"Computer Repaired in Sapporo, BTC's August First Secures Insurance" Saturday, August 1, 2026 Q3 · Issue 91 Aspirin · Cyclical Analysis from a Data Scientist's Perspective No updates these days. Currently residing in Japan, my computer coincidentally got 100ml of tea spilled on it causing a black screen. After struggling, I took it to the official Apple store in Sapporo. Fortunately, it was repaired and all data remained intact. After rebooting, BTC had already slid back from above 64K to around 63K. The machine is restored, but this July rebound is the time to check the components. In July, BTC opened at $58,628, reached a monthly high of $66,955, and closed at $62,891, an increase of about 7.3%. The pullback at the end of the month cut much of the intraday gains that once exceeded 10%. The candlestick is still bullish, but the closing quality is not impressive. Reviewing the mid-cycle years of 2014, 2018, and 2022 again: they did not replicate the same candlestick pattern. Their commonality is deleveraging in the first half of the year, followed by a decent summer recovery; if the rebound fails to reclaim the long-term trend, a re-pricing occurs in the second half. History does not place bets for today but changes the odds. Chasing shorts in July was painful; entering August and mistaking the rebound for a one-sided bull market is equally risky. Chart: BTC Trading Map for August (as of August 1, 2026) 1. In the three mid-cycle years, summer looks like "improvement" The "mid-cycle years" I refer to are the adjustment years after the previous bull market peak in the four-year cycle, coinciding with the US midterm election years 2014, 2018, and 2022. Their declines, news, and bottoming months differ, but the market sequence is very similar. 2014 was an early-maturing cycle. BTC was around $800 in January, dropped to about $351 in April; from May to June, the price rose from around $423 to $674, closing June at about $640. It was easy to think the big drop was over, but July closed at $586, August bottomed at $444, and the fourth quarter touched $289. The capitulation did not happen within the year but dragged into January 2015, closing at about $178 on January 14. 2018’s path was the neatest. The June low was $5,826, July’s high rebounded to $8,424, nearly a 45% increase from the low; August returned to $5,971. September to October ranged between $6,200 and $7,000 with decreasing volatility until November’s support broke, bottoming at $3,585, and December saw $3,191 again. The summer rebound was significant but did not change the bear market structure. 2022 resembled a fast-forward triggered by events. June bottomed at $17,709 under the shocks of Terra, Celsius, and Three Arrows Capital; July rebounded to $24,573, August briefly touched $25,135, about 42% above June’s low. Then prices fell back near $20,000, and November’s FTX collapse pushed the low to $15,599. Looking at the three cycles together, summer rebounds repair oversold conditions and sentiment but do not necessarily restore the long-term trend. Bottoming times are not uniform: December 2018, November 2022, and January next year for 2014. Therefore, I do not infer a guaranteed drop on any day from the "mid-cycle year" but treat Q4 to January as a window to hold cash and hedge. If 2026 can reclaim 64.8K, hold above 69.5K, and long-term yields fall, it will have left this old path. 2. Two yields are more useful than a whole page of statements The Fed maintained the 3.50% to 3.75% rate range with a 9-3 vote, three members requested a 25 basis point hike. No need to repeat the press conference; I focus on two numbers: on July 31, the US 10-year Treasury yield rose to 4.75%, the 30-year to 5.27%, and TLT fell to $82.25 the same day. The central bank did not move, but the bond market has already raised the cost of long-term capital. Continued rise in long-term yields will compress tech stock valuations and BTC risk appetite simultaneously. My observation switch is: if the 10-year stays above 4.75%, the 30-year above 5.25%, and BTC fails to reclaim the 200-week moving average, the rebound is treated as counter-trend. If yields fall below 4.60% and 5.15%, and BTC recovers 64.8K, the summer rebound can continue to expand. 3. 63.6K is the first gate in August BTC is currently about 63.1K, the 50-day moving average is at 63.4K, and the 200-week moving average is around 63.6K. These three prices cluster together, making 63.4K to 63.8K a short-term watershed. Weekend liquidity is thin; a sudden break below might be a fake move; if a four-hour rebound is still suppressed, it means something entirely different. Looking up, 64.8K is near the 50-day exponential moving average recovery line, 67.2K is the first realization zone before July’s high, and 69.1K to 69.5K is the 20/21-week resistance band. Looking down, 62.2K is the structural confirmation level; if lost, then 60.5K and 58.2K follow; deeper support remains at 54K to 55K. Computers can be sent to official stores, but positions have no repair counters. Stop-loss, hedging, and replenishment prices are best written before the "water spill". Data notes: BTC prices and moving averages are from OKX as of noon August 1, 2026; historical cycle monthly highs and lows reference CoinMarketCap and StatMuse; US Treasury yields are from the US Treasury’s July 31 closing data; Fed rate decision is from the July 29 meeting results. The above is personal research and trading plans only and does not constitute investment advice.

