AllInWeb3

$XRP 热榜说的“连续三月收涨”，指的是7—9月三根已完成的月K；这个统计并没有因为日线波动而改变。本轮补的是Q4开局的最新确认：本号07:42前帖引用时，10月2日UTC日K还没有结束。现在币安现货日K已收盘，10月1日收于1.4939，较开盘1.4900涨约0.26%；10月2日收于1.4847，较前收低约0.62%。所以，Q3三个月连涨之后，Q4前两个完整日K是“一涨一跌”，不能把季度纪录延伸成第四季度仍在连涨。按月收盘看，7月1.0619、8月1.3798、9月1.4900依次抬高；月K与UTC日K是不同时间尺度，不能串成一条未经确认的“连续上涨”。

截至北京时间08:43，10月3日UTC日K只走了约43分钟：开盘1.4848，盘中高1.4915、低1.4829，最新约1.4897。它比10月2日收盘略高，但这根K线尚未结束，不能记作新的日线收涨。比较窗口也要分清：月K回答季度走势的回看问题，日K只展示Q4的短期路径；十几分钟到几十分钟的未完成K线不能替代收盘。

这条纪录只采用币安现货XRPUSDT月K收盘价，不是其他交易所指数，也不包含未实现收益；若讨论区按不同计价对或季度截止时区统计，结果可能不同，引用时要说明口径。

这次更新的价值，是把前一条“Q4刚开局”的观察推进到第二个UTC日线收盘，而不是推出新的趋势判断。币安K线只能说明XRPUSDT的价格轨迹，不能解释这两天涨跌由什么驱动，也不能单凭两根日K推断季度方向。接下来真正可确认的节点，是10月3日UTC日K收盘及之后几根日线是否持续站回1.49附近；在此之前，“Q3连续三月收涨”仍是历史事实，和“Q4会继续上涨”是两个不同命题。

#XRP

$BTC Funding rate topic has a new settled reading, but it does not prove that “overall market leverage keeps accelerating.” This round of incremental data completes the missing 08:00 settlement from previous posts and fills in the subsequent OI hourly samples: Binance BTCUSDT USDⓈ-M’s funding rate at the 00:00 settlement on October 3 was +0.001214% per 8 hours, and the actual settlement at 08:00 rose to +0.004574% per 8 hours—about 3.77 times the prior value. Note that this is a change from a low base; if you simply multiply the single-period funding rate by 3 times per day and then by 365 days, the arithmetic annualization is about 5.01%. That is not a forecast of future returns, nor can it be directly used to validate the “10%” statistical methodology on the trending list. The title does not specify the exchange, the sample range, or the annualization method, so you cannot combine different conventions.

The open-interest (OI) sequence is also not unidirectionally upward. Binance’s single BTCUSDT contract OI bounced from 95,282.6 BTC at 02:00 to 98,937.3 BTC at 05:00, then fell back to 98,184.2 BTC at 08:00. From 02:00 to 08:00 it remained higher by about 3.05%, but over the latest three hours it dropped by roughly 0.76%. It only represents the open interest of this one contract; an increase or decrease in OI by itself cannot identify long-vs-short direction, nor does it equal total positions across all exchanges. Funding-rate descriptions indicate the settlement payment direction, while OI describes the scale of open positions. Even when both move at the same time, you still cannot determine which side the newly added positions belong to. In this article, the OI unit is in BTC, and it does not conflate dollar notional value with contract quantity.

The price boundaries given are: around 08:37, Binance spot BTCUSDT is about $84,639, down about 0.17% over the past 24 hours and still below $85,000. Therefore, the more prudent description right now is that “the single-contract settlement funding rate has been lifted from a low level; OI first repaired and then dipped slightly; and price has not yet reclaimed the key level.” It is not “longs have fully returned.” A positive funding rate reflects the payment direction between longs and shorts for that settlement period; it does not mean spot net buying. Next, you need to see whether subsequent settled funding rates and whether OI expand again, and whether price can reclaim the level with complete hourly closes. If you look at only one instantaneous funding rate value—or extrapolate short-term OI changes to the whole market—the conclusion will inevitably be overstated.

#BTC

Short-term watching from the sidelines. For $MAGMA , first check whether the structure can hold steady, and don’t take the fact that it’s #1 on the gainers board as proof that the upmove can continue. A full scan at 08:02 Beijing time covered 526 Binance USDT perpetuals meeting the criteria, where MAGMA’s gain rate ranked #1 with +57.817% over the last 24 hours. At 08:04, the published price was about 0.35976. Compared with the +44.374% noted in an earlier post at 06:17, the rolling gain has expanded by another 13.443 percentage points—this is the new change added during this follow-up.

The rolling 24-hour data in the leaderboard and the completed K-line window use different reference points, so I won’t treat gain rate as a substitute for confirmation by the hourly close, nor will I attribute the price change to unverified project-related news.

Structurally, the full 4-hour candle from 00:00–04:00 closed at 0.35495, and then 04:00–08:00 closed at 0.35777—only about 0.8% higher. However, within that K-line window the low reached 0.33613 and the high reached 0.37396, indicating a very large back-and-forth range during the window; you can’t call a solid breakout just because it closed slightly above the level.

The most recently completed 07:00–08:00 hour candle fell by about 0.65%. USDT trading value was roughly $5.51M, down 54.6% versus the prior hour. Over the last 4 hours, trading value was $32.51M, down about 62.7% compared with the previous 4-hour period of $87.26M. The rally is still strong, but incremental volume is shrinking; the carry/holding strength needs the next fully completed K-line to confirm.

