Orbit Post Sitemap
SNDK｜Recovery rebound after a sharp drop, the key lies at this line
SNDKUSDT, 15-minute chart review.
After a rapid sell-off earlier, the lowest hit 793.7. Once the panic selling was exhausted, a wave of consolidation and recovery followed. The current price is 889.8, with a 24-hour increase of 6.84%.
This wave is an oversold rebound, not a direct trend reversal. The short-term moving averages MA5, MA10, and MA20 have converged, with bulls and bears battling here.
Key levels:
✅ First resistance: 901.7, the most critical current level. Whether it can hold above this will determine if the rebound can continue; the next strong resistance is the previous high at 981.7.
✅ First support: 877.7, followed by the SuperTrend defense line at 871.7. If broken, the rebound pace will slow, and the price may retest the lows.
Market outlook:
Holding above 901.7 gives bulls a chance to expand upward;
If pressure near 901 causes stagnation, a pullback is likely, with priority on testing the support strength at 877.
The storage sector is highly volatile, and the rebound is very much a battle. Focus closely on 901 as the watershed level.#非农今晚揭晓，美股大跌后市场如何押注12月加息？
Nonfarm payrolls are revealed tonight, how will the market bet on a December rate hike after the sharp drop in US stocks?
Can US stocks make a comeback tonight when the nonfarm payrolls are announced?
Last night, the Nasdaq 100 was just cut by 6.36%, the sharpest drop since March 2020, the Philadelphia Semiconductor Index down 12%, and the VIX closed above 28.
This morning before the market opens, it looks lively: Nasdaq futures up 1.1%, Micron up 8.9% pre-market, Nvidia +4.1%. But don’t be fooled by this 1%—Dow futures are still down 0.8%, money is just picking around in chips, not a full return.
Tonight at 9:30 PM, the consensus for nonfarm payrolls is only 65,000, previous value 29,000, unemployment rate 4.2%. Sounds low, right?
But ADP has already done 86,000, initial jobless claims dropped to 189,000 matching a 57-year low, and Goldman Sachs directly calls for 75,000. Here’s a counterintuitive trap: the stronger the data, the harder it is for US stocks.
Strong employment will supply ammo for a December rate hike, while the interest rate swap market’s priced-in hike probability has dropped from 80% a week ago to 68.7%. In other words, the market just priced in a "no rate hike" expectation, if nonfarm payrolls prove otherwise, this expectation will have to be reversed.
So tonight is not a simple "good data = rise, bad data = fall" multiple-choice question, but a question of which comes first: "recession trade" or "rate trade." If the data is a cold surprise, recession fears might push the chip rebound back down; if the data is hot, the rate bomb will suppress valuations. Both sides have risks. $AR
Permanent data storage addresses a problem that conventional cloud services do not always solve: long-term accessibility. Arweave approaches this through its blockchain-based permanent storage model. The interesting question is whether demand for preserving applications, websites, and digital records can translate into lasting network activity. Storage technology matters, but actual usage will determine its economic relevance.#IREN加速转型AI云，业务收入环比增长182% #ModernaQ3财报将揭示疫苗销售趋势 Key focus on two things:
✅ Actual sales of the updated COVID vaccine and overseas order deliveries, which form the current baseline;
✅ Commercial progress of the RSV vaccine, determining whether it can move beyond reliance on a single product.
COVID vaccines have shifted to seasonal administration, with unstable demand and high performance variability. Cost reduction can only improve profits in the short term; long-term value depends on the launch of new pipeline products.
This is just a review and discussion, not investment advice.On November 5, after the US stock market closed, the Nasdaq fell 5.52%, with AI-related stocks leading the decline. Coinbase and Strategy dropped nearly 10% in a single day. Bitcoin weakened in sync, falling below $76,000, hitting a new low since September, with a 24-hour drop of about 5.5%, the largest single-day decline in nearly five months; Ethereum fell even deeper, to about $2,244, down 7.6% in 24 hours. This round of decline is not due to news from the crypto circle itself, but a transmission of risk sentiment from the US stock market—this time Bitcoin is more like a highly volatile tech stock. This wave is a sentiment-driven sell-off, not a fundamental reversal, but don’t rush to bottom-fish in the short term. On Wednesday (November 4), the US spot Bitcoin ETF still had a net inflow of $74.6 million, and the Ethereum ETF had a net inflow of $53 million, just ending a continuous outflow; institutions have not completely exited. The real problem is that ETF money has come in, but the coin price was still dragged down by the US stock market, indicating that the current market leadership is in the hands of the US stock market. At 16:00 today, about $2.65 billion in Bitcoin options will expire on Deribit, with the biggest pain point at $79,000; at 20:30 tonight, the US October employment report will be released. Just watch two numbers: whether Bitcoin can hold the $75,000–$75,650 range, and how quickly it can recover the $79,000–$80,000 range. If it can’t hold, weakness will continue; a quick recovery is just an emotional shock.#非农今晚揭晓，美股大跌后市场如何押注12月加息？
I am the mid-term intelligence guy. Before tonight's non-farm payrolls, the market's main theme is clear: after the US stock market bloodbath, the probability of a 25bp rate hike in December remains around 69%–71%, but US Treasuries are already front-running a "narrowing rate hike space."
If the data is weak: rate hike expectations retreat → risk assets catch a breather,
$BTC first looks to defend 76,000, with a rebound aiming to break 80,000; $ETH's upward momentum is weaker than Bitcoin's, $SOL is recovering quickly from overselling but is the most volatile.
If the data is strong: interest rates rise, Nasdaq drops again, BTC is likely to dip around 73,000, ETH tests 2,200, SOL experiences the sharpest pullback.
