side has not fully started, which is an important buffer against a bubble burst. SK hynix itself listed on Nasdaq in July 2026 in the form of ADRs, with a record fundraising scale, but this was more about valuation re-rating and opening the channel to U.S. investors rather than large-scale sell-offs. The real potential "bloodsucking" pressure comes from U.S. related entities not yet listed. For example, SK hynix's U.S. NAND subsidiary Solidigm (formerly Intel's NAND business) has started Pre-IPO financing with a target valuation exceeding $35 billion and is actively preparing for Nasdaq listing. Once the official IPO releases circulating shares, it may trigger phased profit-taking and valuation volatility. Another potential target is other U.S. memory-related tech companies (such as those focused on enterprise SSDs or new storage solutions), which remain private and have not yet undergone large-scale capital extraction through public markets. The pace of their IPOs will determine the rhythm of capital exit. Until they complete IPOs and fully digest valuations, the industry overall still tends to see capital inflows rather than outflows, lacking the fuse for a bubble burst. Historically, memory cycle peaks are often accompanied by concentrated capacity releases, inventory accumulation, and capital frenzied cashing out. The current situation is completely different: cautious capacity expansion (prioritizing HBM and high value-added products), customers locked into long-term agreements, and although domestic Chinese substitution is accelerating, it is difficult to fill the high-end gap in the short term. From a valuation perspective, even though SK hynix, Micron, and others have risen sharply, forward P/E ratios remain relatively controllable after profit explosions, and the market is trading more on "shortage sustainability" rather than pure bubbles. ChangXin Memory's listing on the STAR Market caused a brief fluctuation but did not reverse the global supply-demand tightness. Of course, risks always exist. If AI capital expenditure slows significantly, geopolitical shocks disrupt supply chains, or new capacity is unexpectedly released early, the cycle may turn earlier. But based on current data, demand in 2026-2027 will still exceed supply, and price levels are expected to remain high. The memory industry is far from dead; SK hynix's outstanding performance is just a leading indicator, not a terminal signal. U.S. related tech companies have not completed their "bloodsucking" IPOs, meaning the capital feast still has chapters to come, and the bubble burst countdown is far from starting. Investors need to focus on supply-demand data, long-term agreement progress, and new capacity timelines, rather than simply watching short-term valuation fluctuations. This AI-redefined memory super cycle can still continue for a while. #闪迪收涨逾8%，长期协议受关注 #高盛称美联储9月加息可能性非常低

Recently, the Strait of Hormuz has once again fallen into a state of effective closure. This US-Iran conflict, which erupted at the end of February 2026, has lasted for over 170 days. Although there were brief memorandums of understanding and limited navigation attempts in between, Iran has clearly stated that the strait will not truly return to normal commercial navigation until the US meets a series of conditions including lifting the maritime blockade, removing sanctions, and unfreezing assets. About one-fifth of global oil trade originally depended on this chokepoint waterway, but now vessel traffic has dropped to single-digit percentages of pre-war levels. War risk insurance premiums have soared to 30 times the usual rate, and Brent crude oil prices have climbed back above $90. The sharp rise in energy costs has directly pushed up global inflation expectations, while the US Treasury's massive debt issuance has compounded this, causing US Treasury yields to rise across the board. The 30-year Treasury yield once touched its highest level since 2007, and the 10-year yield is also approaching multi-year peaks. The bond market sell-off is transmitting to the stock market, putting risk assets under repricing pressure. Against this macro backdrop, the probability of a sharp plunge in US stocks after tonight's opening has significantly increased. Historical experience shows that when energy shocks and rising interest rates occur simultaneously, growth stocks and high-valuation tech stocks often bear the brunt first. The semiconductor sector, as the core beneficiary of this AI rally, has already accumulated huge gains, and its valuation elasticity has correspondingly amplified downside risks. Especially those memory chip manufacturers highly dependent on global supply chains and terminal demand prosperity are more vulnerable to capital withdrawal when risk appetite sharply declines. Once the market enters a risk-off mode, funds tend to prioritize selling liquid and previously high-gain targets, creating a stampede effect. Based on the above logical chain, I recommend focusing on shorting SK Hynix opportunities. As a global leader in HBM high-bandwidth memory, Hynix's stock price has experienced multiple-fold increases amid the AI server demand boom, with its market value once surpassing Samsung to become Korea's largest. However, the current high oil prices may push up data center operating costs, while the high-interest-rate environment will suppress corporate capital expenditure willingness, potentially slowing AI infrastructure expansion. Coupled with weakening overall US stock sentiment, Hynix's US ADR and related derivatives are likely to become concentrated targets for shorts. Whether through direct shorting, using inverse ETFs, or leveraging futures and options tools, a relatively favorable window seems to have emerged timing-wise. Of course, short-term volatility is intense, so strict position management and stop-loss discipline must be observed. It is especially important to emphasize that geopolitical situations can dramatically turn at any time. Once the US and Iran reach a substantive agreement again and truly restore strait navigation, oil prices may fall and risk appetite recover, quickly reversing the current logic. Therefore, shorting operations are more suitable as tactical trades rather than long-term strategic holdings. Meanwhile, the semiconductor industry itself still has strong long-term fundamental support, and AI demand will not disappear overnight. Investment decisions must be combined with one's own risk tolerance; blind following should be avoided. The market is always full of uncertainty, and tonight's plunge expectation may also be interrupted by unexpected positive news. Staying calm and thinking independently is key to navigating cycles. Finally, a reminder: the above analysis only represents my personal observation and deduction of the current macro and market environment and does not constitute any investment advice. Financial markets carry very high risks, past performance does not represent future results, please make decisions cautiously based on your own situation and consult professional advisors if necessary. Brothers, the market changes in an instant; may we all protect our principal amid volatility and seize the opportunities that truly belong to us. $SNDK $SKHYNIX #30年期美债收益率创2007年以来新高