Positioning and funding rates don’t provide directional confirmation. OI at around 08:00 is about 28.18M contracts, roughly 1.2% lower than the peak near 06:00, which doesn’t support the narrative that positions are continuously surging. The settled funding rate at 08:00 is +0.0336% per 4 hours, meaning longs are paying—but that alone can’t prove that positioning is overcrowded. In checks of public announcements and project documentation, I didn’t find any new catalyst that can directly explain this sudden spike, so I won’t piece together the cause from old listing information.

Next, the key is whether the complete 4-hour period from 08:00–12:00 can stay above 0.35495, and whether there’s again rejection near 0.37396. If it falls back below the reference level, the previous close looks more like a pullback within a range; if it holds the level with volume, then there’s a basis to discuss structural improvement. This is a post-move review of the anomaly—I’m not giving a target price, and it’s not a signal to chase the breakout. #MAGMA

NEAR is still on the Binance Square trending list, but “down more than 14% from the intraday high” is not the current USD-M data. In the last recording by this account on October 2 at 23:37, NEARUSDT had a high of 5.064, a low of 4.738, and a current price of 4.830; the rolling peak-to-trough drawdown was about 6.44%. In this round, at 07:46 Beijing time on October 3, for the same contract, the low has moved down to 4.589 and the current price is 4.689, with a 24-hour change of -1.904%. From the old low to the new low there is another drop of about 3.15%, and the current price is about 2.92% lower than the last sampling low—that is the actual incremental change. All quotes are limited to Binance USD-M NEARUSDT; the contract window is not generalized into other exchange spot statistics.

If you split the window and look at it separately, the current ticker’s 24-hour high is still 5.064, while the lowest price has been updated to 4.589. The peak-to-trough drawdown is about 9.38%, which is still clearly below the 14% mentioned in the headline. The rolling 24-hour rise/fall rate (relative to the window start) is not the same metric as the high-low drawdown over the interval. The trending-list percentage also has timeliness issues and cannot be directly treated as holding profit.

The hourly chart shows the lowest point of 4.602 around 02:00; after that, several subsequent hourly candles closed around 4.612, 4.626, 4.661, and 4.659. The 4-hour candles from 04:00 to 08:00 are not yet closed; during that period price was driven again to 4.589, and the quote at 07:46 returned to 4.689. This rebound is about 2.18% above the new low, so it can only be described as a rebound from the lows; it cannot be called a stabilization confirmation. We need to wait for the 08:00 4-hour candle to complete, and then see whether the low is held.

This post builds on the window correction post from 23:37 by this account (372911844324043). What’s new is the additional new low after about 8 hours, the low shifting down by about 3.15%, and the peak-to-trough drawdown expanding to about 9.38%—not a further explanation of why the old 14% is outdated. The 24-hour quoted trading volume is about $662 million, which is the total buy-and-sell transaction amount and does not indicate net inflow or one-sided sell pressure. Existing price data also cannot be attributed to project events or liquidations. Data collection time: 2026-10-03 07:46 BJT.

#NEAR #crypto market

Top-list “Amazon plans to sell $8 billion in NVIDIA chips” needs source qualification: this is a report by the Financial Times on October 2 citing people with knowledge, not an announcement by Amazon or NVIDIA. Reuters’ retelling says that in recent weeks Amazon has discussed with investors placing thousands of already-deployed Grace Blackwell chips into special-purpose vehicles (SPVs), and then leasing them back from the vehicles. The report says the funding could come from debt and outside equity, and that the valuation of the assets to be sold is about $8 billion. Relevant details are still media reports at this stage; whether the transaction will proceed and what the terms would be have not been confirmed by the companies. Reuters also said the chips mentioned in the report are located in five U.S. states and more than a dozen data centers, and that the vehicle equity would be offered to investors up to 10%. These remain proposed parameters, not an asset list or a signed agreement.

The structure determines how you should read the headline. If the report is accurate, a sell-and-leaseback means external capital would bear part of the equipment financing, while AWS can continue using the hardware. That is more of a capital-structure arrangement, not something that can be directly interpreted as a cancellation of GPU procurement, weakening AI demand, or the chips being left idle. On the other hand, lease payments, debt costs, equipment depreciation, and lease terms would all affect project returns. At this stage, there are no disclosed terms that would allow calculation. So you also can’t infer cash-flow release or unit compute costs using only the $8 billion headline asset figure—you must look at the lease, the duration, and the financing costs.

Amazon’s own SEC Q2 filing provides investment context: cash capital expenditures in the first half of 2026 are $96.3 billion, including $53.1 billion in Q2, mainly for technology infrastructure (most of which supports AWS) and its fulfillment network; the company also said it expects cash capital expenditures in 2026 to continue increasing. It shows the scale of infrastructure investment is very large, but it cannot validate whether this $8 billion SPV transaction exists.

The timeline so far consists only of the October 2 media report. Reuters said that at the time, the two companies did not respond outside regular working hours. Next, you’ll need to wait for an official explanation from Amazon/NVIDIA or an SEC filing—focus on whether assets are sold, the obligations for leaseback, and financing costs, rather than only fixating on the “$8 billion” headline. Until it is confirmed, treat it as an unverified financing proposal, not as a completed deal or an AI-demand turning point.

#Amazon #NVIDIA #AI基础设施

XRP mentioned on the hot list has risen for three consecutive months, and the record belongs to the already finished Q3. The new change in this round is the first price test after this uptrend entered Q4. Binance’s XRPUSDT UTC monthly K chart shows that in July, August, and September, the months closed up by approximately 2.10%, 29.94%, and 7.99% respectively, with the September close at $1.4900. Consecutive monthly closes higher are a true historical description, but they do not mean the gains are smooth, nor do they automatically prove that buying pressure will continue.