"The Seven Giants Evaporated $797 Billion in One Day, Why Does BTC Still Hold at 64K?" Sunday, July 26, 2026 Q3 · Issue 90 Aspirin · Cycle Analysis from a Data Scientist's Perspective On Thursday, the US stock market laid a very expensive bill on the table: the seven major tech giants lost about $797 billion in a single day, with the Nasdaq down 2.2%. Google's cloud revenue grew 82%, yet the market chose to sell; Tesla set a new delivery record, but its stock plunged 14.5%. Investors began to ask the same question: how long will it take for the hundreds of billions invested in AI to steadily convert into free cash flow? Friday's recovery was uneven. GOOGL only rebounded 0.65%, TSLA fell another 2.1%, QQQ dropped 1.1% again; BTC remained around $64,100 by the weekend. Over the past month, BTC, considered a "high beta tech stock," did not replicate the Nasdaq's decline this time. Chart: July cross-asset returns, correlations, and BTC key price levels (as of July 24/25, 2026) 1. A Set of Easily Overlooked Divergence Data I compared from July 1 to the latest closing prices on the same basis: BTC rose about 6.9%, SPY fell 0.9%, QQQ fell 5.7%, GOOGL fell 11.5%, TSLA fell 26.4%. Roughly calculating the 30-day correlation coefficient of daily returns on common trading days, BTC and QQQ have about 0.34, BTC and SPY about 0.41. This reading does not indicate decoupling but shows a significant loosening from the high synchronization at the start of the year. The current price divergence has a clear source. US stocks are re-evaluating AI capital expenditures. Google raised its 2026 capex guidance to $195 billion to $205 billion, with free cash flow turning negative for the first time in Q2; Tesla's operating profit dropped 57% year-over-year, with AI, robotics, and production line investments continuing to squeeze margins. Stockholders discount every future return to today. BTC has no income statement, no depreciation, and no capital expenditures. It more directly trades dollar liquidity, real interest rates, crypto market leverage, and cycle positioning. Therefore, when the decline stays within the "AI investment return rate" industry logic, BTC can temporarily follow its own rhythm. This divergence is not solid. Brent crude has already risen above $100, and the Fed will hold a meeting on July 28-29; the market even prices in about a 36% chance of a rate hike. Rising oil prices push inflation expectations and bond yields higher, escalating tech stocks' capex anxiety into a market-wide discount rate shock. At that point, BTC and the Nasdaq will likely move together again. 2. BTC Trapped Between 63.5K and 69.5K BTC's current position is straightforward: below is the 200-week moving average at about $63,500, above is the 20/21-week resistance band around $69,300 to $69,700. The price is squeezed back and forth between these two lines, with longs chased near 64K and shorts near 65K, neither with a favorable risk-reward. History also reminds me not to call the July rebound a reversal. In 2018 and 2022, clear recoveries appeared in July, only to give back gains in August and September. This current rebound has already entered the end of the month, with the next week overlapping heavy earnings reports from FOMC, Microsoft, Meta, Amazon, and others, opening a volatility window. 3. My Two-Way Order Sheet BTC Long Conditions If a four-hour candle shows a long lower shadow between $63,500 and $64,000 and closes back above $64,200, one-third of the planned position can be placed between $64,000 and $64,300; stop loss at $62,700, first target at $66,800, second target between $69,300 and $69,700. After reaching the first target, move the stop loss near cost. BTC Short Conditions If the daily close is below $63,300 and then fails to rebound above $64,000, a small hedge short position can be built between $63,600 and $63,900; stop loss at $65,200, targets sequentially at $60,500 and $57,500 to $58,000. If not triggered, continue to wait. XGOOGL and XTSLA GOOGL holding between $317 and $320 on Friday indicates stronger support than Tesla. After reclaiming $324, targets are $332 and $342; if it falls below $311, the rebound plan is canceled. TSLA remains weak; only consider longs after recovering $328; if it falls below $306 and fails to rebound, downward targets are $296 and $285. Under the same account risk, TSLA's nominal position is only half of GOOGL's. OKX's XGOOGL, XTSLA, and BTC can all be traded within the same account. Tokenized US stock spot supports 24/7 trading, but spreads may widen outside US stock regular hours, so I only use limit and conditional orders. Maximum single loss is controlled at 0.5% of account equity, and total event risk for the three asset types before FOMC does not exceed 1%. The current opportunity depends on how long the divergence lasts. Between 63.5K and 69.5K, patience is more valuable than direction; once a boundary is broken, let orders take over judgment. Data Notes: US stock data as of July 24 close, BTC data as of July 25; correlation coefficients estimated from daily returns over the last 30 common trading days. #财报观察员：谁能看懂谷歌和特斯拉这次的真实答卷？ #新手必看：这里有你需要的一切

"Google's Net Profit Surges 298%, Tesla Drops Nearly 14%: Wall Street Begins Calculating AI Payback Period" July 24, 2026 · Friday Q3 · Issue 89 Aspirin · Cycle Analysis from a Data Scientist's Perspective After reading two earnings reports, the first thing I did was cross out the "net profit" line. Alphabet's net profit soared 298% year-over-year, Tesla's revenue grew 26%. According to typical earnings narratives, both numbers should have triggered a strong bullish candle. However, on the first trading day after the reports, GOOGL fell about 6.6%, and TSLA dropped nearly 14%. The AI narrative hasn't disappeared, but now there's a stopwatch alongside it. In the past two years, models, computing power, robotics, and autonomous driving could be valued ahead of time; starting this quarter, every dollar of capital expenditure must answer a very practical question: how long until payback? 1. Google: AI revenue is real, but the 298% net profit surge shouldn't be taken at face value Alphabet's Q2 revenue was $119.8 billion, up 24% year-over-year; operating profit increased 30%, with an operating margin around 34%. This part is solid; the core business remains strong. Google Cloud revenue reached $24.8 billion, up 82% year-over-year; operating profit was $8.8 billion, with margin rising from 20.7% to 35.6%. Enterprise customers are paying, and AI's scale effects have emerged. Net profit rose to $112.1 billion, a 298% increase; about $98 billion came from unrealized gains on equity investments, mainly related to SpaceX's valuation changes after going public. This profit won't recur every quarter. Excluding this, Google's earnings report is still good, just not as exaggerated as "tripling profits." Free cash flow was about negative $5.9 billion, and the full-year capital expenditure guidance was raised from $180 billion–$190 billion to $195 billion–$205 billion. Search and cloud businesses are profitable, but data center bills are coming faster. The stock price drop reflects this timing gap. 2. Tesla: More cars sold, but profit cushion is thinning Tesla's Q2 revenue was $28.236 billion, up 26% year-over-year; deliveries reached 480,126 vehicles. Demand is decent, but operating profit was only $398 million, down 57% year-over-year, with an operating margin around 1.4%, and automotive gross margin at 16.9%. R&D expenses grew 49% year-over-year, and the company expects capital expenditures to exceed $25 billion in 2026. Automotive cash flow is buying time for AI, robotics, Robotaxi, and energy storage, but the current business provides only a 1.4% profit buffer. The longer the payback period, the greater the valuation discount. Unrealized gains from SpaceX investments and taxes make net profit appear steadier than operating profit. Tesla still holds 11,509 BTC, with no sales this period; unrealized losses on digital assets were about $334 million in the first half. These amplify earnings volatility but don't fully explain the nearly 14% single-day stock drop. The market is focused on automotive margins, AI investment, and payback cycles. 