My mid-term view: high interest rate pressure currently has no resolution yet. 🌙 Uncle's Night Talk｜11.06 ▍CRYPTO｜Touched 75,000 then climbed back to 76,000 BTC hit a low of $75,131 today, the lowest since mid-September, currently back near $76,100. Over 240,000 liquidations in 24 hours, with liquidation amount reaching $1.26 billion. Funding logic: Today, the crypto market is not the main character but a supporting role. It was hammered along with AI hardware during the day and then followed the rebound of US stock futures at night. It currently lacks its own independent narrative and is mostly following Nasdaq's risk appetite. Next focus: Whether $76,000 can hold overnight. If it can't hold, the weekend may continue to see tug-of-war around $75,000. ▍A-share Review｜4700 stocks down, volume shrinks by 460.2 billion: More painful than a crash is no one stepping in to buy Shanghai Composite -1.10%, Shenzhen Component -2.57%, ChiNext -3.96%, turnover 1.66 trillion yuan, down 460.2 billion yuan from yesterday. Computing power hardware is lining up on the decliners list: Cambridge Technology, Aohong Electronics hit limit down, big names like Industrial Fulian and Shengyi Technology also couldn't hold up. Counter-trend gainers are banks, power, and agriculture: Chengdu Bank +1.61%, Agricultural Bank +1.38%, power sector also had stocks hitting limit up. Funding logic: The most notable today is not the decline but the volume shrinkage. A decline with increased volume often means growing divergence and chip exchange; a decline with shrinking volume indicates reduced market participation willingness and insufficient buying support. Computing power funds are retreating, but still238,600 people liquidated in 24 hours, $1.25 billion, with longs accounting for 88%.
But the position data is moving in the opposite direction: long accounts increased by 25% in one day, short accounts decreased by 26%. The long-short ratio shifted from 58:42 to 70:30.
To translate — what's liquidated is leverage, what's entering the market is position size. Shorts are running, retail traders are catching the knife at 75,000.
The problem is 75,800 is the average cost line for short-term holders, and it has now been broken. This means there is a large group of losing traders above, and every rebound is their window to sell.
Tonight at 21:30 is the Non-Farm Payrolls, expected at 59,000, with a 68.7% chance of a rate hike. If the data is strong, we look directly at 73,000; if weak, it only gives the trapped longs a chance to escape.
The real bottom never appears when "everyone thinks it's cheap." #韩股KOSPI暴跌触发熔断，半导体集体重挫 The recent sharp drop in KOSPI triggered a trading halt mechanism, mainly due to a global AI computing power valuation reassessment combined with market structure amplifying selling pressure. Samsung and SK Hynix have very high weights, with memory chips as the core pillar of the index. HBM was previously the main theme of this AI rally, with concentrated capital bets; once expectations reverse, the downside elasticity is very strong.
The trigger came from the US midterm election results, with the market worried that the Democratic Party will tighten data center approvals and raise computing power construction costs, directly weakening the long-term demand expectations for memory chips; combined with high US Treasury yields and the upcoming non-farm payrolls tonight, foreign capital is withdrawing early from the highly elastic Korean tech stocks. Meanwhile, the sector had large prior gains, accumulating significant profit-taking positions, with concentrated Korean leveraged ETFs and retail margin financing; after the drop, this triggered a chain of forced liquidations, intensifying the sell-off.
The market is now repricing two core risks: slowing capital expenditures by major manufacturers and gradual release of memory capacity, with HBM premiums falling. However, fundamentals have not completely deteriorated; long-term AI demand remains, representing a valuation correction rather than a cyclical reversal.
Going forward, focus on tracking US Philadelphia Semiconductor Index, HBM pricing, and foreign capital trading trends. If tonight's non-farm payrolls are hawkish, pressure on Asia-Pacific tech stocks will continue; if employment cools, risk assets may see a recovery window. $BTC $ETH $SNDK *Trump media transfers 4259 BTC, not confirmed as a sell-off!*
*Core:* Trump media transferred *4259 BTC, about $344 million* to http://Crypto.com, in two transactions: 2140 + 2119 BTC.
*Key clarifications:*
- On-chain data only proves a *transfer*, not a *sale*; no official confirmation of a sell-off
- This amount is almost equal to its convertible bond collateral of *4260.73 BTC*, facing a maximum redemption pressure of about *$1 billion* on November 30
- The company still held *9477 BTC* at the end of June, not a full liquidation
*Market linkage:*
- After the midterm elections, Trump-related tokens plummeted: TRUMP -21%, MELANIA -13%, WLFI -11%
- The Philadelphia Semiconductor Index in the US stock market recorded its largest single-day drop since March 2020
- BTC ETFs saw a net outflow of about *$1.1 billion* in the past 7 days, with BTC pressured to fluctuate near 80K
*In a nutshell:* This is not a confirmed run, but a warning signal before debt maturity; the market is betting on whether it will be forced to sell coins to repay debt.
> Reminder: Avoid saying "liquidated 99.99%" as it can be misleading; transfer ≠ sale. Today's second short draft post, recommend posting soon:
Fidelity's Bitcoin ETF just received $237 million on Wednesday, but $199 million flowed out on Thursday. In one day, over $400 million went back and forth.
The whole market is following this rhythm: U.S. spot Bitcoin ETFs had a net outflow of $430 million on Thursday, with none of the 12 funds attracting money. November has only had 4 trading days so far, with a cumulative outflow of $870 million, and the $74.6 million on Wednesday was the only net inflow day this month. (Source: Farside, SoSoValue)
My view: Wednesday's "inflow" looks more like short-term funds reacting to election market conditions, not long-term buyers returning. If real buying pressure existed, it wouldn't be given back in one day. Bitcoin holding at 76,000 now relies on the breathing room after leverage clearing, not ETF support.
From now on, I only watch one thing: consecutive net inflows for two or three days before discussing whether the rebound has a foundation. Don't take single-day inflows too seriously.
#BTC现货ETF重回流入，ETH资金持续流出 $BTC NVIDIA evaporated over $440 billion in one day, is AI really about to collapse this time?
Last night, the US stock market took a heavy hit. NVIDIA plummeted 9.3% in a single day, closing at $181.60, wiping out about $446.6 billion in market value in one day.
Even worse was AMD, which dropped 17.3% straight down. Just a few days ago, everyone was discussing how strong AI chip orders were, but suddenly the entire sector got crushed.
But I think there’s something worth pondering about this round of decline.
NVIDIA’s chip demand hasn’t suddenly disappeared, and giants like Microsoft, Google, and Amazon haven’t announced stopping data center construction.
What the market worries about is something else: money is being spent too aggressively.
Several tech giants plan to invest at least $720 billion in capital expenditures this year, most of which is related to AI infrastructure. Now data center construction is also facing pressures from power, costs, and policies.
If projects get delayed, NVIDIA’s order delivery schedule could also be affected.
However, NVIDIA’s current expected price-to-earnings ratio has dropped to about 14 times, even below the S&P 500 average. Just looking at this number, it’s indeed not as scary as before.
