Ahr999 has fallen back below the 0.45 bottom-fishing line. It’s cheap in the long run, but too many people are trying to catch a falling knife right now.
Ahr999 is at 0.44 today, down from 0.49 two days ago.
Retail accounts across all four exchanges are heavily bullish: Binance +51.9%, Bitget +57.5%.
But large traders on Gate are net bearish at -16.6%, and large traders on OKX are also bearish at -5.7%.
Bitfinex has about 90,557 BTC in long positions, while short positions are down to just 125 BTC. Leverage is almost entirely stacked on one side.
The indicators say it’s cheap; the positioning data says the market hasn’t finished shaking out yet. Long-term investors can buy in gradually, but in the short term, wait for retail longs to cool off. If 76,000 doesn’t hold, another leg down is possible.
$BTC Gave back 6% in a day, and last night's dream of breaking through 80,000 has been completely dashed.
Binance spot fell from 80,648 to 75,776 in 24 hours, hitting an intraday low of 75,208.
$ETH fell even harder, down 8.2% over 24 hours—more than two percentage points steeper than BTC's decline.
When altcoins fall harder than Bitcoin like this, it usually means money is pulling out, not simply changing hands.
My take: in the short term, watch whether 75,200 holds. If it does, that would be a breather after the sharp drop. If it breaks, don't rush to buy the dip—wait until it gets back above 78,000 before talking about a rebound.
$XAUT Gold is already expensive and has even reached +2σ. Why is analyst Benjamin Cowen still bullish?
When many people see gold prices surge and logarithmic regression reach +2σ, their gut reaction is to wonder: Is the bull market over? In his framework, the answer is usually: this rally has overshot and needs to consolidate; it doesn’t mean the entire bull market is over.
Since 2020, he has treated logarithmic regression as fair value over roughly 50 years: below the line means undervalued, above it means overvalued. He expected this cycle to become severely overvalued in 2025–26, then gradually return toward fair value. So reaching +2σ around 5,600 marks an extreme extension of this parabolic move. He has also stated explicitly: even in overvalued territory, “that doesn’t mean gold can’t rise further.” Overvaluation can persist for a long time; expensive ≠ topped out.
He has since more often drawn comparisons with the kind of rhythm seen in the 1970s: 1. First leg up → retest the regression band 2. Second major rally (the one that just ended, surging to around 5,600) 3. A major correction in between (mostly in 2026, around 20–30%, forming a higher low) 4. Another leg up, lasting toward the end of this decade (often citing a target of 6,000+) First digest the extreme extension, then talk about the next wave—that’s how the sequence makes sense.
Logarithmic regression (current price around 4,120; fair value around 1,741, approximately +1.82σ)
Key level: $BNB . If it holds above the hack-day low of 680, look for a move back to 722. A break below means the market hasn't fully absorbed the news.
Have your assets moved this week? Comment “Moved” or “Stayed put.”
The scariest Halloween mask is the one you wear for yourself.
Tomorrow is Halloween. The streets will be full of vampires and pumpkin heads. Everyone knows they’re fake, so it’s fun.
Trading masks are trickier, because even you believe them.
You say, “I’m in it for the long term,” but one drop and you sell. You say, “I’m just testing with a small position,” but then you keep adding more. You say, “I have a stop-loss,” but it only exists in your head.
Wear a mask long enough, and usually you’re the one who gets scared in the end.
This weekend, why not take it off and have a look: does your $BTC position still match the plan you started with?
Which mask do you wear most often? Comment “fake long-term” or “fake small position.”
Key levels: If $SKHYB holds, watch for a move to challenge the previous high at 158.5. If it falls below 150, don’t rush to chase.
The memory shortage is now expected to last through 2030. Those still sitting on the sidelines waiting to buy the dip after a cyclical crash may have to wait until retirement to get on board.
Image: 4300streetcar/Wikimedia Commons, CC BY 4.0 #SK海力士 #記憶體 #AI
Key level: if $TSLAB holds, watch for a challenge of 398, the previous high; if it breaks below 390, don’t rush to buy yet.
There isn’t even a shadow of merger terms yet, but the stock price has already eaten up half of the premium. If you rush in now, be careful not to become the bagholder helping others exit.
Image: Maurizio Pesce/Wikimedia Commons, CC BY 2.0 #特斯拉 #SpaceX #USStocks
Retail traders are piling in long, whales are relatively cold, and $BTC is fluctuating just above this week’s opening price of 80,157.
At 20:19 tonight, data shows retail accounts across four exchanges are all leaning long, with Bitget the most bullish at +44.7%.