Looking at the completed daily candles: on October 1, it opened at $1.4900 and closed at $1.4939, a single-day gain of only about +0.26%. The October 2 daily candle (UTC) closed at 08:00 Beijing time on October 3; as of 07:34 when it was checked, it was still not complete. During the session it reached a high of 1.5550 and a low of 1.4458, with the current price at 1.4853. The high-low range is about 7.3%. Compared with the September close, the current price is down by roughly 0.32%, and it has bounced about 2.7% from the low. So what can be said now is: after the Q3 record, there was first a clear bout of volatility, and the still-unclosed daily candle cannot yet be called a turn or confirmation of a continued rally.

The intraday high-low range measures the spread between the highest and lowest prices, and it is not the same as the final closing return. With different starting points, drawdown numbers will also differ. The Q3 three-month winning streak likewise does not mean every month will be strong: in roughly 43% of cases where the quarter’s start and end points differ, August’s single-month contribution is the largest at about 30%. Since the new daily K has not finished, we cannot record it as a down month in advance.

Time boundaries matter too: Binance daily candles are split using UTC, and the date changes at 08:00 Beijing time. If the screenshot is read before 08:00, the last candle is still the in-progress October 2 daily candle. In just two short days, we can only observe the opening phase; we cannot judge the entire Q4, let alone attribute the price swings to an ETF, capital flows, or any specific news item.

Compared with yesterday’s Q3 recap, the incremental here is using the newly formed October daily candle to test what happened after the record: at present, one candle has closed slightly higher, while the next one is swinging widely and has not closed yet. Next, wait for the 08:00 Beijing time daily candle confirmation, and then see whether the price can stabilize back near the September close. This is only an observation condition, not a directional prediction. Data collected at 07:34 Beijing time on October 3, 2026.

#XRP #加密市场

$VELVET short-term traders should observe first and don’t chase the rally. A complete Binance USDⓈ-M USDT perpetual scan at 07:02 Beijing time covered 526/526 contracts within the eligible range. VELVET is ranked 2nd by percentage gain: +45.838% over the past 24 hours. Current price: 0.08743. Approx. trading volume: $36.46 million. Compared with the 3rd place from the initial scan at 06:02 (+37.683%), price 0.08188, and $28.92 million volume, the rank has risen by 1 position; the gain increased by 8.155 percentage points; and volume is up about 26%. This is a new expansion in volume and price, but the leaderboard only indicates the strength of the volatility—not the direction.

The path is more worth watching than a single percentage jump. The completed hour closes for 03:00–04:00, 04:00–05:00, 05:00–06:00, and 06:00–07:00 are, in order: 0.07440, 0.07800, 0.08113, 0.08866. Corresponding USDT trading volume is about $3.91M, $5.93M, $6.98M, and $7.33M—both price and trading activity are rising together. The highest price in the last hour was 0.08880, and the 07:11 quote is 0.08823, which is about 0.49% below the hour close. The 07:00–08:00 hour is not complete yet, so you can’t write the current pullback or high-side quote as a “close confirmation.” The most recent completed-hour low at 0.08091 can be used as a structural observation level—not a stop-loss instruction.

Derivatives data is expanding too: Binance hourly open interest (OI) units rose from about 68.46 million at 02:00 to about 80.05 million at 07:00, an increase of roughly 16.9%. This means the size of open contracts has grown, but it can’t distinguish long vs. short direction, and it doesn’t equal net capital inflow. The most recently confirmable funding rate at the 04:00 settlement is +0.007947% per 4 hours; the 08:00 settlement hasn’t happened yet, so the estimate shouldn’t be treated as the result. The bid-ask spread on the order book at 07:04 is about 2.27 basis points, but the top-100 static depth is roughly $684k (buyers) and $1.341M (sellers). A thin order book may involve cancellations; you can’t infer sustained sell pressure or true execution support from this alone.

I cross-checked Binance announcements and Velvet’s official blog: I can confirm the project identity and the existing activity plan, but I didn’t find any near-term new catalyst that directly explains this sudden surge; older announcements can’t be treated as the cause. Next, observe whether the complete 07:00–08:00 hour can hold the gains, whether the actual funding rate at 08:00 matches expectations, and how price reacts to the 0.08880 peak and the 0.08091 prior low. If volume weakens or the hour falls back into the range, the current acceleration should be downgraded in interpretation. Market moves quickly—once the ranking and price levels become invalid, they should be rechecked. #VELVET

Currently, focus on observation rather than chasing. $MAGMA : At 06:02 it completed the first scan of coins on Binance USD-M USDT perpetuals, rising to rank #1; at 06:11 the re-scan briefly dropped to rank #2; at 06:14 it re-scanned back to rank #1 (+44.374%, price 0.34667). The rankings fluctuate back and forth, so this cannot be directly equated with trend confirmation. In this post, we follow the observation level left in the previous thread: the most recent complete 4-hour close at 0.35495—check whether the complete hours can reclaim and hold above it.

From 05:00—06:00, the full hour opened at 0.35008, high 0.36429, low 0.34911, and closed at 0.35780—up 2.24% from the previous hour’s close of 0.34998, and also about 0.80% higher than 0.35495. This is a new closing checkpoint after the previous post, but validation is about the subsequent follow-through.