I won’t conclude that the AI rally is over just because of one day’s plunge, but I’m also not rushing to bottom-fish. The November 17 earnings report and management’s judgment on subsequent orders and deliveries will be very critical.
If NVIDIA’s performance continues to grow but the stock price keeps falling, do you think the problem lies in valuation, or is the market starting to doubt whether AI investments can pay off?$WLFI
WLFI rose by more than 3 points. Looking at today's two pieces of news together, I still hesitate to call it a buyback rally.
On one hand, the House Oversight Committee may investigate World Liberty Financial in February 2027; on the other hand, the USD1 supply is reported to have dropped to 3.63 billion USD. Both pieces of news require follow-up details; an investigation plan is not an investigation conclusion, and changes in supply cannot be directly equated with fund withdrawals.
OKX spot is about $0.04682, 24-hour low $0.04355, high $0.04887; perpetual positions about $8.22 million, fee rate 0.005%, leverage positions are not thick. The price has returned from the low to the middle, but this range is insufficient to define the news.
I prefer to wait until the reasons for the USD1 supply change and the scope of the investigation are clear before judging which side this matter falls on. For now, just note the news and do not rush to answer for the market.
#美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 I always thought hedge funds closed because they lost money. This time it's different—Room40 Capital returned all external funds to clients and quit.
The reason is a bit harsh: the profits weren't enough to justify the risks. The boss, Paul Yablon, also mentioned the $6.3 billion Binance hack, specifically saying he didn't keep assets there.
To translate: he's not afraid of a drop; he just feels the table no longer has good cards. Then he went back to Moore Capital to do macro trading, switching venues to keep playing.
So is he too conservative, or have we, still watching the $BTC candlestick charts, gotten used to mistaking no opportunity for opportunity? $BAM
Does an increase in assets under management mean Brookfield's profits grow proportionally?
Brookfield Asset Management is entering its Friday earnings report window. Fee-paying assets, management fees, performance fees, and exit proceeds are not measured on the same basis. Total AUM growth may come from valuation changes or new capital inflows.
If fundraising converts into recurring management fees and exit proceeds support distributions, the earnings quality is better; if profits rely on a few exits or valuation adjustments, I would reduce extrapolation of recurring revenue.A fund has shut down, but not because of performance. Room40 Capital returned all external capital, keeping only its own funds and one institutional partner. Smaller in scale, it speaks more freely.
At the helm is Paul Yablon, former portfolio manager at Moore Capital. His explanation is brief: the risk-reward ratio is asymmetric, and profitable crypto opportunities are scarce. To put it plainly—he can't find a worthy table to bet on anymore.
He casually pointed to the $BNB-related $6.3 billion hack and specifically emphasized that Room40 holds no Binance assets. This is worth pondering: it’s not an excuse after losing money, but a clear disassociation before leaving.
What’s more interesting is where he’s going. Not to another crypto fund, but back to his old firm as a macro portfolio manager. The direction is clear: money isn’t flowing from Bitcoin to another coin, it’s flowing out of the entire asset class.
Judgment: one fund’s exit doesn’t define the market, but it is a sample. When professional capital sets the tone with "opportunities are scarce" rather than "prices are cheap," it’s not a valuation issue, it’s a structural one—repeated security incidents, narrowing profit paths, and a declining tolerance for volatility.
One thought to leave with: the price of $BTC is never decided by this kind of news, but its foundation writes one rule—how many institutions are willing to allocate. If this rule is slowly crossed out, it won’t drop that day, but it will gradually become harder to rise later.75200 was tested twice but not broken; will tonight's non-farm payrolls decide life or death? Yesterday I was still dreaming above 80000, but tonight I'm focused on the 75200 line, feeling quite anxious. Overnight $BTC touched 75200, a seven-week low; Nasdaq fell 5.5% in one day, Nvidia dropped 9.3%, AMD fell 17.3%, marking the worst day for US stocks since April 2025. Over $1 billion liquidated across the network in 24 hours, 90% of which were long positions.
My inclination is: 75200 might really be a critical threshold this time. It was tested once overnight and again in the afternoon at 75300, both times not broken; the shorts even took a hit in the afternoon, with nearly 877,000 liquidated shorts versus only 178,000 longs in the past 4 hours, those chasing were taken out. After a full day of decline, the price still managed to stand back at 76000, so the shorts aren't having it easy either.
On the flip side, a bounce back doesn't mean a reversal. It hovered above 76000 all afternoon, but couldn't hold 76300; if the non-farm payrolls at 21:30 tonight cause US stocks to fall again, the 75200 level might not hold a third time. If it breaks, the next stop is 73000.
So tonight, a straightforward question: do you think holding 75200 twice means a bottom and it's time to get in, or do you think this rebound is just giving shorts a position, and breaking 75200 will send it straight to 73000?
Just a market record, not investment advice
#美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 Damn! The Nikkei has officially entered a bear market 😂
US stocks AI sector crashed, and Asian chip stocks all went flat. The Nasdaq dropped about 5.5%, the S&P about 4.4%, and the Philadelphia Semiconductor index fell over 12%. The Nikkei closed down about 6%, at 57434, down more than 20% from the June high, textbook bear market. South Korea's KOSPI once fell over 8% intraday, triggering a circuit breaker.
BTC is still hovering around 76,000. The spot Bitcoin ETF saw a net outflow of about $430 million yesterday, and 24-hour futures liquidations were about $1.26 billion. US stocks are coughing, Asia is feverish, and the crypto world is shaking along.