3. Two earnings reports, the market faces the same question Google has provided a sample of AI monetization, with answers reflected in Cloud's 82% growth and 35.6% margin; Tesla has greater imagination but showed no operating leverage this quarter. Both companies must prove that new revenue can keep pace with rising capital expenditures. When interest rates remain high, valuations are based on certain cash flows, and long payback periods are discounted. The next phase of AI winners will need to keep revenue, margins, and free cash flow all in sync. Missing any one will make stock prices more selective. 4. My trading map: Four numbers are more useful than sentiment As of the close of U.S. markets on July 23, GOOGL was $319.53, TSLA was $322.74. The first big bearish candle after earnings only releases sentiment; the second pullback better reveals whether funds are willing to support. XGOOGL: 315 and 332 315 to 320 is the first defense line. If price shows a four-hour reversal in this range and closes back above 323, I will deploy one-third of the planned position on a pullback between 320 and 323, with a stop loss at 311, targeting 332 first, then 342; if 332 holds, the next target is 355. If the four-hour chart breaks below 315 and fails to rebound to 320, the spot bottom-fishing plan is paused, and the next observation zone shifts to 300–305. Earnings gaps rarely fill all at once, so there's no need to keep averaging down during a decline. XTSLA: 318 and 341 318 to 320 is short-term support. If the four-hour chart recovers to 328 and then pulls back to 322–326 without breaking, I can open one-quarter to one-third of the planned position, with a stop loss at 312, targeting 341 and 357; after 357 is reclaimed, we can discuss the gap near 374. If 318 breaks and the rebound is weak, the next observation zone is 300–305. TSLA's volatility is clearly higher than GOOGL's; under the same account risk, nominal position size is only half that of GOOGL. OKX's XGOOGL and XTSLA support 24/7 trading, but order books outside U.S. market hours may be thinner. I only use limit and conditional orders, with initial positions not exceeding one-third of planned funds, and single-trade maximum loss controlled at 0.5% of account equity. After the first target is hit, stop loss is moved near cost. Keep these four numbers on your screen: XGOOGL's 315 and 332; XTSLA's 318 and 341. They are closer to trading answers than the next AI news. After triggers, I will update entry, cancellation, and stop-loss moves on the platform. Data source: Alphabet's Q2 2026 earnings materials; #财报观察员：谁能看懂谷歌和特斯拉这次的真实答卷？ $XGOOGL $XTSLA

"Gold Returns to 4050: The Low Point Window for Gold Has Opened, Silver Is Still Early" Friday, July 24, 2026 Q3 · Issue 88 Aspirin · Cyclical Analysis from a Data Scientist's Perspective In the past 24 hours, XAU dropped from around 4156 to 4056, hitting a low of 4045; XAG fell from 60.2 to 57.8. Both gold and silver are declining, but the trading implications differ: gold is testing mid-term support, while silver is still undergoing more intense deleveraging. Treating them as the same bottom-fishing trade risks underestimating silver's volatility. 1. The Low Point Window for Gold Has Opened According to OKX daily data, XAU has retraced about 28% from the January high. The World Gold Council reports that since 1971, gold has experienced eight retracements exceeding 20% after reaching record highs, with an average retracement of about 36% and a median of about 29%. The current retracement is approaching the historical median, indicating the panic phase is not far off, though the lowest point still requires price confirmation. In terms of timing, the mid-term low in 2018 occurred in mid-August, and in 2022 it was in late September. July to October remains the main window for observing this cycle's low, with August to September weighted more heavily. The World Gold Council's mid-year framework views around $4100 as the macro consensus center for the second half of the year, with normal fluctuations roughly ±5%; if it breaks below 3860, technical selling pressure may trigger another round. 2. Three Lines Determine Trading Direction 4115: Bull Confirmation Line This is near the 4-hour 200 EMA, with a short-term previous high at 4168 above. If XAU closes steadily above 4115 on the 4-hour chart and then retests 4080 to 4100 without breaking down, you can open the first long position with 30% of planned capital, placing a stop loss at 4038; targets are 4168, 4250, and 4400 sequentially. After reaching 4250, move the stop loss to near cost to prevent the trend trade from turning into a losing position. 3990: Bear Trigger Line If the price breaks below 3990 effectively on the 4-hour chart and fails to reclaim 4000 to 4020 on a pullback, you can open a small short position or hedge existing gold longs. Stop loss is at 4070, with targets first at 3950, then 3860. The first target is for partial reduction, not rushing to guess the bottom; observe volume and closing structure near 3860. 3860 to 3800: Support and Acceleration Zone This area is near the long-term bull market support band; breaking this key level may trigger a final acceleration. Do not enter perpetual longs directly during the decline; wait for the 4-hour chart to reclaim 3900 or for a daily volume surge with a reversal candle, then place stop loss below the new low. For long-term allocation, if scaling in, only use 20% to 30% of planned capital, keeping reserves. Between 3990 and 4115, I do not chase trades. This range is close to both stop loss levels, and frequent trading risks being stopped out repeatedly. 3. Silver Has Not Yet Earned the Right to Lead XAG has retraced over 50% from the January high, dropping about 26% in the past 60 days; during the same period, the gold-silver ratio rose from around 58 to about 70. Silver's larger drop does not mean it will bottom earlier. When gold continues to weaken, silver often amplifies the decline. Going long on silver requires two conditions simultaneously: XAU holds above 4115, and XAG closes above 61 on the 4-hour chart and retests 59.8 to 60.2 without breaking down. Then open only 30% of planned capital, stop loss at 57.8, targets at 64.5 and 68. If XAG breaks below 54.8 and fails to rebound above 56, treat it as a continuation of weakness, stop loss at 58.2, targets at 52 and 49.5. The area below $50 is a deep value zone worth studying but does not imply an immediate V-shaped reversal. 