This wave of $KAITO is pretty much what I expected, it has bottomed out Today the market sentiment is good, all coins and stocks are rising. Actually, you can still enter a small position now, with a stop loss set at 0.4 Or continue to buy at 0.4, the positive news has already been fully released, the Kaito project team has been selling off continuously Empty empty empty, living in the palace #BTC成交萎缩，ETF买盘能否回暖

#闪迪长期协议成焦点，开盘表现待验证 Friends who follow the trades are truly "fortunate." Look at this set of data: the 7-day return rate has been slashed by more than 70%, the trader himself lost over 70,000 U, and followers collectively suffered huge losses totaling over a million, even nearly 1.6 million U. Usually, profits are small and trivial, but when losses occur, there is an unlimited deep drawdown, with the maximum drawdown steadily stuck around 70%, and a pitifully low win rate of just over 16%. What’s even more ironic is that despite this performance, the platform still collects a 20% profit share as usual. They boastfully show off orders and recruit people in private groups, while publicly continuing to exploit followers—typical dual-line harvesting. Such sustained massive losses at this level make it hard not to suspect a dual-account hedging operation behind the scenes: one account deliberately dumps the market, creating high drawdowns to attract followers, while the other account profits in the opposite direction simultaneously, not only securing the spread for itself but also earning an additional 20% profit share. The real risk is entirely borne by the followers, while the trader profits regardless of market conditions. This "sure-win, no-loss" harvesting model turns following trades into a one-way blood transfusion. The data is clear, and anyone with eyes can see how deep the water is. $SNDK #闪迪财报双超预期，新增140亿美元回购授权

Tribute to the legendary Air Force King, initially said the stop loss was 1700, but now says the stop loss is 1800, probably going to be directly blown up #闪迪长期协议成焦点，开盘表现待验证 $SNDK

5x small snacks a little bit, still on the way to break even $KAITO is expected to pump a bit

KAITO is currently around $0.337–0.340 (24h down about 3.8–6.5%, range roughly $0.335–0.367), ARX (Arcium) around $0.107–0.110 (24h down about 8%, range roughly $0.105–0.117). Trading Data Update ARX: The stop loss at **0.1120 has been triggered. The current price has fallen below this level and continues to weaken (24h low touched about 0.105), consistent with the recent ongoing downtrend. Positions have been exited according to the stop loss; further observation is needed to see if there will be a rebound pullback or further decline before making new plans. KAITO: Entry price **0.3860, stop loss set at 0.3560. The current price is significantly below the entry price (floating loss about 12%+), and is near or has reached the stop loss zone. The 24h high is about 0.367, still not far above the stop loss, requiring close monitoring; if it falls below 0.3560, the stop loss will be triggered. Overall, there is a clear pullback from a higher position (recent weekly-level decline is significant), short-term is weak. Brief Market Observation KAITO (AI InfoFi / Base ecosystem): Market cap about $81–82M, 24h trading volume relatively high (around $50M level), circulating supply about 241 million. Price has been oscillating downward after a significant correction from previous highs, RSI and other indicators show oversold signs, but short-term moving averages are bearish; resistance is around 0.36–0.37, support near 0.33. Overall sentiment is cautious, waiting for volume increase and stabilization signals. ARX (Arcium, privacy/confidential computing related): Market cap about $22–23M, 24h volume about $3–4.5M, circulating supply about 209 million. Recently continuously declining, just hit a near low point, rebound is weak. Technicals are weak; short-term focus is whether it can hold above 0.11, otherwise it may continue to seek a bottom. Project narrative (Solana ecosystem confidential computing, etc.) remains, but price performance lags. Overall market volatility remains high, both of these assets have a bearish structure recently. Your ARX position has been stopped out as planned, KAITO continues to strictly observe the stop loss at 0.3560, prioritizing risk control. If there are new positions or need for more detailed levels (such as 4h/1h structure, volume), feel free to mention anytime, continuing to follow the previous style. Pay attention to position management and stop loss discipline. $ARX $KAITO