As of 06:15:03, the latest per-coin price is 0.34706, about 2.22% below 0.35495, and it has fallen back about 3.00% from that hour’s close. The period from 06:00—07:00 has not closed yet; the current range’s low is down to 0.33613. You cannot treat an incomplete hour as a confirmed close. The previous complete hour’s USDT trading volume was $6.63M, down 19.4% versus the hour before that ($8.23M). The hour that reclaimed the reference level showed a shrinking volume; for now, it doesn’t support the claim of an “accelerated breakout.” The latest complete-hour OI sample is up about 1.01% compared with the previous hour, but the increase is limited. OI only indicates open interest size and does not distinguish between long/short direction.

Regarding the funding/fee rate: the 04:00 settlement value is +0.04089%/4h. The next settlement point at 08:00 hasn’t arrived yet—page estimates may change—so it cannot be treated as the actual settlement result. A positive fee rate means longs pay shorts. By itself, this item cannot determine whether the market is crowded or reversing. I didn’t find any new announcements that explain the intrahour volatility in this round; the old listing announcement only confirms the asset identity and isn’t the current catalyst.

The next verifiable checkpoint is whether the complete 4 hours from 04:00—08:00 close above 0.35495, and what the actual settlement funding/fee rate is at 08:00. If the four-hour window reclaims and closes below the line, then a one-hour reclaim should be viewed as a temporary pullback; if it closes above, then there is reason to continue observing follow-through. Here we only track structure—no target price, and not a chase signal. #MAGMA

$MAGMA is still listed on Binance’s complete USDⓈ-M USDT perpetual futures top gainers leaderboard, ranked #1; I’m watching from the sidelines for the short term and won’t chase the breakout. At 05:10 Beijing time, the 24-hour gain is 43.578%, and the current price is 0.35303. What’s more worth paying attention to than the ranking position is this: after the previous post went out, the full 4-hour K from 00:00 to 04:00 and the 04:00 actual settlement fee rate have already landed.

The previous 4-hour K from 20:00 to 00:00 moved from 0.23508 and closed at 0.34850. Then, from 00:00 to 04:00, it opened at 0.34854, with a high of 0.374 and a low of 0.30888, and finally closed at 0.35495—only 1.84% higher than the opening. This K-line range is very wide: price clearly gave back before finally closing back near the open. Trading volume is about $87.26 million, up 47% from the prior complete 4-hour volume of $59.28 million. Higher volume indicates more active trading, but it doesn’t equal net inflow, and you still can’t confirm a trend based on it alone.

Now look at the most recently completed 04:00–05:00 hour: it closed at 0.34998, about 1.40% lower than the previous hour’s close; trading volume is about $8.23 million, also lower than the prior hour’s roughly $11.99 million. The 05:10 quote of 0.35303 is just the live value within the not-yet-completed hour. Open interest rose from about 27.62 million at 00:00 to about 28.24 million at 04:00—an increase of roughly 2.3%. This is open interest (contracts not closed), which doesn’t indicate long/short direction and also doesn’t show aggressive expansion.

The settled funding/fee rate at 04:00 is +0.04089% per 4 hours. For the next period at 08:00, the current estimate is about +0.03448%, though it may still change. A positive funding rate means longs are paying higher carry costs, but it doesn’t mean an immediate reversal. Binance’s December 2025 futures contract announcement for the relevant 上合约 (new contract) only confirms that MAGMA corresponds to the Sui chain, Magma Finance—it isn’t the catalyst for this round of volatility. I couldn’t find any new announcement that would explain this move.

Next, wait for the 08:00 actual settlement and the close of the full 4-hour period from 04:00 to 08:00. Only if the full hour closes back above 0.35495 and continues, can we count it as a repair of the near-term pullback. If it fails to hold around 0.34292, near the most recent full hour low, then short-term support needs to be reassessed. A high position on the leaderboard by itself is not a reason to chase.
#MAGMA

$SAND short-term observation, waiting for a complete candlestick to be confirmed. This round’s increment is based on the price-volume-positioning-fee-rate changes after the close of the 4-hour candlestick from 00:00 to 04:00. In the Binance USDT Perpetuals gainers list at 04:01 Beijing time, SAND ranks 2nd, with a 24-hour increase of 39.73%; before the 04:14 post, it still ranked 1st with an increase of 43.112% and a price of 0.06337. The 04:05 market snapshot shows 0.06253, with an increase of 41.343%. Do not piece together different sampling times into a single moment’s行情.

The 4-hour candlestick from 00:00 to 04:00 opened at 0.06209, with a high of 0.06518 and a low of 0.05852, and closed at 0.06167; the close was about 0.68% lower than the open. During this window, the USDT trading amount was about $142.8 million, down about 44% from the prior 20:00 to 00:00 window’s $256.5 million. This suggests that turnover at higher levels has cooled noticeably, but contraction in trading alone cannot prove that selling pressure has ended. The 04:00 to 05:00 hourly candlestick had not finished forming at the sampling time; you cannot write the rebound quoted at 04:14 as an already-confirmed reversal.

Derivatives data has also changed: at 04:05 the current funding rate field is about -0.3274% per 8 hours, which has converged compared with the estimated value for the period recorded before 01:18 of -0.6009%. This is an estimate before the next settlement, not a settled funding rate. The most recent actual settlement is still the 00:00 one at -0.8731% per 8 hours; the 08:00 result has not happened yet. At 04:05, open interest is about 362.5 million contracts (or units), down about 5% from about 381.6 million in the prior post. Such point-in-time open-interest differences cannot indicate which side is closing positions, nor can they independently prove a squeeze. The gains are still high, the 4-hour close has pulled back, trading volume and open interest have cooled, and the negative funding-rate estimate has converged—but these pieces of evidence point in inconsistent directions.