4. How to Execute on OKX XAUUSDT and XAGUSDT perpetual contracts track spot gold and silver prices respectively, suitable for two-way swing trading and position protection. My approach is to pre-set breakout long and breakdown short triggers, canceling the opposite side once one side fills; all use isolated margin with leverage controlled between 1x and 2x, and single trade stop loss not exceeding 0.5% to 0.75% of account equity. Silver is more volatile, so nominal position size is no more than half that of gold for the same stop loss distance. Perpetual contracts have no expiry but still carry funding fees, margin requirements, and gap risks from traditional market trading hours. Avoid maxing out positions before weekends and major data releases; use mark price for stop loss triggers and check product availability in your region in advance. Gold confirms, silver amplifies. During the low window from July to October, preserving margin is more important than catching the first rebound. Next, I will focus on four numbers: XAU's 4115 and 3990, XAG's 61 and 54.8. Whichever triggers first dictates the trade direction; I will continue updating order cancellations and trailing stops on the platform. Data Notes: OKX market data and technical indicators as of July 24, 2026, 01:15 (Japan Time); historical and macro ranges reference public market data and the World Gold Council's "2026 Global Gold Market Mid-Year Outlook." The above is solely a personal trading framework and risk management record, not investment advice. Product permissions may vary by region; please refer to your actual account display. #Gold #Silver #XAU #XAG #OKX #TradFi

"SOXL Rebounds to 150: A Rally Is Possible, But the Bottom Is Not Yet Confirmed" Tuesday, July 21, 2026 Q3 · Issue 87 Aspirin · Cyclical Analysis from a Data Scientist's Perspective After a continuous plunge in semiconductors, a decent rebound has finally arrived. During the regular US stock market session, Ambarella rose 6.24%, Teradyne increased 3.54%, Marvell gained 3.32%; however, SOXL touched 148.47 intraday before clearly pulling back, closing with only about a 0.9% gain. SMH and Nvidia also closed slightly higher. Subsequently, OKX's XSOXL and SOXL perpetual contracts retraced back to the 149 to 150 range during the Asian session. This indicates buying interest has returned and short covering is amplifying volatility. It is enough to form a trade but insufficient to prove that semiconductors have bottomed. 1. Looking at the breadth of this rebound If this were a trend reversal, the first sign would be a broad sector lift. Currently, the gains remain concentrated in a few oversold names: SOXL shows large intraday swings, SMH and Nvidia have limited follow-through, and the S&P 500 is not strengthening in sync. Capital seems more like it is covering crowded short positions while probing valuation lows ahead of earnings, rather than significantly adding to the entire chip sector. Technical structure supports this view. As of 15:30 on July 21, SOXL perpetual is about 149.9, the 4-hour 20 EMA is at 140.0, and the 50 EMA at 150.9; the daily 20 EMA remains at 172.5, and the 50 EMA at 192.8. The price has just hit the first short-term resistance, still some distance from mid-term trend recovery. MU and MRVL’s 4-hour RSI have risen to 76-77. The sharper the rebound, the less it should be interpreted as a "new trend beginning" simply because of a bounce from extreme oversold conditions. 2. I will handle this with three scenarios First, breakout continuation. If SOXL closes above 151.5 on the 4-hour chart, then retests 147 to 149 without breaking below, while SMH holds above 577.5 and MU stabilizes above 915, the sector is confirmed. At this point, 30% of the planned position can be established as the first long, with a stop loss at 143.5; first target at 158, second at 167, and 172 to 173 corresponding to the daily 20 EMA, suitable for further trimming rather than temporarily increasing conviction. Second, pullback entry. If the price fails to break out and falls back to 139.5 to 142.5 with volume contraction and stops falling near the 4-hour 20 EMA, the risk-reward ratio is better than chasing 150. Handle in three parts with ratios 40%, 30%, 30%; a valid 4-hour break below 133.5 indicates this rebound structure has failed. Rebound targets are 149, 158, and 167 respectively. Third, rebound failure. If SOXL breaks below 140 on the 4-hour chart, rebounds to 142-145 but fails to hold, a small short or hedge position can be used with a stop loss at 149.5, targeting 134, 129, and near the previous low at 117. Those already holding semiconductor longs should focus on reducing net exposure, not opening full long and short positions simultaneously. Between 140 and 151.5, I will just observe. This range is directionally unclear, and frequent chasing of rallies and sell-offs is the easiest way to give back rebound profits. 3. Earnings reports should focus on capital expenditures, not just EPS In the coming weeks, Alphabet, Microsoft, Meta, AMD, and Nvidia will release earnings. The market’s real concern is whether large cloud providers are cutting AI capital expenditures, and whether chip companies’ orders, gross margins, and inventory guidance continue to deteriorate. If capital expenditures are not cut and SMH can hold above 577.5, the rebound may upgrade from short covering to fundamental recovery. If profits beat expectations but capital expenditures are cut, or MU falls back below 860, the market will likely return to a deleveraging logic. 4. How to implement on OKX For bullish bias, consider XSOXL, XMU, XNVDA, and XQQQ. XSOXL is spot-based but tracks SOXL, which is already a triple-leveraged daily ETF, so no need to layer additional 5x or 10x contract leverage. A more prudent portfolio is to use XQQQ or XSPY as the base position, with XSOXL as a satellite position. For bearish bias or to protect existing positions, use small-sized SOXL-USDT perpetual isolated margin orders with pre-set conditional stop losses. Only execute either the long plan above 151.5 or the short plan below 140, canceling the other once triggered. Single stop loss should control 0.5% to 0.75% of account equity, which is more important than guessing a single candlestick correctly. Long-term semiconductor demand has not been overturned by two days of sharp declines, and short-term deleveraging will not end with a single rebound candle. Trade continuation above 151.5, defense below 140, and patiently wait for the market to show direction in between. Whichever price triggers first, I will continue updating execution and cancellation logic on the community. Focus on price, not narratives. Data notes: US stock regular session price changes are from July 20 closing public data; OKX prices and moving averages are as of July 21 15:30, calculated from public market data. #SOXL #Semiconductors #USStocks #OKX #TradFi #ChipStockRebound, US stock short positions hit record highs