If you didn't drop more than 5% today, come out and check in. Brothers and sisters, today Pharaoh glanced at AAOI's earnings report while holding his coffee and almost thought his presbyopia was acting up: Q3 revenue soared 128% year-over-year to $269.9 million, setting records for six consecutive quarters, with data center revenue more than doubling. Logically, with such an "AI computing power selling shovels" performance, the US stock market should be celebrating with fireworks. So what happened? After-hours stock price dropped first, falling more than 10% at one point. Pharaoh put down his coffee and broke it down for you; the impact on the future trend of the US stock market lies entirely in these three points below: 1. Even explosive revenue growth can't save the "gross margin bottleneck" life. AAOI's data center revenue is indeed strong, with visible acceleration in demand for 800G and 1.6T optical modules. But the problem lies in the profit side: gross margin stubbornly stuck around 28.2%, barely moving; GAAP operating level actually still lost $1.4 million, and the net profit on the books is propped up by "other income." In plain language: business is growing bigger, but the hard-earned profit per dollar sold hasn't increased. The US stock market's patience with AI hardware stocks is running out; capital wants "revenue up, profits up"—just relying on scale narrative is no longer enough. 2. Capital is moving from "AI infrastructure" to "value stocks," and AAOI is just a microcosm. Recent US stock market signals are clear: institutions are systematically withdrawing from capital-intensive AI infrastructure and shifting to cyclical value stocks and high-quality defensive assets. AAOI's Q4 guidance midpoint is $40.25 billion, below the disincreasing, the rate hike expectations will be However, last night's drop was more about selling on expectations. As the director analyzed earlier, if the non-farm performance is good, it might actually cool down rate hike expectations, giving the market a chance to make a V-shaped rebound. Regarding the Democratic Party's bill, it is currently just a proposal, representing a signal of tightened AI regulation and uncertainty, which has amplified risk-off sentiment. But no matter who takes office24-hour spot fund data released! Short-term bottom-fishing funds aggressively enter the market, with BTC and ETH seeing significant inflows over 6-12 hours, while the daily level still maintains an outflow pattern
📊 Fund statistics for each coin:
✅ $BTC
6 hours: net inflow of 164 million U
8 hours: net inflow of 159 million U
12 hours: net inflow of 133 million U
24 hours: net outflow of 392 million U
✅ $ETH
6 hours: net inflow of 69.8584 million U
8 hours: net inflow of 59.706 million U
12 hours: net inflow of 67.3631 million U
24 hours: net outflow of 87.3382 million U
✅ $SOL
6 hours: net inflow of 2.7036 million U
8 hours: net inflow of 3.0491 million U
12 hours: net inflow of 3.9861 million U
24 hours: net outflow of 7.8391 million U
Core analysis (based on market updates):
At the daily level, BTC and ETH overall still show outflows, and medium-term selling pressure has not been fully released.
However, short-term inflow signals are strong, with significant capital entering for bottom-fishing at the 6 to 12-hour level. SOL also shows a slight inflow, with bottom-fishing funds in the market beginning to build positions in batches.
In simple terms, it’s a process of selling while absorbing; daily selling pressure is still being released, but short-term bottom-fishing funds have already started entering to accumulate.High-Leverage Whale Shorts Exposed! $BTC , $ETH & $SOL All in Profit Another major whale's futures portfolio has been revealed, with a total position value of approximately 57.2 million $USDT. The whale is simultaneously holding short positions in $BTC, $ETH, and $SOL using high leverage. All three positions are currently showing substantial unrealized profits, while also generating additional income through funding fees. 📊 Position Breakdown 🔹 $BTC | 40X Short - Position: 269.59 BTC - PositAI massive retreat???
Friday's Taiwan stock market sell-off was quite brutal, with the weighted index closing down 4.51% at 39,701, and TSMC falling 4.69% to 2,135 TWD. The morning session was even more extreme, with the weighted index plunging 3,737 points at one time, marking the largest intraday point drop in history, hitting a low of 37,837; TSMC opened hitting the limit down at 2,020 TWD.
The trigger was the overnight heavy plunge in U.S. semiconductor stocks, with the market lowering expectations for U.S. data center expansion, causing Taiwan stocks and the Philadelphia Semiconductor Index to collapse in tandem.
After-hours policy was swiftly implemented: the National Stabilization Fund activated its market support for the 10th time, and from November 9, the short selling limit on borrowed shares was reduced from 30% to 3%, and margin requirements for short selling were raised to 130%. Foreign investors' selling pressure was fierce, with net sales exceeding 994.3 billion TWD in October, and another 524.4 billion TWD sold through November 6.
Short selling restrictions plus market support can only suppress volatility; pessimistic expectations have yet to be cleared. In the short term, Taiwan's semiconductor stocks remain bearish at high levels, U.S. market sentiment is unstable, and the Asian chip supply chain continues to face pressure and volatility. #美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 $ARB is displaying powerful bullish momentum with aggressive continuation potential.
Structure remains firmly intact as buyers dominate every minor dip.
EP
0.1650 - 0.1690
TP
0.1760
0.1840
0.1920
SL
0.1590
Buying pressure is intensifying rapidly above the crucial breakout baseline, paving the way for a sharp acceleration toward upper liquidity zones.
#OKXOrbitTopics
Let's go $ARB
#OKXOrbitTopics $AKE has been consolidating for so long; is it accumulating or waiting for retail investors to take over?
During this AKE market trend, many people in the market were eager to short every day when the price was steadily declining. When the price suddenly surged, they started doubting whether the main force was pumping to dump. The most frustrating thing is not daring to buy on the dip and not daring to chase the rise, ending up just watching the price move up in segments.
But I have never recommended blindly shorting at such levels. What really needs attention now is not just how much the price has risen, but whether funds have withdrawn and whether upcoming token unlocks will create selling pressure.
November 21 is an important date for the next AKE token unlock. The project team may attract attention through market expectations before the unlock, but unlocking does not necessarily mean a dump; the key is to observe on-chain funds and actual selling activity.
Currently, AKE is oscillating repeatedly between 0.0045 and 0.0057. According to existing on-chain data, if funds continue to flow in without significant large transfers out, this consolidation cannot be simply regarded as dumping; there is also the possibility of fund rotation and re-accumulation of chips.
My short-term strategy:
0.0043–0.0048: Focus on observing the support situation; consider gradually trying long positions after stabilization.
0.0057: Short-term breakout level; only with volume and stable hold can the upper space open.
0.0064: First target area.
#美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 $SKHYNIX $BTC
Recent capital movements entering the market require special attention. After a sell-off, the US stock market has dropped sharply.
Funds have clearly started shifting towards US Treasury bonds for safety, with the 10-year Treasury yield falling back to 4.82%, indicating that market risk appetite is cooling down.
Although expectations for interest rate hikes have decreased, funds are not rushing back into US stocks or the crypto market.
The risk-off sentiment remains.
Therefore, I tend to remain bearish on mainstream and US stocks, and I am not in a hurry to bottom-fish before the non-farm payrolls release. Any rebound should be seen as a correction first.