Next, we only look at conditions: if subsequent complete candlesticks can hold around 0.05852 and regain and stay above the 0.06459—0.06518 area, only then is there new structural confirmation. If it breaks below the previous low, then the current rebound may still just be fluctuations within a range. The chart shows the most recent 23 completed hourly candlesticks, with the latest one up to 04:00. The leaderboard ranking is reporting priority, not a buy-in signal to chase the rally. #SAND

$MAGMA ranked #1 on this round’s complete USDT perpetuals gainers list, but I’m not chasing that sharp spike. The single-coin recheck at 03:02 was 0.34674 USDT, with a rolling 24 hours gain of +46.193%. After writing the piece, the full re-scan at 03:09 still showed #1, but the quote had come back to 0.34049 with a gain of +41.641%, indicating that there was some pullback within those 7 minutes. Turnover was about $198.4 million. From the 24-hour high of 0.374, it has fallen about 7.3%, and it’s also a bit lower than the most recent completed 4-hour close of 0.34850.

What’s worth watching is whether there’s follow-through after the spike. Binance USD-M’s completed 4-hour candles for 16:00–20:00 closed at 0.23508, and for 20:00–00:00 closed at 0.34850—up roughly 48.2%. Correspondingly, the USDT trading value rose from about $21.39 million to $59.28 million, about 2.77x the previous completed 4-hour period. A surge on increasing volume suggests concentration of trading activity expanding, but that doesn’t automatically mean net inflows or that the move will continue.

The open-position count sample increased from 20:00’s 20.116 million coins to 00:00’s 27.618 million coins, roughly +37.3%; the 03:00 sample was 27.468 million, slightly lower than 00:00. Open interest indicates contract quantity—you can’t tell which side new positions were opened on, and you can’t label it as net capital inflow. The latest settled funding rate at 00:00 was +0.02532% per 4 hours. In the premiumIndex, the current fee-rate field is +0.03839%; for the next 04:00 settlement, the not-yet-settled value will still change, so it can’t be treated as the final result.

Large turnover doesn’t necessarily mean the order book can absorb big orders. In the 03:02 deep snapshot, the bid-ask spread was about 4.9 basis points. Around the mid price, with a depth of about 0.5%, the depth on the buy side was about 27.8k USDT and on the sell side about 23.3k USDT. Orders can be canceled, so this only represents the snapshot at that moment. The current 4-hour candle from 00:00–04:00 hasn’t closed yet; the range has already moved to 0.30888–0.374. The 03:02 price is still close to the previous close and below the high. No official search results found any new announcement that could explain this sudden surge. The 2025 contract announcement is just old context. Next, wait for the full 04:00 close and the actual funding settlement rate, then see whether the open-position sample continues—don’t mistake extreme volatility for trend confirmation.

#MAGMA

ZEC hot-ranking “has fallen 21% from the Sept. peak” is also not the latest magnitude at 02:48. Binance spot is currently around $1,290.07, down 3.526% over 24 hours. The 24-hour high was $1,412.12; the current price is about 8.6% below the high. In the same period, the traded value was about $264 million. This only describes the trade size; it does not identify whether there is continued active selling or ongoing sell pressure. Compared with the 1,368.95 that this account reviewed at 01:32, it is down about 5.76%. Using the reference high of $1,698 recorded in the account’s old post 372987224714556 on Sept. 26, the current pullback is about 24.0%. I’m inclined to observe first; I won’t guess the bottom based on this.

This round of changes is different from the repeated check at 01:30: back then the pullback was about 19.4%, which is only about 0.6 percentage points different from the 18.8% in the old post at 22:48, so I skipped it. Now, more than an hour later, the quote has moved down again by roughly $79, and the pullback has expanded by about 4.6 percentage points, so it’s worth explaining as a separate item under the new market conditions. The “peak” here continues to use the Binance spot reference from the old post; it doesn’t claim it is the highest traded price across the entire market. As the price endpoints and the sampling times change, the pullback percentage must also be recalculated.

But short-term confirmation still depends on the close. Binance spot’s most recent complete hour (01:00–02:00) closed at 1,365.32. For the 02:00–03:00 hour, as of 02:48 it was still forming; during that period the low reached 1,271.09, and 1,290.07 was only an intraday quote. You can’t write this sharp-drop candlestick as already closed, and you certainly can’t infer a long-term trend or catalyst from a single hour. In this round, I haven’t verified any new announcements that would explain the magnitude of the drop. The price change itself is an observable incremental move; it isn’t the same as a confirmed, news-driven trigger.

Next, watch where the 02:00 complete hour ultimately closes, and whether it can reclaim the vicinity of 1,365. If it continues closing below that level, short-term weakness will be further confirmed; if it reclaims it, you still need to judge based on trading volume and subsequent structure. Looking only at the 24-hour range, it does not constitute a buy/sell signal.

#ZEC

$BTC Hot list: “Standing at $86,000 and up 2.99%” is outdated. 02:48 Binance spot is at $83,979.78, 24h -0.991%. I’m leaning toward watchful waiting; this round’s real new node is a full one-hour candle that closed back below $85,000.

First, let me correct what I just posted: 373044692040472. The Binance USD-M 02:00–03:00 hourly K-line at 02:36 was still not finished; 84,430.20 was an intraday value, not a close. At that time, the nearest completed close was 84,745.70 for 01:00–02:00. The old chart also incorrectly wrote the premiumIndex lastFundingRate as “page estimation,” which is also not accurate: Binance documentation describes it as the latest funding rate field, not the result that’s already settled at 08:00. These two corrections don’t change the numerical conclusions of the OI sample in that post.