"ETH Stands at the $1900 Threshold: The Dollar Has Risen, But the Coin-Based Level Hasn't Cleared Yet" Monday, July 20, 2026 Q3 · Issue 86 Aspirin · Cyclical Analysis from a Data Scientist's Perspective 1. ETH's Structure Is Slightly Stronger Than Its Sentiment As of 11:50 on July 20, OKX's ETHUSDT is around $1879, with a 24-hour range of $1850 to $1891. The 4-hour 20 EMA is at 1861, the 50 EMA at 1842; the daily 20 EMA and 50 EMA are at 1808 and 1817 respectively. The price has already returned above the short-term moving averages, the daily RSI is about 59, and the current funding rate is approximately +0.0044%. This is a tradable rebound, not yet crowded enough to require contrarian moves. The upside is not empty. $1945 is a high point reached earlier this month; the 20-week moving average and 21-week EMA cluster between $2005 and $2043, and the daily 200 EMA is still near $2199. $1900 is only the first gate; $2005 to $2043 is the resistance zone that this rebound must respect. BTC has simultaneously reached 64.8K, almost touching the daily 50 EMA at 65.0K, with the previous high near 65.6K. ETH's current rise still carries a clear BTC beta. If BTC fails to break above 65.6K for a long time, ETH's sustainability in breaking through $1900 alone will be discounted. 2. 0.0300 Explains More Than 1900 ETH/BTC is currently about 0.02896, with a recent 4-hour high at 0.02908, and the 20-month simple moving average around 0.0299. These three numbers are close together, indicating that the coin-based level is exactly at a directional choice point. If ETH rises above $1900 and ETH/BTC also closes above 0.0300, the market begins to actively assign ETH a higher valuation, and rebound positions can be held longer. If ETH rises but the coin-based level remains below 0.0291, the dollar gains mainly come from BTC's rise, and it is wise to take profits near $1945. The opposite is also true. If ETH falls but ETH/BTC holds above 0.0287, it looks more like systemic volatility brought by BTC, and it is not suitable to chase shorts at support. If ETH breaks below $1835 and the coin-based level simultaneously loses 0.0287, that signals renewed altcoin risk spreading. 3. Bulls Wait for One More Confirmation Trigger conditions | ETH closes above 1895 on the 4-hour chart, holds support on a pullback between 1875 and 1890; ETH/BTC closes above 0.0291 simultaneously. If the coin-based level further stabilizes above 0.0300 on the daily chart, the bullish confirmation level rises one notch. Entry and stop loss | After pullback confirmation, use 25% of planned funds to establish tactical long positions between 1880 and 1895. Isolated margin, leverage no more than 2x; stop loss at mark price 1848. Tiered take profit | Take 30% profit at 1945, another 40% at 2005, and hold the rest targeting 2040 to 2045. After the first tier is filled, move the protection level to near the entry price. If the daily closes above 2045 and ETH/BTC holds 0.0300, move the remaining target up to 2110 to 2140. Time stop loss | If the next three 4-hour candles fail to close above 1915 after entry, reduce position by half. Without momentum, continuing to hold only turns a good entry into a choppy position. 4. Bears Wait for 1835 Breakdown Trigger conditions | ETH closes below 1835 on the 4-hour chart, fails to reclaim 1835 to 1850 on a pullback; ETH/BTC simultaneously breaks below 0.0287. If BTC also closes below 63.9K, this bearish scenario gains higher reliability. Entry and stop loss | After pullback failure, use 20% of planned funds to establish tactical short positions between 1835 and 1848. Isolated margin, leverage no more than 2x; stop loss at mark price 1882. Tiered take profit | Take 25% profit between 1815 and 1805, another 35% at 1775, and hold the rest targeting 1725. After the first tier is filled, move the protection level down to near the entry price. Time stop loss | If the next three 4-hour candles fail to close below 1815, reduce position by half. If there is no follow-up selling after breaking support, the quality of the breakdown is insufficient. 5. 82% of CLARITY Is No Longer Valid Some still share the old market saying "CLARITY bill's probability of passing this year has dropped to 82%" today. Polymarket's Yes price on July 20 is about 37.5%, down about 8 percentage points in the past week. The market's judgment on completing legislation this year has dropped below 50%. The bill passed the House in July 2025 by 294 to 134; the Senate Banking Committee advanced the text in May this year by 15 to 9. It has not yet passed the full Senate, nor reached the President's desk. Committee advancement and formal law are separated by scheduling, amendments, floor votes, and final signing. CLARITY will affect the industry's long-term discount. The boundary between SEC and CFTC will be clearer, and compliant trading, custody, and on-chain finance may benefit. But it cannot help ETH break above 0.0300, nor protect long positions' stop losses. If news causes ETH to spike briefly while ETH/BTC remains below 0.0300, I treat it as a profit-taking window. Only if a clear Senate floor schedule appears later and the price simultaneously breaks the coin-based level is it worth increasing positions. Note: CLARITY bill procedures are from Congress.gov and the US Senate Banking Committee public information; prediction market prices are from Polymarket, with probabilities changing in real time. Historical paths are for scenario analysis only and do not guarantee future repetition. #CLARITY年内通过预期降温

"BTC Falls Back to 200-Week Moving Average: Bulls Wait for 63.8K, Bears Wait for 62.2K" Friday, July 17, 2026 Q3 · Issue 85 Aspirin · Cycle Analysis from a Data Scientist's Perspective BTC is back at 63K. Bulls tend to treat the 200-week moving average as an automatic rebound button, while bears are prone to chase at support. Today, I placed new contract orders outside the range: go long on a 4-hour close above 63.8K; go short after a failed rebound on a 4-hour close below 62.2K. Whoever confirms first is the one to follow. 1. The area around 63K is critical but direction is not yet clear OKX's BTCUSDT is about 62.9K, with a 24-hour low at 62.5K. According to OKX K-line estimates, the 200-week moving average is about 63.1K, the 20-day daily moving average is about 63.3K; the 4-hour 50 EMA and 20 EMA cluster between 63.6K and 63.8K, and the daily 50 EMA is near 65.0K. The price just dropped below long-term support, with a group of short-term moving averages pressing overhead. Contract data shows no typical panic liquidation. The current funding rate is about +0.0044%, and the long-short account ratio is about 1.77. When the price falls, long accounts actually increase, indicating the market is not short of buyers. If 62.5K is tested again, be wary of a round of long stop-losses. 