#US stocks suffer a sharp sell-off: Philadelphia Semiconductor Index records its largest single-day drop since March 2020 #非农今晚揭晓，美股大跌后市场如何押注12月加息？
Korean, Hong Kong, and A-shares closing market breakdown (2026.11.06): Storage giants bloodied, did someone secretly step in at the close?
Starting with Korean stocks, today’s open was a slaughterhouse. The KOSPI plunged over 8% intraday, directly triggering a circuit breaker and halting trading for 20 minutes.
Samsung Electronics closed down 6.35%, SK Hynix plummeted 10.44%, with two-day cumulative drops of 13% and 18% respectively. But note the close: the National Pension Service brought in support funds, forcibly pulling the index back from below 5000 points to close at 5059, narrowing the decline to 6%. The storage sector’s blade is still out, but there are already takers catching the falling knives.
On the Hong Kong side, the Hang Seng Index closed down 1.27% at 24,481.
The AI hardware industry chain was hit hard, with storage chips, PCBs, and optical communications all sharply down. Interestingly, northbound funds bucked the trend with a net inflow of 2.4 billion, and Li Ning led blue chips up 5% due to a generous 6 billion special dividend. Money is shifting from AI hardware into consumer and high-dividend sectors.
A-shares were the most resilient, with the Shanghai Composite down only 1.1%, but the ChiNext was dragged down nearly 4% by computing hardware.
Cambridge Technology and Aohong Electronics hit the daily limit down. Top three net inflows by main funds: power +863 million, high dividend +567 million, highways and railways +541 million. Capital is withdrawing from the AI narrative, clearly clustering in power, transportation, banking, and dividends.Tonight, the U.S. Bureau of Labor Statistics will release the October nonfarm payrolls. The market expects about 65,000 new jobs, an unemployment rate of 4.2%, and average hourly earnings up 0.3% month-over-month. Once hourly earnings reach 0.4%, even if employment meets expectations, it won't save Bitcoin. Why hourly earnings? The Federal Reserve's dual mandate is price stability and maximum employment. The October meeting held rates steady but had a hawkish tone (economic activity was upgraded from "solid" to "strong," emphasizing that inflation remains high). CME interest rate futures show a 68.7%-82% probability of a 25 basis point rate hike in December, with about 80% already priced in. In September 2026, nonfarm hourly earnings exceeded expectations, causing Bitcoin to drop 2.32% within 30 minutes. No scenario is favorable: strong employment (100,000+) strengthens rate hike expectations, bearish; weak employment (below 40,000) raises recession fears, also bearish; overheated wages (0.4%+) imply sticky inflation, the most bearish. How does overheated wages hit Bitcoin? Prolonged high rates → U.S. Treasury yields rise, increasing opportunity cost of holding cash → funds shift to yield-bearing assets → accelerated ETF outflows (BlackRock's IBIT and Fidelity's FBTC recently saw single-day net outflows exceeding $100 million) → high-leverage longs get liquidated, creating a "liquidation—decline—re-liquidation" cycle. Pricing saturation means amplification effects: with 80% priced in, marginal room is limited; data meeting expectations leads to muted reactions, while significant deviations cause sharp repricing. This was demonstrated on November 5: the three major U.S. stock indices collectively plunged (Dow -3.44%, S&P -4.36%, Nasdaq -5.52%), Bitcoin briefly fell below 76,000, 24 $DOGE has no total supply cap and issues new coins every year.
This is completely different from Bitcoin, so don't apply scarcity logic to it.
It relies on merged mining with Litecoin, so miners mine it incidentally.
Cheap, abundant, and fast transfers make it convenient for small tips.
But the coin holding concentration is high; when big holders move, the market shakes.
It’s widely listed on exchanges, has good liquidity, and is easy to chase pumps and dumps.
The community culture is quite fun, but culture alone can't support long-term price.
If you want to treat it as a piggy bank, first think clearly about inflation.
Playing is fine, but don't over-allocate and don't mistake jokes for fundamentals.When the boss shouted this time, many friends directly got liquidated.
To be honest, even the most skilled experts can make wrong judgments. Everyone's position size, leverage, and trading cycle are completely different, so you can't just copy someone else's strategy. $BTC
Take my trade yesterday as an example, I prepared two backup plans in advance:
If BTC breaks through 82000, decisively cut losses and wait for a shorting opportunity in the 84000-84800 range;
If it doesn't break 82000, hold on and take profits when available to observe.
I've shared technical indicators and methods in posts for a long time, but trading mindset can only be honed by yourself gradually.
There is no shortcut in a trading system; only through repeated trial and error and accumulation can you find a path that suits you. This path is really tough.
⚠️This does not constitute investment advice$SOXL That turnaround was way too fast
Texas data center policy signals ease, panic sentiment significantly recovers. According to market news, the Texas governor-elect has clearly stated that all currently permitted Texas data centers will be allowed to complete construction. Market participants expect this to basically cover all capacity through 2027 and likely most of 2028. Previously, concerns about new data center construction suspensions after the Democratic victory have been greatly alleviated; the AI hardware industry chain collectively rebounded pre-market, with Lam Research up over 9%, SanDisk up nearly 8%, Micron Technology up over 9%, AMD up over 8%, and TSMC up nearly 8%.Elliptic: About $50 million XRP mainly exchanged for Bitcoin
On November 6, according to The Defiant citing on-chain analysis firm Elliptic, addresses related to the Binance hack have exchanged about $50 million XRP for Bitcoin, mainly completed via THORChain, with about $226 million XRP yet to be transferred. Bitcoin obtained from other stolen asset exchanges nearly doubled within one day to 1,650 coins, with major ETH and BTC holdings still not transferred. #美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 $BTC $SNDK $SOL So it turns out that every midterm election
is paid for by the crypto community$SNDK 24-hour large short positions at 241M crushing long positions at 31M, tonight's nonfarm payroll data is expected to fall short of expectations, pushing down the December rate hike expectations a bit. Hopefully, SanDisk can make a good rally 🥹$ENB
Should Enbridge's dividend stability be judged solely by its yield?
The public calendar lists Enbridge's earnings window on November 6. The defensive nature of pipeline assets comes from contracts and cash flow; the dividend yield alone does not prove payment ability. Debt maturities, growth investments, and distribution coverage are equally important.
