Crypto KOL Community & Market Trends

The biggest illusion over these past two days is—that the bottom has finally been hit.
The thing I most want to do is—to get back to break even fast, and get out of the squeeze fast.
My biggest wish is—to go back to a month ago.
The most desperate thing is—that none of the three things above can be achieved.
😂

BTC worth $64,900—what are you waiting for?

First glance: the rate-hike blade is hanging overhead, and big money is hesitating.

After bouncing from 58k in July to 66.7k, it fell back; the current price is barely holding on in the 63,000-67k range. Up 1.4% in 1 day, down 1.9% in 1 week, up 7.7% in 1 month. It’s trading below the 50-day EMA, far under the 200-day moving average at 71k. With macro pressure, weak demand, and unclear direction, everyone is waiting for a clear signal.

First thing: Is FOMC “no more hikes” bullish? Don’t be naive.

On July 29, the Fed held rates unchanged with a 9-3 vote—pausing hikes is indeed good news, but look closely: three regional Fed presidents publicly voted against it and demanded a 25bp hike.

Since Waller took office, the emphasis on fighting inflation has been even stronger. The June dot plot already revised up end-of-year rate expectations. The market says out loud that “pausing is bullish,” but in its heart it knows: the probability of another hike at the September meeting isn’t small.

After the news was released, BTC surged briefly and then quickly stabilized—no violent push up, suggesting “smart money” doesn’t view this as a signal of a trend reversal. Oil prices are still rising, and inflation could easily return.

Second thing: ETF flows are running—this is the most honest signal.

In June, ETF net outflows were about $4.0-$4.5 billion, setting a historical record. Mid-July saw a brief inflow, but from July 23-28 it flipped back to consecutive net outflows of hundreds of millions. Cumulatively since 2026, net outflows are about $5.0-$6.0 billion.

In the short term, funding rates have cooled, new opening interest has retreated from highs, and on-chain active capital inflows have stalled—clearly thinning demand.

Third thing: a technical signal has emerged that must be watched.

On the daily timeframe, after the July rebound topped out around 66.7k, it immediately pulled back—confirming that rebounds below the 200-day MA (71k) are just “dead cat bounces.” RSI is below 50, trading volume keeps shrinking, and MACD is neutral-to-weak.

More unsettling is this: the 63,000-67k range has been consolidating for nearly a month. This kind of “high-range consolidation + shrinking volume” is usually a precursor to trend selection.

The battle of bulls vs bears—judge for yourself

One side is:

FOMC pauses hikes; the “hanging chad” is temporarily off the table

Higher share of long-term holders, lower exchange reserves—tightening on the supply side

Whales show signs of recent accumulation

Spot ETF cumulative net positive about $51.0 billion; the long-term narrative remains unbroken

The other side is:

Three officials voted against and demanded hikes; uncertainty still hangs over the September meeting

June ETF net outflows of $4.0-$4.5 billion; repeated net outflows in July

Price is below the 50-day EMA, far under the 200-day moving average

Volume is fading, funding rates cooling, demand thinning

Down 45% year-over-year from ATH 126k

Key levels

Overhead resistance: 65,000-65,200 → 65,500-66,700 (July high + supply zone) → 68,000-68,500

Below support: 64,300-64,500 → 63,500-63,800 → 62,000-62,500 → 60,000 (psychological line)

For short-term traders:

Light long on a pullback to 64,300-64,500; stop loss below 63,000; target 65,500-66,700. If it rebounds to 65,500-66,700 and stalls, you can lightly try a short; stop loss above 67,000; target 64,500-63,500.

For swing players:

If price breaks out above 66,700 with volume and then pulls back to confirm, chase longs looking toward 68,000+. If it effectively breaks below 63,500 with volume, follow the short looking toward 62,000-60,000.

For long-term believers:

Accumulate via DCA in batches below 60,000. 2026 is the year when ETFs first undergo a clear bear-market pressure test—while the long-term thesis (digital gold, anti-censorship, AI/sovereign reserves) hasn’t broken, short-term suffering is hard to avoid.