2. Both cycle clocks still have some time to go Counting from the last low, this cycle is about day 1333; the previous two lows were around day 1430. Counting from this cycle's high, it is about day 283 now; past bear market lows often appeared between days 364 and 406 after the high. These two samples suggest about 80 to 120 days remaining. This is only a cycle sample and cannot be used as a countdown to a specific day. If the market undergoes concentrated deleveraging, price capitulation could bring the low forward; if volume remains light, time digestion may extend to late Q3 or early Q4. Therefore, I split my funds into two layers: tactical positions for the remaining July window, and cycle positions remain in cash. Take rebounds for short-term trades; if structure breaks, short is also possible; do not use all chips on a single 4-hour K-line for long-term positions. 3. Bullish route: enter after reclaiming 63.8K Trigger condition | 4-hour close above 63.8K, then a pullback holding between 63.3K and 63.6K. A single breach does not count; both close and pullback are required. Entry and stop-loss | After pullback confirmation, use 20% of planned funds to build tactical long positions between 63.5K and 63.9K. Isolated margin, leverage no more than 2x; mark price stop-loss at 62.4K. Partial take-profit | Take one-third at 64.9K; another third at 65.6K; remaining position targets 67.3K. After the first take-profit, move the protection level near the entry price. If the daily closes steadily above 65.6K, the rebound has a chance to challenge 69.8K to 70.4K. Time stop-loss | If the three 4-hour K-lines after entry still fail to close above 64.5K, reduce half the position. Lack of momentum after returning to support often means the breakout quality is insufficient. 4. Bearish route: wait for a rebound after breaking 62.2K Trigger condition | 4-hour close below 62.2K, then a rebound failing to hold between 62.2K and 62.6K. Do not short on the first touch of 62.5K; give support a chance to react. Entry and stop-loss | After rebound failure, use 20% of planned funds to build tactical short positions between 62.2K and 62.5K. Isolated margin, leverage no more than 2x; mark price stop-loss at 63.4K. Partial take-profit | Take one-third at 61.3K; another third between 60.0K and 59.5K; remaining position targets 57.8K. After the first take-profit, move the protection level near the entry price. Time stop-loss | If the three 4-hour K-lines after entry still fail to close below 61.8K, reduce half the position. If the breakout does not continue, do not keep paying the cost of volatility. 5. Tokenized US stock spot adds an extra confirmation layer for BTC On July 16, OKX opened the first batch of 24 tokenized US stock spot trading pairs in four phases. The list includes XMU, XNVDA, XTSLA, as well as stock and ETF exposures like XSPY, XQQQ, XSOXL, all quoted in USDT and supporting 24/7 trading. I treat XSPY and XQQQ as risk appetite anchors, and XNVDA and XSOXL as high-beta thermometers. During regular US stock trading hours, if BTC reclaims 63.8K while XSPY and XQQQ hold the day's open and XNVDA does not lag significantly, the bullish breakout is more credible; if BTC rises alone while XQQQ and XNVDA weaken, the move looks more like short covering, and 64.9K should be taken more readily. Bears have corresponding validation. If BTC closes below 62.2K and XQQQ and XNVDA simultaneously break the day's low, risk assets weaken in resonance, and short rebounds can be executed as planned. When tokenized US stocks remain stable but BTC breaks down alone, it is more likely internal deleveraging in crypto markets, so the first take-profit should be more aggressive. These products provide price exposure to stocks or ETFs without shareholder rights like voting, and currently do not support single-asset leverage. After US stock market closes, pricing references the latest close and market valuation, so spreads may widen. I only use limit orders during the initial launch of new pairs, with trial positions no more than a quarter of regular size. Although XSOXL is spot, it tracks SOXL, a triple-leveraged ETF, so volatility does not disappear just because it is spot. Final sentence Between 63.8K and 62.2K, I do not guess; if it stands back above 63.8K, go long on the rebound; if it closes below 62.2K, wait for a rebound to short. XSPY and XQQQ verify the quality, BTC's close triggers the order. Tradable paths are always those with room after triggering. #在OKX交易代币化美股现货

"BTC Like 2018, Funding Like 2019: Three-Step Trading Plan After CPI" July 14, 2026 Q3 · Issue 83 Aspirin · Cycle Analysis from a Data Scientist's Perspective The market is running on two clocks simultaneously Looking only at price structure, 2026 resembles a less volatile version of 2018: a low in February, a higher low formed from late March to early April, a lower high in May, and a sweep of the February low again from late June to early July. We are currently at the mid-July pullback level, with even the weekly rhythm very similar. The funding environment, however, is closer to 2019–2020. Manufacturing, semiconductors, or other cash-flow-supported assets can strengthen, but funds have no reason to automatically flow into every altcoin. The price cycle determines which stage BTC is in, while the business cycle determines which asset classes receive capital. These two clocks can run simultaneously. As of 17:00 on July 14, OKX's BTCUSDT is about 62.7K. The 200-week moving average is about 62.8K, the 20-week moving average about 69.9K, the 21-week EMA about 71.5K; the 50-day moving average on the daily chart is near 64.6K. Long-term support below is very close, and medium-term resistance above is clear. This is a compressed market phase with direction yet to be confirmed. 1. Tonight's CPI, focus on three closing prices The US June CPI will be released tonight at 20:30, followed by PPI tomorrow night. The first 5-minute or 15-minute candle after the data release mostly reflects liquidity and stop-loss orders. I will not chase prices at that moment but wait for the 4-hour close before taking action. 64.7K | Rebound confirmation. A 4-hour close above 64.7K, followed by a pullback to 64.0K–64.5K without breaking, is the tactical long entry condition. Stop loss at 62.4K, with targets at 67.3K, 69.9K, and 71.5K respectively. 62.8K | Bull-bear pivot. Price oscillates around the 200-week moving average. Spot can be accumulated in small amounts; contracts should not be used to bet on data direction here. 61.3K | Plan invalidated. A 4-hour close below here cancels breakout longs. The next observation zones are 60.3K–59.5K and previous lows at 58.0K–57.5K. Do not chase shorts near support. 2. Divide funds into 100 parts 10 parts | Build BTC spot base near 62.8K, no leverage. 10 parts | Place second-tier spot orders at 60.3K–59.5K. 10 parts | Reserve for sweeping previous lows at 58.0K–57.5K. 20 parts | Use only after 4-hour close above 64.7K and pullback confirmation for tactical rebounds, isolated margin, no more than 2x leverage. 50 parts | Keep cash reserves. If August retraces July gains, only late September to early October will offer space to build more significant cycle positions. This allocation allows me to be wrong on the short term without missing the long term. If price goes straight up, I already have 30 parts base and 20 parts confirmed positions; if the market dips further, half the cash remains. 