If distributable cash consistently covers dividends and financing arrangements are solid, the return base is relatively good; if growth mainly relies on higher leverage, the apparently high yield may include risk compensation.His Trump Media company has started liquidating Bitcoin, selling as much as 99.99% of its holdings! The company sold 4,259 BTC, equivalent to $343.67 million. The timing of this large-scale sell-off is very delicate: the Democratic Party is preparing to investigate the crypto industry and the financial connections between the Trump family and crypto assets. This batch of Bitcoin holdings once peaked at a market value of $1.4 billion in 2025, but now has been almost entirely cashed out. #美股遭大幅抛售：Looking at the xSNDK to USDT pre-market chart, the price sits at 911.0 with a daily gain of 12.26 percent after hitting a low of 794.8. Looking at old up and downs, the chart previously crashed heavily from a high near 1,714.4 down to these lower levels, showing a 30D drop of 46.51 percent. Even with today's green bounce, past trends show sharp drops usually face heavy resistance near moving averages. My best prediction of all time is that after this brief bounce, it will test lower support aga SOXL at $49, do you dare catch the falling knife?
A single-day plunge of 38.7%, crashing from 56.95 to 40.59, 500 million shares sold in panic, Korean retail investors bottom-fishing got crushed by leverage—just now, a pre-market rebound of 17% back to $49. Is this a dead cat bounce or a golden pit?
First, look at the surface: The semiconductor sector suffered its worst single-day drop since the 2020 pandemic, with the Philadelphia Semiconductor Index down over 10%, no component stock spared. But pre-market has already bounced back to 48-49, with explosive 24-hour perpetual contract volume. The candlestick tells you: RSI on the daily chart is extremely oversold and now recovering, the selling climax may be over. The question is, do you dare pick up chips on the wreckage of a leveraged ETF?
First thing: The 38.7% plunge doesn’t mean AI is dead, it means you got played by leverage.
On November 5, SOXL opened at 56.95, closed at 40.87, with an intraday low of 40.59. A 38.7% drop in one day.
Multiple triggers: Midterm elections led by Democrats, market worries about tighter AI data center regulations; UBS estimates 30% of new capacity by 2028 is at risk; Qualcomm’s earnings missed expectations; mechanical sell pressure from leveraged ETF rebalancing amplified the drop.
It’s not that semiconductor companies are failing, it’s that triple leverage got trampled in panic.
You bought 3x daily returns, a 10% drop means a 30% drop for you, 11% means 33%, plus forced selling from rebalancing—this isn’t investing, it’s riding a roller coaster without a seatbelt.
Second thing: AI demand isn’t dead, your position is.
ASML EUV equipment customers are booked through 2029-2030
TSMC and ASE advanced packaging (CoWoS) expansions are becoming new bottlenecks
Microsoft and Google hyperscalers’ AI capex remains high
Sector PE after the crash is near historical mid-to-low levels
The AI infrastructure story hasn’t changed a bit. What changed is your leverage multiple and cost basis.
Korean retail investors previously bought SOXL heavily to bottom-fish, but got crushed by leveraged rebalancing—they got the direction right but died on the tool.
This is the harshest truth about leveraged ETFs: even if the direction is right, you might not survive to the end.
Third thing: Technicals tell you this is a rebound after extreme overselling, not a confirmed reversal.
Daily chart: massive bearish candle, volume over 500 million shares, typical panic selling plus leveraged stampede. RSI extremely oversold then recovering, MACD bearish but showing bullish divergence.
Support: 40-42 (previous low + psychological level) → 35-38 (next downside target)
Resistance: 52-55 (previous intraday high) → 60 (dense volume) → 70-75 (previous platform)
$49 is a rebound zone, not a safe zone. Only a volume breakout above 55 can talk about a short-term reversal; breaking below 47 means the falling knife keeps dropping.
Bull vs. bear, you decide:
On the bullish side:
Long-term AI demand intact, ASML orders booked through 2030
Sector valuation near historical mid-to-low after crash
RSI extremely oversold then recovering, selling climax may be over
Pre-market rebound 17-20%, shorts covering pushing price up
Binance perpetual funding rate slightly positive, open interest rising
On the bearish side:
Midterm election policy risks not priced in, data center regulation looming
Leveraged ETF negative compounding effect, long-term holding in high volatility causes losses
Qualcomm earnings bomb, more earnings risks ahead
Fed hawkish bias, rate hike expectations suppress valuations
Breaking below 40 targets 30-35, falling knife has no mercy
Trading strategy
Short-term traders:
Light long positions near 49, stop loss at 46.5-47 (exit if breaks previous low structure), first target 52-55 to take half profits. Add on volume breakout above 55, target 58-60. Don’t heavy load, don’t hold through losses, leveraged ETFs don’t give second chances.
Swing traders:
Wait for daily close above 50 with volume, then enter longs, target 65-75, stop loss 40. If breaks 40 with volume, reverse to short targeting 30-35. Until confirmed, cash is also a position.
Long-term believers:
Wake up. SOXL is 3x daily leverage, negative compounding means it’s unsuitable for holding beyond 1-2 weeks. If you really believe in semiconductors, buy SMH, SOXX spot ETFs, or individual stocks like NVDA, TSM. Using leveraged ETFs for long-term is like hauling cargo with a sports car—the wrong tool crashes no matter how good the road.
SOXL now is like semiconductors in March 2020—
99% got scared by the crash, but it multiplied several times a year later. But don’t forget, how many who bottom-fished SOXL in March 2020 got shaken out along the way?
$BTC $SOL $SOXL The more you lose, the more desperate you are to break even—is it really the "market's fault"? Or is it that the moment the wallet shrinks, people's risk appetite is exposed most clearly? Recently, I've noticed a very typical sentiment: some people take the money they prepared for their children to gamble, and as $SNDK, $BTC, and $ETH all pull back together, their mindset collapses immediately. On the surface, it looks like bad luck, but it actually feels more like a signal—when stories of "life money forced into the market" start appearing around you, it often means risk appetite has already been squeezed, not still spreading. My own observation is that this round is different from the "dare to chase anything" state at the beginning of the year. The rebound rhythm of BTC and ETH is becoming more selective, altcoins' sustainability has shortened, hotspots remain, but the relay is clearly slower. This is not just simple sector rotation; it feels more like capital reprioritizing: survival first, then offense. The fact that $SNDK, a storage/semiconductor-related asset, is dragged along also shows retail investors' attention is scattered—they grab whatever is hot, lacking a main theme. There is also a bullish perspective: if BTC can hold the key range and ETH follows, the suppressed risk appetite has a chance to slowly return, with altcoins and themes feeling the warmth first. But the risk lies in that if the "life money entering the market" sentiment continues to spread, the rebound can easily become a window for reducing positions rather than the start of a new trend. What really needs watching is not a particular candlestick but whether people still dare to hold and for how long. My current judgment is: this looks more like a test of risk appetite contraction, not a full retreat, but definitely not yet... CORE project official announcement translated point by point: Which are just empty promises, which are truthful
Regarding the official announcement on the b14g theft incident, we break it down to identify: which are facts already implemented, and which are just verbal plans.