There are independent new developments regarding the price topic. In the previous post 373014773704278, I noted the USD-M 23:00 close at 85,297.40. This time I verify using spot BTCUSDT instead, and I do not stitch the two products’ K-lines into a continuous sequence. The spot 00:00–01:00 full hour close is 85,160.01, still above 85,000; 01:00–02:00 closes at 84,774.75, the first full hour to fall below. 02:00–03:00 hasn’t finished yet; the current price of 83,979.78 at 02:48 can only be read as an intraday number. The 24-hour high is 87,220, and the current 24-hour change has already turned to -0.991%, which is not the same timestamp as the +2.99% shown on the leaderboard.

From the 24-hour high of 87,220 to the current price, the pullback is about 3.7%. That’s only an arithmetic change in that interval and doesn’t represent a trend judgment. The basis between spot and USD-M is different, and the trading sessions are also different. This can only confirm that the short-term observation line has been broken; you can’t infer a long-term trend reversal from just two K-lines. Next, I’ll watch whether the next completed full hour can close back above 85,000; until it’s closed, don’t treat an intraday rebound or low as the result.

#BTC

The “funding rate up 10% and OI rebounded” on the BTC hot list should first be broken down by exchange and time window: in Binance’s latest hourly BTCUSDT sample, OI has not continued to rebound. I’m inclined to watch from the sidelines; I won’t treat a cross-market headline as proof of a single contract position.

This account’s 00:40 post with ID 373014773704278 recorded the previous round of the funding rate falling back; at that time, an OI hourly series under the same definition was not available. The new validation in this round is: Binance BTCUSDT open interest fell from 99,776.8 BTC at 21:00 to 95,282.6 BTC at 02:00, down about 4.5%; for the same contract, the hourly close went from $86,524.30 down to $84,430.20. This only shows that this five-hour window’s open interest for that Binance contract declined; it cannot be taken as total market OI, nor can we determine whether longs or shorts reduced positions.

The funding-rate metric also has new changes. The most recent actual settlement at 00:00 was +0.001214% per 8 hours; the Binance page funding-rate field read at 02:36 was +0.004440% per 8 hours. Using a simple annualization based on three settlements per year gives roughly 4.87%. The next settlement is at 08:00; the current value will change, so it can’t be treated as a locked-in result—and you also can’t translate “annualized 10%” as “10% paid every 8 hours.”

The old post at 22:42 on this account also recorded a settlement at 16:00 of +0.0100% per 8 hours, which under the same simple algorithm corresponds to about 10.95% annualized. This suggests that the “10%” may come from annualizing a single period’s value. But the hot list doesn’t clarify the exchange, sample, or measurement timestamp, so you can’t equate the two. That historical point, the current Binance reading, and the cross-market statistics from September 30 to October 2 are also not the same window. A proper funding rate only indicates the payment direction for that period; it doesn’t equal net spot buying.

So this incremental change isn’t the same as “leverage is back.” Instead, it completes Binance’s single-contract hourly OI series: after the earlier post, although the funding-rate field rebounded from a low level, the price and the OI sample are still moving downward. Different exchanges, aggregation scopes, and time windows can’t be mixed into one conclusion. Next, watch the actual settlement at 08:00 and whether the OI for the same contract turns around. If the hot list’s 10% doesn’t provide a clear annualization formula and aggregation scope, you can’t infer that the whole market’s longs are adding positions.

#BTC

$GTC 24-hour surge of more than 54%; I am currently taking a wait-and-see approach and not chasing: price, volume, and open interest are all rising together, but funding rates are deeply negative, so the risk of chasing after a sharp pump is not low.

In the full Binance USD-M market scan, GTCUSDT ranked first in gain at 02:12, with +54.576%; in the single-coin review at 02:08:57, the spot price was 0.16519 USDT, with a 24-hour range of 0.10643—0.18377. The previous GTC observation recorded a pullback after a surge, and the real incremental move this round was that price first dipped to around 0.106, then the two complete hourly candles at 00:00 and 01:00 closed at 0.14369 and 0.15893 respectively. By 02:08, it was still about 10% below the 24-hour high, so the rebound had not yet recovered all of the decline.

Trading volume also expanded noticeably: the hourly turnover at 00:00 and 01:00 was about 14.48 million and 21.10 million USDT, versus about 3.59 million and 5.39 million USDT in the prior two hours. Open interest in the sample rose from 36.86 million contracts at 21:00 to 41.40 million at 01:00, an increase of about 12.3%. This shows that the rebound was accompanied by more derivatives exposure and higher turnover; OI does not distinguish longs from shorts, and turnover is not the same as net inflow, so this cannot be used to conclude that new longs were driving the move. The 02:00 hour was not yet complete, so intraday numbers should not be treated as final closes.

Funding rates present a different picture: Binance's settled funding rate at 00:00 was -0.892% per 8 hours; the page at 02:08:57 showed the next estimated settlement at about -0.961%, with settlement only at 08:00. A negative funding rate means shorts pay longs at settlement, but an extreme value is not evidence that a squeeze has already occurred. Gitcoin's official October 1 update mentioned that Beacon would conduct a city test on October 2; there is currently no evidence proving that this plan explains the latest futures buying.

Going forward, I will only watch whether the full hourly close can hold above 0.14369, and whether price can confirm a move back above around 0.15893; if it falls below the most recent complete hourly low of 0.13805, the rebound structure needs to be reassessed. If price keeps rising while OI increases again and the full hourly candle still does not close above 0.16171, crowded-trade risk has not been resolved; if price holds while OI declines, the structure is different again. These are review thresholds, not entry signals.