$ARX $KAITO Kaito (KAITO) and ARX (Arcium), these InfoFi/AI+privacy computing narrative coins, have had a clear pattern in the past: after official positive news or listing events create FOMO, there is a rapid price surge, followed by unlocking and profit-taking leading to a gradual decline. Currently, both have significantly corrected from their highs (KAITO has dropped over 60% to around $0.35, ARX has fallen more than 75% from its ATH to about $0.10-0.11), sitting at relatively low levels. After volume shrinks, technical rebounds are likely. The core driving logic for possible gains today: Technical recovery after overselling + sentiment return After continuous decline, prices are near short-term support (KAITO around $0.35, ARX near $0.10), with RSI and other indicators entering oversold zones. During a relatively quiet weekend in terms of capital flow, even a small buy-in or a rise in social heat can trigger short-covering and short-term bottom-fishing, leading to a quick rebound. Historically, similar projects have seen 15%-30% technical rebounds within 1-2 weeks after deep corrections following positive news. Weekend residual heat of AI/privacy narratives + community expectations Kaito’s previous data collaboration with X and the Katalyst rewards platform remain hot topics, with the creator reward mechanism returning to stakers continuing to attract attention; ARX’s mainnet computing volume data (already exceeding one million confidential computations) and its positioning in Solana’s privacy/AI compute ecosystem also prompt some funds to revisit “real adoption” discussions over the weekend. If social heat bursts during the weekend, it can quickly push low-position chips higher. Pre-unlock game window KAITO has a large proportion unlocking next week (around August 20), but the common altcoin script is “pump before unlock”—major holders or early chips may use this opportunity to push prices up to create liquidity for subsequent selling. Short-term funds will position in advance, forming today’s upward momentum. ARX’s circulating supply remains low; any mainnet progress or partnership rumors could also amplify volatility. Overall altcoin sentiment and capital rotation When BTC is stable and the altcoin sector shows localized recovery, the AI/InfoFi/privacy track, as a previously strong narrative, tends to absorb returning funds first. Its high volatility and low market cap characteristics mean that once volume expands, gains can be more exaggerated than the broader market. Summary: If there are gains today, it is most likely a combined effect of “oversold recovery + narrative residual heat + pre-unlock game,” rather than entirely new major positive news. These coins pump fast and pull back fast, suitable for short-term speculation but strict stop-loss is essential—historical patterns show that sustained rises still require new substantial adoption data or larger-scale positive catalysts; otherwise, the rebound may still be followed by a continued decline. Real-time market conditions, funding rates, and on-chain data are the ultimate judges. #ETF买盘反转，BTC杠杆仓位回升

I hope the United States will not hinder the progress of humanity, but instead work together to bring the benefits of super AI to all mankind sooner. Recently, Reuters revealed that the U.S. plans to send letters to about 35 countries, demanding a "choose one between two" stance in the China-U.S. AI camps: those joining the China-led World Artificial Intelligence Cooperation Organization (WAICO, currently about 29 countries) will be excluded from the U.S. "Pax Silica" alliance (about 24 countries). This is a typical move to turn technological cooperation into geopolitical alignment. But the reality is: China's leading AI models have caught up to the "range" of the top U.S. level, with performance gaps shrinking from generational to marginal; meanwhile, their prices are only 1/57 to 1/2 of the U.S. side. The market votes with its feet—around June this year, the global usage volume of Chinese models surpassed that of the U.S., occupying about two-thirds of the share, completing a reversal within a year. Open source, inclusiveness, and cost-effectiveness crush competition, enabling more developers and countries to truly afford and effectively use AI. Human progress should not be torn apart by camps. AI is not a zero-sum game. Once superintelligence (AGI and beyond) arrives, it will reshape all fields including healthcare, energy, scientific research, and education, shortening the timeline to solve cancer, climate change, poverty, and other problems. Division will only repeat supply chain barriers, standard conflicts, and waste of talent and computing power, making the wait for super AI longer and more costly for all humanity. True "Pax Silica" should not be an exclusive wall but an open and collaborative bridge. As a long-time leader, if the U.S. can let go of the "choose one" obsession and jointly explore security, ethics, and open-source standards with China and other countries, it could accelerate breakthroughs and share the benefits. Competition can drive innovation, but cooperation can amplify the dividends—just like scientific communities in history, which often made the greatest breakthroughs by crossing boundaries. Don't let political fear block humanity's common path forward. I hope the U.S. will rethink: rather than building walls, let's light up the lamp of super AI earlier and brighter together. All humanity is waiting for that day. #AI基建融资升温，英伟达英特尔路径分化