3. Aggressive moves possible late July, default realization in August Similar rhythms appeared in 2018 and 2022: rebounds in July, giving back gains in August. The current path also allows for a mid-July pullback followed by another rise, roughly late July to early August. If the rebound reaches 67.3K, I will take one-third off; at 69.9K, another third; near 71.5K, handle remaining positions based on weekly strength. If the weekly close is above 71.5K and holds 69.9K the following week, the baseline scenario of August retracement fails, and remaining positions turn into trend holds. If price never surpasses 64.7K or breaks out but quickly falls back to 62.8K, the rebound is treated as a short-term repair. I lean toward placing this cycle's low window in late September to early October. If a concentrated deleveraging occurs, timing may advance; if the market continues to drift down, patience matters more than price levels. 4. BTC holds support, altcoins may continue bleeding In 2018, BTC consolidated around 6K for months, while altcoins fell from August through October. BTC not making new lows only means BTC itself is temporarily stable, not that it can guarantee altcoins. My altcoin tactical positions total no more than 5% of the strategy account, no leverage, and no averaging down just because of bigger drops. Survivors must have users, revenue, or clear products; projects relying on sentiment and new buyers to maintain price are most vulnerable in tight liquidity phases. 5. Use gold and semiconductors to judge rebound quality BTC rising, SOXL or SMH strengthening simultaneously, and gold weakening indicate risk appetite is returning, making the rebound more sustainable. BTC rising while semiconductors weaken and gold strengthens suggests a short-covering rally. Above 67.3K, I would be more willing to take profits. BTC, semiconductors, and gold all falling means the market is scrambling for liquidity. Cash is king; no "long-term bullish" view justifies high leverage. SOXL is already a high-volatility instrument; I treat it as a thermometer and avoid adding a second layer of leverage. 6. Set the roadmap in advance on OKX I will set five alerts: 61.3K, 62.8K, 64.7K, 67.3K, 69.9K–71.5K. Put BTCUSDT, SOXLUSDT, and XAUUSDT on the same TradFi watchlist to quickly identify whether funds chase risk, defend, or simply cover shorts after data releases. Wait for 4-hour confirmation before breakout trades; stop loss triggered by mark price; tiered take profit only reduces positions. Manage spot accumulation and contract tactical positions separately with different stop losses. Before the July 29 Fed meeting, if BTC remains below 69.9K, I will continue reducing elastic positions. After CPI and the late July rebound path trigger, I will continue updates on the OKX platform. Just follow along; no need to spend the whole night staring at the screen waiting for a single data candle.

"July for the rebound, August to protect profits: BTC Summer Trading Roadmap" July 12, 2026 Q3 · Issue 81 Aspirin · Cycle analysis from a data scientist's perspective For the next six weeks, I will treat July as a rebound trading window and August as a profit defense period. Only add to BTC positions if the daily chart holds above 65K; start taking profits at 67-68K, actively reduce at 73-74K; if it falls below 60K, tactical positions exit. I've written extensively about the four-year cycle. How to act in the next six weeks deserves today's focus. As of July 12, BTC spot on OKX is about $64,000. The 20-day moving average is near 61,900, the 50-day near 64,950, and the 200-day near 73,900. The price has reclaimed the 20-day MA and is facing resistance at the 50-day MA; this is a rebound that has started but has not yet confirmed the trend. Historical samples also provide a clue suitable for trading plans: in the mid-term years 2014, 2018, and 2022, BTC July returns were approximately -5%, +38%, and +20%, averaging about +17.7%; however, August and September in all three samples closed lower. The sample size is small and not for fortune-telling, but enough to remind us: July is for offense, profits should not be carried intact into August. 1. Mid-July: Do not chase 64K, wait for data to give direction The US CPI on July 14 and PPI on July 15 are the first key points. BTC is currently right around the 50-day MA; chasing a bullish candle before data release has poor odds. 61.5-62.3K is the first support zone, also where the 20-day MA lies. If it pulls back and reclaims this area, small positions can be added; if the daily closes below 61.5K, reduce tactical longs by half. 64.8-65.5K is the confirmation zone. If the daily closes with volume above 65K, I will add tactical positions up to half of the planned maximum; only after breaking 67.3K will I add the remaining portion. Intraday spikes do not count as confirmation. 67-68K is the first profit-taking zone, corresponding to the recent monthly high. Take one-third profit here first. 73-74K near the 200-day MA is the main resistance zone for this summer rebound; take another one-third to one-half profit here. Profits must be realized at resistance levels, not left for the market to decide. 2. July 29: Use the Federal Reserve meeting as a position dividing line The Fed will meet on July 28-29. If BTC has broken 67.3K before the meeting, part of the trend position can be retained; if still pressured between 65-67K, the rebound is likely just a correction in a downtrend, so I will reduce leverage and altcoin positions before the meeting. A practical judgment is whether semiconductor assets like SMH and SOXL strengthen alongside BTC's rise. If both rise together, risk appetite is indeed returning; if BTC rises alone while semiconductors remain weak, it looks more like internal short covering in crypto, which tends to sell off after the spike. If CPI is hot, crude oil and gold strengthen simultaneously, and semiconductors weaken, the market is trading inflation and risk aversion, and BTC's high beta characteristic is vulnerable. At that time, cash and gold take priority, reducing crypto elastic positions. 