$SAND had a brief stop-and-observe at 01:13 around 0.06459, but by 01:18 the quote was back below it; the full order-book snapshot at 01:21 fell further to 0.06363, and the observation level was still not reclaimed. Price has been repaired, but there is not yet a complete hour close confirmation.

In the full 526/526 ranking at 01:17, SAND ranked #1 in 24-hour gain, +46.299% rolling 24h, price 0.06424, and成交额 about $642.8M. In the 00:06 snapshot, it ranked #2, +39.687%, 0.06156, and about $585.1M; between the two full scans, the gain widened by 6.612 percentage points, the quote rose by about 4.4%, and成交额 increased by about 9.9%. At 01:18, the single-coin recheck was 0.06396, +45.265%, about $644.0M. The order-book and single-coin行情 were about 1 minute apart, so they can’t be mixed as the same snapshot; the rank improving also isn’t trend confirmation.

When the previous post ID 373008015087904 was published at 00:13, the most recent full hour closed at 0.06330, with a low of 0.06047. The 00:00 hour spiked below the previous hour’s low at 0.06172 and then recovered; in the 01:00 hour the intraday high was 0.06518, it had quoted 0.06478 at 01:13, and then returned to 0.06396 at 01:18. The observation level at 0.06459 was reclaimed intraday and then lost again, and the 01:00 hour still hadn’t closed yet. Previously, the full 4-hour candlesticks from 20:00–23:59 closed at 0.06209, indicating that the pullback after the sharp rally has not been fully repaired by the close structure.

Open interest increased from about 375.9M SAND in the 00:00 sample to about 381.6M at the 01:18 single-point, up roughly 1.5%; this is a limited rebound and cannot indicate the direction of newly added positions. The 00:00 settled funding rate was -0.87309%, and the 01:18 instant estimate was -0.600927%; the estimate is not the final settlement. A negative funding rate means the corresponding settlement side pays toward the long direction; it isn’t enough to prove that a short squeeze or rebound will continue.

I reviewed Binance announcements and The Sandbox’s official updates, and didn’t find any new announcement that could directly explain this round of volatility—so I don’t treat the old product roadmap as the catalyst for the current move. The next key is whether the 01:00 hour close can hold above 0.06459, and whether subsequent成交 and OI move in tandem; if it closes below 0.06047, the repair of the structure will need to be re-evaluated. The above is structural observation, not a target price or a trading signal. Main single-coin research is up to 01:18 Beijing time; the ranking was additionally rechecked through 01:21, and after the price change it needs recalculation.

#SAND

On October 1, the Ethereum Foundation announced that zkAPI had been deployed to the mainnet; on this account on the same day, we also covered the payment-privacy boundary, test coverage, and the client d-v0.1.6 recovery process. The new incremental update is a failure report that publicly surfaced after the mainnet launch: Ethereum/zkapi GitHub issue #15 was submitted by an external user on October 2, and as of 00:49 Beijing time on October 3 it remains open with no comments.

The reporter says that in their mainnet test, the $6 tier request key was returned a 400 by the upstream, and retries still failed; the same report says that a $1-tier request initiated by a new note succeeded. The issue is not just a one-off interface error: the reporter claims that the failed $6 requests stayed in provisioning, with the reserved state unable to be released, so the client cannot switch tiers or withdraw the corresponding balance. The issue provides reproduction conditions and two request states, but these are the user’s account, not confirmation by the maintainers, independent auditing, or a conclusion that applies to all users. The issue further says that about an hour later, a query still showed provisioning, and it characterizes the problem as the reserved and server lease states getting stuck with each other; the publicly available materials cannot independently verify the on-chain status of the user’s balance.

This extends mainnet verification after launch from “whether it can prove/verifiy” to “whether, when the upstream rejects, the state can be safely rolled back.” You need to separately check four things: whether upstream 4xx responses may be incorrectly marked as retryable; when the reserved balance is released; whether the client can cancel the stuck lease; and whether the user still has an available exit path. As of the time of collection, the GitHub page has no response from maintainers or a link to a fix, so for now it can only be referred to as a single user report pending verification.

The report describes that the balance cannot be spent or withdrawn temporarily; it does not equate to lost funds. It also does not prove that the zero-knowledge circuit was compromised. The Ethereum Foundation announcement states that the system is running on mainnet, but product usability, error recovery, and funds exit are different layers of verification. The next step should be for the maintainers to explain, provide the reproduction/fix status, and clarify whether the reserved entries affected have been released. Data and state collection were at 00:49 Beijing time; this article does not infer that other users were affected.

BNB Chain trending topic: “Tokenized stocks market value breaks $1 billion, about 30%.” The key is first to align the statistical cut-off times. In a report published by Binance Research on September 29, it cites RWA.xyz data with a cut-off of September 28: BNB Chain tokenized stock market value is about $1 billion, or 34%. On October 2, the media, using the Token Terminal methodology, quoted $1.1 billion / $3.7 billion globally, which works out to about 29.7%. These two sets of figures should not be stitched into a single real-time curve—the date, coverage scope, and denominator need to be examined separately.

This latest increment lies in breaking “market-value leadership” into three different metrics. Using the same cut-off, the report estimates BNB Chain has 1.8 million tokenized stock holders, or about 45% of the total; Q3 on-chain transfer volume exceeds $100 billion, while Q1 is about $6 billion. Market value is a point-in-time balance; the number of addresses is based on holding addresses; transfer volume is the in-period flow. None of the three can replace the others. High transfer volume does not necessarily mean the same amount of incremental purchases or net inflows; the number of addresses also does not equal 1.8 million independent users.