✅【True, can be trusted】
1. The total amount stolen this time is 97.05 million CORE, a hard fork was used to adjust the state once and recover 68.15 million CORE, accounting for 70.2% of the total stolen amount, verifiable on-chain.
2. The official statement "no rollback of historical transactions, no token minting, total supply unchanged" is technically valid: only the hacker's address balance was zeroed out, past transaction records were not erased.
⚠️【Unresolved, leaning towards empty promises, no evidence of implementation yet】
1. "A transparent allocation plan will be formulated later": recovered funds are only stored in the recovery address, the distribution rules for affected users and the compensation channel have not been launched yet, and the timing and proportion of distribution are all unknown. $CRCL's earnings just beat expectations, but the stock price immediately dropped about 12%, quite a contrast.
On November 5 during U.S. stock trading hours, CRCL fell about 12% at one point, Coinbase dropped about 10%, and Strategy was also sold off. BTC fell below $76,000, and ETH dropped more than 7%.
But Circle's problems are not just market panic.
Its recently announced quarterly revenue reached $865 million, which looks good, but about $182 million of that came from ARC token presales, while USDC reserve interest income actually declined by 8%.
This is interesting. The market originally expected stablecoin scale growth to bring sustained revenue, but now it is starting to reassess how much profit actually comes from repeatable business.
Meanwhile, Coinbase's third-quarter trading revenue fell 44% year-over-year, indicating that exchanges have not escaped the impact of the crypto market downturn.
I think this round of decline exposes a problem: the U.S. stock market is willing to give crypto financial companies high valuations, but once growth quality is questioned, valuation cuts can be even harsher than BTC declines.
Stablecoin and exchange businesses are both developing, but the capital market has already started demanding they demonstrate more solid profitability.
#美股遭大幅抛售：费城半导体指数创2020年3月以来最大单日跌幅 The non-farm payroll data will be released tonight at 21:30. What really kills you is usually not the direction judgment, but the execution speed in those few seconds.
A very realistic scenario: you decided to go long before the market opens, the market rallies three seconds after the data is released, but you are still entering price and quantity repeatedly in the fourth account. By the time you click confirm, the price has already moved away. The direction was right, but you didn’t make money. This is not a judgment problem, it’s operational friction.
Friction is not noticeable in normal times, only magnified in very short windows. It’s especially obvious for people with multiple accounts: the same 0.3 BTC split into two accounts means you have to cut twice, enter twice, confirm twice, each time taking hundreds of milliseconds to a few seconds. The more volatile the market, the more expensive this delay becomes.
Solving this doesn’t require stronger predictive ability, just fixing variables in advance. Accounts are grouped first, ratios set in advance, for example split 0.3 BTC into 0.1 and 0.2 BTC at a 1:2 ratio. When placing orders, only one variable remains: direction and price. When the data comes out, submit with one click, multiple accounts enter simultaneously. The same applies for the opposite direction: one-click batch close, cancel orders, adjust take profit and stop loss.
My judgment is that among people with similar win rates in this kind of market, the difference basically comes down to two things: whether the positions are pre-divided, and whether the actions are completed in one go. $BTC’s volatility has never been the problem; the problem is whether you have compressed the process into one click before it arrives.
Here’s a question worth thinking about: the last opportunity you lost because you were slow, when reviewing it afterward, did you remember it as a wrong judgment or as being too slow in operation? The 24-hour increase has reached 32.91%, but the funding rate remains negative. In the OKX 12:18 (UTC) snapshot, $ONE spot is quoted at 0.001474 USDT, only about 1.73% below the 24-hour high of 0.001500 USDT; the range low is 0.001027 USDT. The price surged to a high, yet the perpetual funding rate is about -0.0136%, showing a divergence between long and short pricing.
In the most recent full hour, the spot price rose 7.07%, with trading volume increasing by 27.34% compared to the previous period; the perpetual rose 6.60%, with trading volume nearly doubling. Over the past 24 full hours, spot and perpetual trading volumes were approximately 874,700 and 7,366,300 USDT respectively. Current open interest (OI) is about 829 million ONE, equivalent to 1.235 million USD, but a single OI snapshot cannot prove that positions are still expanding, and a negative funding rate does not necessarily mean a continued short squeeze.
If the full hour holds above 0.001447 USDT and trading volume continues to expand without the funding rate quickly turning positive, the strong structure is more credible; if it falls back below 0.001311 USDT with increased volume, I would first guard against deleveraging after the rally.
* NFA, DYOR$SNDK #US stocks suffer heavy sell-off: Philadelphia Semiconductor Index records largest single-day drop since March 2020 Democratic victory → US domestic data center approval thresholds raised, new computing power construction slows → Two branches:
1. Short term: AI server storage demand is expected to be downgraded, which is the main reason for the drop in Sandisk, Micron, and SK Hynix that day (market's first reaction: demand weakened, sell storage stocks)
2. But there is a second layer of reverse expectation (which is what you said favors price increases):
If new data centers cannot be quickly launched domestically in the US, the capital expenditure willingness of domestic manufacturers to expand local storage to support production will also cool down; coupled with the already very long NAND expansion cycle (fab construction + ramp-up takes 2–3 years), the global pace of new bit supply is delayed.
Supply release slows, demand only "slows down rather than disappears," supply-demand gap lengthens, mid-to-long-term NAND contract prices and enterprise SSD prices are more likely to remain strong and not drop quickly.
In simple terms: restricting downstream computing power expansion indirectly suppresses upstream storage manufacturers' motivation to actively expand production, supply is released more slowly, supporting storage prices. $SNDK After CORE was hacked, 68 million were "recovered" — where did the remaining 29 million go? Can your money come back?