BTC right now is like itself back in June 2022—

Everyone asks, “Is this time different?” Turns out long-term holders are quietly counting their chips.

If 63,500 holds, you’ll regret not building your position in batches in this range.

But if 63,500 can’t hold, don’t blame the market for not giving you a chance to escape.

What’s your BTC cost basis?

At 64,900—are you adding or cutting? #美联储维持利率不变 #USD1持币生息最高8% #三星电子涨4.5%领涨韩股 $BTC $SOL $ETH

Tune in for a live stream tonight to watch the PCE and GDP data together.
Let’s give July a perfect finish.
$XAU $BTC
{spot}(BTCUSDT)
{future}(XAUUSDT)

Huge 🔥

I asked a friend
Stocks lost two million
How do you sleep?
He said he sleeps like a baby
Sleep a bit
Then you wake up and cry for a long time😭

$BTC is holding firm near the $64.5K support zone after the Fed maintained interest rates with no major policy changes.

At the same time, U.S. spot Bitcoin ETFs continue seeing net inflows, reflecting steady institutional interest.

The current range remains intact, with a break above $65.5K or below $63.5K likely to set the tone for the next move.#FedHoldsRatesSteady #GateCardUpTo8%Cashback #EthereumInstitutionalClosesInitialFunding

ETH worth $1,919—what are you hesitating about?

First, look at the surface: major bullish news, yet the price isn’t moving up.

Yesterday, the FOMC kept rates unchanged with a 9-3 vote, easing immediate tightness fears; ETH bounced on the news. Morgan Stanley launched an ETF trust that includes staking yield, with net inflows of $5.15 million on day one. A mega whale withdrew 40k ETH from exchanges, and about 430k large transfers followed—by all logic, this string of positives should have pushed ETH to 2,000.

But what’s reality? ETH is just ranging at 1,900-1,920, like a lion asleep. Over the past 24 hours, it’s only up 0.2%-0.7%, barely even catching its breath.

First thing: the FOMC paused rate hikes, but a “pause” doesn’t mean a “turn.”

A 9-3 vote kept rates unchanged, and 3 people still insisted on hiking. The statement is full of hard lines like “inflation remains high” and “geopolitical risks.” The market is trading this as a “pause,” but the Fed is literally telling you: don’t get too excited—there’s a chance September keeps coming with hikes.

Historical pattern: every first rebound after an FOMC “pause” is a false move. The real trend only shows up once the CPI data confirms it about two weeks later.

Second thing: institutional action is coming fast—this is the real signal.

Morgan Stanley launched MSSE Ethereum Trust—includes staking yield with extremely low fees; net inflows were $5.15 million on day one. BitMine keeps adding, with total holdings now at 5.79 million ETH, most of which is already staked. A mega whale withdrew 40k ETH from CEX, and roughly 430k large transfers appeared on-chain. Staking-included ETFs solve the hard flaw of traditional ETFs having “no yield.”

Third thing: a technical signal has appeared that you must take seriously.

Price repeatedly grinds in the 1,900-1,970 range, forming an ascending channel. RSI is around 55-58, neutral; MACD is near the zero axis, with flat momentum. This isn’t a continuation of a downtrend—it’s accumulation, with bulls waiting for the fuse.

But short-term pressure is heavy around 1,925-1,940, and 1,960-1,970 is the swing high. Above 2,000 is a psychological line, and the 200SMA near 2,100+ looks like a wall.

Either there’s a volume-backed breakout above 1,970 that opens the door to 2,100, or a break below 1,850 that declares the rebound is over.

Bull-bear duel—judge for yourself.