3. August: Default to protecting profits, only continue long if conditions are met Employment data on August 7 and CPI on August 12 are the second key points. How August is handled depends on where July closes. If July closes above 67.3K and then pulls back without breaking 65K, BTC trend positions can be retained. Consider adding elasticity to ETH only after it stands above $1,850; SOL needs to break $84. ETH, SOL, and XRP combined should not exceed 10% of the strategy account and must follow BTC confirmation. If BTC consolidates between 60.5-68K, 10%-15% of funds can be used for spot grid trading within that range, with 12-20 grids. Stop the grid if it breaks below 58K. This phase is not suitable for futures grid trading, as a one-sided drop quickly turns "profit from volatility" into "holding the trend." If the daily closes below 60K, reduce tactical positions by half again; if it breaks 57.8K, exit all tactical longs and grids. Whether to keep long-term base positions depends on the ability to endure volatility for over a year, but do not use leverage to lower cost. 4. Divide the account into 100 parts, I will allocate as follows 40 parts in BTC spot base positions, 20 parts for July tactical trades, 30 parts kept in USDT, and the remaining 10 parts for ETH, SOL, XRP, or TradFi elastic positions. Before BTC firmly holds 65K, keep the last 10 parts in cash. Leverage in the strategy account is controlled within 2x, using isolated margin; single stop-loss losses should not exceed 1% of total funds. Those already holding high-leverage semiconductor, gold, or crypto longs are even less suited to add four or five same-direction longs near 64K. 5. Pre-set the plan into orders on OKX I will set four price alerts: 62K, 65K, 67.3K, 74K. 62K for defense, 65K for confirmation, 67.3K to start taking profits, 74K to actively reduce positions. Set breakout, pullback, and stop-loss orders in advance; do not chase prices during data releases. Spot grids only open after range confirmation and stop when price leaves the range. Tools do not replace judgment but prevent emotional plan changes. I will continue updating this roadmap's triggers on OKX Planet after July 14, July 29, and August 12. Bullish or bearish views can change with data, but price and position discipline must not be changed on the fly. Final sentence The task in July is to earn from the rebound; the task in August is to take the earned profits away. The direction may be wrong, but positions must not get out of control. #沉睡比特币案迎行业机构介入 #BTC与ETH现货ETF终结八周流出

"Signal Contradiction, Interest Rates Unchanged: July Rebound Still a Window" July 9, 2026 Q3 · Issue 80 Aspirin · Cyclical Analysis from a Data Scientist's Perspective Key Judgment: The contradictory signals between the US and Iran indicate that geopolitical risks have not yet exited; the Federal Reserve minutes lean hawkish, indicating liquidity has not loosened. The July rebound of BTC can be respected, but it looks more like a window trade in a weak cycle rather than a confirmed trend reversal. These two news items together are more useful than just looking at the candlestick charts. The first is about the US and Iran. Both sides are sending negotiation signals while continuing to threaten each other, yet the market reaction is relatively restrained. BTC did not plunge in panic, and crude oil did not price in the worst-case scenario all at once. The second is the Federal Reserve minutes. On the surface, rates remain unchanged, but the message is not dovish. Inflationary pressure persists, and there is even discussion that "if inflation remains high, further rate hikes may be necessary." Together, these two facts represent the market's most realistic position now: geopolitical risk is not out of control, so risk assets still have room to rebound; interest rates have not loosened, so there is still a valuation ceiling. 1. Why the Market Is Not Panicking Regarding the US-Iran situation, the key is not who talks tougher but whether energy and shipping are under real pressure. If the Strait of Hormuz and the Mandeb Strait are repeatedly threatened but shipping and energy prices do not spiral out of control, the market treats this as a "risk premium" rather than a "liquidity crisis." This explains why BTC remains resilient today. The market does not believe geopolitical risk has disappeared but is not pricing a full escalation as the main scenario for now. However, we should not be too optimistic. The biggest feature of geopolitical risk is that it usually acts like background noise but immediately becomes a main variable once it affects energy prices and inflation expectations. 2. The Federal Reserve Is the Ceiling for This Rebound What truly caps risk assets is interest rates. The most important point in this meeting's minutes is not "no rate hike" but "why no rate hike." Officials have not confirmed that inflation is resolved; they have chosen to continue observing. If energy, AI investment, tariffs, or supply shocks continue to push prices higher, the Fed will find it difficult to deliver the easing scenario the market wants. This has a direct implication for BTC: panic easing below allows for a rebound; real interest rate pressure above makes a swift trend reversal difficult. 3. So the July Rebound Can Be Watched but Should Not Be Chased Too Aggressively My judgment on BTC remains unchanged: July is a brief strong window within a weak cycle. Similar rhythms appeared in 2018 and 2022: lows formed around June, rebounds occurred from July to mid-August, but much of the gains were given back in August and September. This cycle is similar. The market's short-term focus should not be on emotional slogans but on three levels: whether the 57K low remains valid, whether the pullback can form a higher low, and whether the 200-day moving average can be firmly held. Unless all three conditions are met simultaneously, I will not upgrade the July rebound to a trend reversal. 4. How I Will Execute on OKX First, place BTC, CL, BZ, and XAU on the same watchlist. For US-Iran risk, watch crude oil; for dollar credit and safe-haven demand, watch gold; for BTC, watch whether risk appetite is willing to flow back. Second, do not chase emotional trades on BTC; wait for pullbacks and confirmation. If it can hold key resistance, consider adding positions; if it reaches resistance but volume is insufficient, prioritize reducing leverage. Third, high-volatility assets like ETH, SOL, and XRP should not be rushed ahead of BTC. Before BTC confirmation, altcoin rebounds are mostly liquidity spillover, suitable for short trades, not heavy faith-based positions. Fourth, watch semiconductors and AI chains together. The Fed minutes place AI investment, energy, and inflation in the same pressure framework, meaning volatile assets like SOXL and MU must be evaluated not only on their stories but also on interest rates. Fifth, I will use conditional orders and price alerts rather than chasing trades live. Range markets can use small-scale grids, but near the 200-day moving average, crude oil volatility, or Fed expectation changes, grids must follow directional judgment. Today's Conclusion US-Iran is the source of volatility; the Fed is the valuation anchor. Geopolitics has not fully escalated, so BTC can rebound; interest rates have not truly loosened, so the rebound cannot be directly considered a reversal. What I care most about now is not today's price moves but whether the market can hold a higher low on the next pullback. If it holds, the July window continues; if not, a retracement in August-September must be guarded against. I will continue to synchronize this cross-asset observation and trading plan on the OKX platform. For how I apply macro events to BTC, gold, crude oil, and semiconductor positions, follow my homepage updates. Risk Warning: Contracts, copy trading, grids, leveraged ETFs, and TradFi contracts all carry risks. The above content is for personal research and trade review only and does not constitute investment advice. #美联储纪要：讨论过加息，仍一致维持利率 #美伊信号矛盾，市场反应克制