A total of 97.05 million CORE tokens were stolen in the b14g theft incident. A hard fork recovered 68.15 million tokens, leaving about 29 million still unresolved. Many holders are most concerned about the whereabouts of these funds and the prospects for compensation.
Before the hard fork was executed, some tokens had already been transferred out by the hacker, swapped cross-chain for other assets, and left the Core chain network. These assets are beyond the scope of intervention by this hard fork and cannot be directly recovered by modifying the on-chain state. The project team has offered a bounty plan to encourage the hacker to return the assets ethically, but whether the funds can be recovered after the deadline remains uncertain; a small portion of the funds is frozen on exchanges, and the team is communicating with exchanges and law enforcement agencies to recover them.
Even the 68 million tokens already recovered are currently only stored in a recovery address. The distribution details and official compensation channels have yet to be established, so affected users cannot access the tokens for now.
This incident, combined with the controversy over the hard fork ledger change, the founder's exit from the DAO, and Bithumb's imminent delisting, has further eroded market confidence. Investors need to face reality: there is a possibility that the 29 million tokens are permanently lost. Even for the recovered portion, when and how they will be distributed remains unknown. Do not assume that "most recovered = losses are controllable"; the trust damage caused by this event far exceeds the numbers on the ledger.November 6 Crypto Evening Report
$BTC falls below 76,000, major coins collectively plunge
In terms of market structure, BTC weekly and daily charts remain in a retracement pattern; the 4-hour chart shows a rebound but continues to be suppressed by moving averages. Key resistance above is at 77,500-78,000; only a recovery there allows room for repair. Support below is at 75,000-75,150; breaking this may open deeper correction space.
Bulls remain the absolute main players
Previously, high-leverage longs accumulated above 80,000 were triggered en masse after the price broke key levels. Stop-loss orders turned into selling pressure, creating a typical cascading stampede.
Sharp single-day reversal, internal institutional conflicts
After Wednesday's midterm elections, BTC and $ETH spot ETFs recorded a combined net inflow of about $128 million, ending a streak of outflows. On Thursday, the situation reversed sharply: BTC ETF saw a single-day net outflow of $430 million, ETH ETF net outflow about $100 million, reversing the previous day's inflows.
Fidelity's FBTC just saw $237 million inflow on Wednesday, then $200 million outflow on Thursday; BlackRock's IBIT had net outflows for five consecutive trading days, totaling over $1 billion since October 30. The overall ETF volume turned positive only due to teammates filling the gap; institutional directions are not unified.
The tech stock crash triggered high-leverage longs in crypto, ETF funds reversed sharply within a day, institutional divisions are clear, and tonight's non-farm payrolls are the key node for the next directional choice.
#非农今晚揭晓，美股大跌后市场如何押注12月加息？ Chainalysis has backtracked: Initially claimed "no tampering with credentials," now fully admitted
On October 30, a week ago, Chainalysis first reported the hack, stating "no evidence that investigation content, credentials, or session tokens were tampered with." Now they have changed their statement.
On November 2, they discovered: all users' full names, usernames, password hashes, and valid login tokens were obtained by the attacker. On November 4, they added: another "man-in-the-middle attack" directly intercepted some customers' passwords.
Services have also been progressively taken down: Reactor and Data Solutions were suspended on November 2; on November 4, abnormal activity was detected, and KYT's transaction monitoring interface and API, address screening products, and login page were all taken offline. Some KYT accounts may have been compromised. Only today did they begin phased restoration of KYT and address screening; Reactor and Data Solutions remain offline. All users must reset passwords, change API keys, and whitelist trusted IPs before returning.
The company still insists one thing remains unchanged: investigation data was not accessed, and address attribution data was not tampered with. Who the attacker is and how they got in has not been disclosed yet.The ORE project team says "no rollback" — but a hard fork changes the state, is that different from a rollback?
After the b14g theft incident, the CORE official repeatedly emphasized: this hard fork did not roll back historical transactions, nor did it issue additional tokens; it was just a one-time adjustment of the on-chain state, modifying the balances of 3 hacker addresses to recover over 68 million CORE. From a technical definition perspective, the two are indeed different, but for token holders, the impact on trust is almost the same.
✅ Technical distinction
A traditional rollback means undoing transactions that have already been packaged on-chain, resetting the entire chain back to a past block, erasing all transaction records in between. However, CORE's operation this time retains all transaction records intact, only directly rewriting the final balances of specific addresses, allowing asset transfers without the address private key signature.
❌ The core pain point is the trust crisis
The foundation of public chain narratives is that private keys equal assets, and the ledger cannot be arbitrarily tampered with. Since the project team and validators can identify certain addresses as hackers and modify balances, outsiders naturally question: can this logic be applied to ordinary user addresses in the future? Who defines the boundary for asset intervention?
Coupled with the founder's exit from the DAO and the impending delisting from Bithumb, negative sentiment in the market continues to rise. Even if technically it is not a classic rollback, it has already severely damaged decentralized consensus. Many investors realize: once a public chain allows human intervention in account assets, no matter how attractive the BTC staking narrative is, the fundamental value will be shaken.MRNA two K lines consumed one-third of the entire day's volume.
At 19:00, there were 10,473 shares, which is more than ten times the 991 shares at 18:00; the price moved from 237.8 up to 261 and closed at 259. At 20:00, it further touched 276.01 and closed at 265.19, with an upper shadow of 10.8 points.
The previous thirteen 1H candles were grinding between 227.5 and 247, with an average volume of 900 shares totaling 16,000. These two candles alone traded 20,500 shares, effectively repeating thirteen hours of activity.
The box top was broken through, but the first candle gave back 40% of the gains.
The MA20 at 241.9 is 23 points below, quite far, so a pullback could be severe.
If the pullback holds between 259 and 261 without breaking and closes with volume and a bullish candle (≥1500 shares), go long with a stop loss at 254.5, target at 276.01 (2.3x), and breakout at 285 (3.7x). Note that 259 is the 19:00 close and 20:00 open.
If the rebound between 272 and 276 fails and closes with volume and a bearish candle (≥1500 shares), go short with a stop loss at 281.5, target at 258.2 (2.1x), and breakout at 241.9 (4.3x). If 258.2 breaks, both high-volume candles are fully absorbed.
At 20:00, about 10,000 shares traded in twenty minutes, similar to the entire 19:00 candle. Money hasn't stopped.
$MRNA
(Not investment advice; trade at your own risk)