Large-holder positions are mixed: Binance +25.8%, Gate -10.9%, and OKX -4.0%.
The whale-to-retail gap is Binance -0.52 and OKX -1.10, showing large holders are clearly cooler than retail traders.
Bitfinex leveraged long positions stand at 89,713 BTC, while short positions are only 173 BTC.
Funding rates every 8 hours are OKX +0.0098% and Gate +0.0070%, so chasing the price still isn’t overly hot.
Ahr999 is 0.49, still in the dollar-cost averaging zone, only 0.04 away from the 0.45 dip-buy line.
On-chain data (10/25): short-term holder cost basis is 75,704, and the True Market Mean is 77,771.
STH MVRV is 1.06, meaning short-term holders are making only about 6% on average.
$BTC 80,390 (Binance spot, 10/26 20:38 Taipei)
Key level: if $BTC holds 79,220, look for a rebound toward 81,070; if it breaks below 79,220, don’t catch the falling knife yet.
Retail long positions are lined up longer than a night market queue, while large holders are sitting back and watching the show. Jumping in now could easily make you exit as bagholders.
Breaking: Musk warned on X today that the AI industry is headed for a massive amount of “stranded compute.”
He said that compute will be stuck inside data centers that are only half-built.
In another post, he added that if even one item is missing from the construction checklist, the compute won’t run.
GPU rental rates, however, keep getting more expensive. a16z cited Ornn data showing that every model costs more to rent than it did at the start of the year.
$NVDAB ’s H200 rental rate is +143% above its price under straight-line depreciation over 6 years; the B200 is +111% higher.
Even the old A100, a card from 6 years ago, rents for +9% more.
Apollo chief economist Slok points out that AI has generated zero financial returns outside the tech industry.
For the 493 S&P 500 companies excluding the Magnificent Seven, revenue per employee and profit margins are flat.
AI cloud stocks are taking the first hit: $CRWVB fell from 65.08 at the 10/20 close to 54.39, down 16.4%.
Key levels: If $NVDAB holds 210.5, look for a rebound to 222.8. If it falls below 210.5, don’t try to catch the falling knife.
Old GPU rental rates are rising more aggressively than Taipei landlords’ rents, while those chasing AI stocks are the ones who’ll move into half-built data centers’ server-room studio apartments first.
Image: Carl Lender/Wikimedia Commons, CC BY 2.0 #AI #輝達 #DataCenters
Key levels: If $SPYB holds above 759, look for a rebound to 775. If it breaks below 759, don't try to catch a falling knife.
Oil prices are going through the roof, and being stuck holding losing stocks is bad enough—now you even have to check the candlestick chart before filling up your scooter.
Image: Quintin Soloviev/Wikimedia Commons, CC BY 4.0 #美股 #油價 #CrudeOil
$GOOGLB 4 days down nearly 7%—what is everyone fleeing before earnings? $GOOGLB Current price: 321.92 (24h -0.61%)
Closed at 344.97 on 10/19; now it’s down to 321.92 Today’s 24h low of 320.69 is the lowest in nearly 30 days
Next Wednesday, 10/28, after the U.S. market closes, Alphabet and Microsoft report earnings on the same night Alphabet doesn’t provide quarterly guidance, so the market is watching Google Cloud most closely Last quarter, Cloud revenue grew 82% year over year to $24.8 billion
Its rival that same night: Microsoft itself forecasts Azure growth of around 45% $MSFTB Current price: 524.40 (24h +1.43%)
The next day, 10/29, it’s Amazon’s turn $AMZNB Current price: 248.88 (24h -0.53%)
If $GOOGLB holds above 320.69, it could rebound to 333.49 If it breaks below 320.69, it could hit a new low before earnings
Which of the three cloud giants do you think will report the strongest earnings? 👇
$METAB Revenue hits a record high, but the stock has fallen from 748 to 681. What is the market afraid of? $METAB Current price: 680.88 (24h: -1.72%)
Meta's Q3 revenue was $64.6 billion, up 26% year over year AI capital spending was $34.9 billion—more than operating cash flow of $32.8 billion Free cash flow turned negative at -$2.12 billion, a rare occurrence Meanwhile, headcount fell 15% year over year as money shifted from people to machines
To raise funds without diluting Zuck's voting power, Meta will issue non-voting Class C shares, with one share issued for every share held starting next January
Most of this money is going toward AI servers and data centers $NVDAB Current price: 217.49 (24h: -1.33%)
$METAB If it holds today's low of 674, it could retest 716 A break below 674 could mean a new short-term low
Burning cash to get ahead in AI—visionary or bubble? 👇