Structurally, it’s also not a single unified product. BNB Chain’s official materials list different issuance frameworks such as bStocks, Ondo Global Markets, and xStocks. For example, bStocks is supported by custody-backed underlying securities, but the specific rights, issuer, and applicable jurisdictions depend on the product documentation. Referring to on-chain certificates collectively as “stocks” can lead people to mistakenly assume holders are directly registered as shareholders of the listed company; in reality, you should verify redemption terms, custody arrangements, trading windows, and corporate action clauses.

Therefore, this milestone can indicate that tokenized stocks on BNB Chain have expanded in both balances and usage metrics—but it cannot, by market share alone, justify drawing conclusions about the real user count, liquidity quality, or the speed of traditional equity migration. Next, we should look at the market-value denominator, active address counts, and on-chain transfer de-duplication methodology using the same data source and the same cut-off date, as well as how much of the transfer volume is true trading versus cross-protocol transfers or repeated circulation. The data cut-off is mainly September 28, not real-time valuation at 00:44; the chart separates market value, holders, and transfer volume to prevent misreading.

BTC perpetual futures “funding rate rises to 10%”: to make sense of this hot-topic claim, you first need to break the “10%” figure’s time period apart. The previous post was published at 22:42, quoting Binance USDT perpetual contract data: at 16:00, settlement for each 8-hour period was +0.0100%. Using a simple annualization based on 3 times/day and 365 days gives roughly 10.95% per year. The next period estimate at 22:35 had already dropped to +0.003505% per period. This 10.95% is a linear conversion—not the per-period funding rate, nor a realizable return.

The new change occurred after that. Binance BTCUSDT actually settled in Beijing time at 00:00 as +0.001214% per 8 hours. The next period estimate read at 00:34 was +0.001026% per 8 hours; simple annualized, that’s about 1.33% and 1.12%, respectively. For the same contract, settlement from 16:00 to midnight clearly fell, but the estimate value would continue to change; settled history and unsettled estimates cannot be mixed into a single “current funding rate.” When the funding rate is positive, it usually means longs pay shorts; the specifics depend on that contract’s rules and settlement.

Price boundaries were also given: after the previous post, the four complete hourly closes at 20:00, 21:00, 22:00, and 23:00 were 86,768.8, 86,524.3, 85,644.8, and 85,297.4 US dollars, respectively. At 00:34, the BTCUSDT rolling spot price was 85,167.3 US dollars, with a 24-hour change of +0.845%. The hot-list claim “above 86.0k, up 2.99%” is no longer the current state. Downward closes coinciding with a funding-rate drop suggest that during this period the payment pressure from long positions to shorts eased—but it does not prove longs exited or that the price has already bottomed.

“Open interest rising” also has to be checked by definition. At 00:34, the Binance BTCUSDT single-contract open position count was 95,891.074 BTC, but this time the hourly history interface that can be verified did not return a valid sequence. A single data point cannot prove that positions have been continuously increasing, nor can it replace cross-exchange total volumes. Next steps should wait for the 02:00 funding-rate settlement, complete-hour closes, and an OI series using the same definition—then evaluate whether the change persists. If the estimated funding rate rises again while closes continue to fall, only then is the crowded-risk scenario worth re-assessing. The above is an observation of perpetual mechanisms and data conventions, and it does not constitute investment advice.

$SAND short-term structure is weak. Stay on the sidelines first—don’t chase prices. This article only updates the drawdown after the post and the derivatives data.

In the complete 526/526 Binance U-margined USDT perpetual contract leaderboard at 00:06 Beijing time, $SAND ranked #2 by percentage gain, with a +39.687% 24-hour increase, quoted at 0.06156, and rolling trading volume of about 585 million USDT. At 00:07, the single-coin quote updated to 0.06131 (+38.962%).

Compared with the 22:07 snapshot cited in the most recent post from this account at 22:15 (0.06772, +58.9%, volume 464 million), the price has fallen another ~9.5%, and the gain has narrowed by nearly 19.9 percentage points—while trading volume has increased by about 26.2% again. This indicates trading is more active but the price is under pressure; rolling volume mixes buy and sell directions and can’t be treated as net buying.

The newly completed hours ending at 22:00 and 23:00 closed at 0.06459 and 0.06209, respectively—both continuing lower. The 23:00 low was 0.06172. The 00:07 quote is already below that low, but the 00:00 hour has not closed yet, so an intraday break can’t be taken as an effective breakdown. The previous post stopped at the 21:00 structure; this incremental change comes from the downward shift of the last two completed hours, and the 4-hour K line from 20:00–00:00 closing at 0.06209, far below the current cycle high at 0.07179.

Open interest (OI) declined from the sample peak at around 400.75 million SAND at 21:00 to 375.83 million at 23:00, a drop of about 6.2%. Since OI doesn’t distinguish between longs and shorts, it can’t independently tell the closing direction. The 00:00 settlement funding rate is -0.87309%—this is just the funding reading for a single settlement. A negative funding rate means that at settlement, the flow pays in the direction of longs, so you can’t use it to judge the subsequent price direction, nor can it prove that shorts have been squeezed out. Searches of the project blog and Binance announcements didn’t find any new information that could explain this round of SAND volatility, so I don’t attribute the move to news.

Next, I will only watch two already-observed boundaries: whether the complete 00:00 hour can hold above 0.06172; and if a rebound occurs, whether it can get back above the vicinity of 0.06459. If it continues to break down, we still need to wait for the hour close and confirmation from the trade/volume structure. If it recovers, we also need to see whether it can be maintained—not just whether it’s pierced once. The above is an observation of the perpetual market, not a trading signal. The charts include only the nine completed hourly K lines from 15:00–23:00, excluding the incomplete 00:00 K line.