One side is:

- The FOMC pauses rate hikes; the risk-asset breathing space opens
- Morgan Stanley’s staking-included ETF launches; the institutional entry channel is wider
- A whale pulls 40k ETH; large whale transfers on-chain lean bullish
- Multiple times holding near 1,900; support at the lower channel edge is valid
- Staking ratio above 30%, circulating supply continues to be locked

The other side is:

- YTD: down from 4,950 to 1,919, a 61% drop; trend still leans bearish
- High rates + sticky inflation; September hike odds aren’t low
- Pressure at 1,925-1,970; momentum is insufficient
- The 200SMA long-term moving average suppresses, and the “death cross” backdrop hasn’t changed

Key levels

- Overhead resistance: 1,925-1,940 → 1,960-1,970 → 2,000 → 2,100+
- Support: 1,900-1,905 → 1,875-1,885 → 1,850-1,860 → 1,730-1,800

For short-term traders:

- Buy in batches on pullbacks to 1,875-1,905; stop loss 1,850-1,865; targets 1,970-2,000
- If it breaks above 1,970, add on; watch 2,100-2,150

For swing traders:

- Wait for a daily close above 1,970 before entering on the right side; target 2,100-2,150
- If there’s a valid breakdown below 1,850 with volume, exit unconditionally; then wait for 1,730-1,800

For long-term believers:

- DCA in batches below 1,900. It’s down 61% from ATH; staking yield is 3-4%; institutional staking-included ETFs have just started; the RWA scale of tens of billions remains in the ETH ecosystem.

ETH is like BTC in 2023 right now—

Everyone was complaining, “Why isn’t it up after institutions come in?” After the ETFs passed, it doubled directly.

The day it breaks 1,970, you’ll realize:

It’s not that ETH is bad—it’s that you got scared out by the 61% drawdown.

What’s your ETH cost basis?

At 1,919, are you adding or cutting? #美联储维持利率不变 #USD1持币生息最高8% #三星电子涨4.5%领涨韩股 $BTC $ETH $SOL

I have ten thousand idle funds in hand—what should I buy for the best value for money?
1. $SUI
directly competes with the existence of SOL. The top one is that the Foundation has a big-picture mindset: it can pull up the ecosystem, and after three years it can be above 10.
2. $ONDO
has the most imaginative RWA track seed candidate. Everything is in place—it's just waiting for a round of super FOMO, aiming for double-digit numbers.
3. The top five exchange platform tokens are related to the dollar rate-cut cycle. In one bull market, the minimum is starting at 2x.

If you have ten thousand idle dollars, what should you buy for the best cost performance?
1. $SUI : Directly targets SOL. The best part is that the Foundation has big-picture strategy; it can pull the ecosystem up, and in three years it could be above 10.
2. $ONDO : Has the most imaginative seed candidate in the RWA track. Everything is ready—only waiting for a wave of super FOMO; look for double-digit gains.
3. The top five exchange platform tokens: related to the dollar rate-cutting cycle; in one bull market, the floor is at least 2x.

If you have 10,000 in spare money, what should you buy with the best value for money?
1. $SUI : Directly positioned against SOL. The biggest strength is that the Foundation has a large enough vision—once the ecosystem can be built up, it could be above 10 after three years.
2. $ONDO : The most imagination-filled seed player in the RWA track. Everything is ready—just waiting for a wave of super FOMO. Look for double-digit upside.
3. Top five exchange platform tokens: This is related to the dollar interest-rate cut cycle. In one bull market cycle, it starts at a minimum of 2x.

Damn, the target they picked is pretty good—too bad they didn’t pick the right chain. Even though BSC’s jacket managed to get the coin dividend increase by going through Mars’ Alpha, I still feel uneasy about it.

Is Gold XAU worth keeping your nerve at $4,085?

First, take a look at the surface: the news is chaotic, and both bulls and bears get killed.

Last night’s FOMC held interest rates steady with a 9-3 vote. Surprisingly, three members voted against, demanding a rate hike. Waller went hawkish: inflation isn’t solved yet, and further tightening can’t be ruled out. Gold surged first to 4,116, then got smashed to 4,042. It was a roller coaster of $74 up and down—both sides blew up and liquidations hit everywhere. From the 5,600 historical high in January, gold’s year-to-date drop is already over 25%.

First thing: not raising rates is supposed to be bullish—so why did gold fall instead of rising?

Retail logic: no rate hike = the dollar drops = gold rises, and people pile in mindlessly.

But last night’s FOMC message was extremely twisted: a 9-3 vote means serious division within the FOMC—3 votes support a rate hike, and that hawkish “no” camp is the most hawkish opposition lineup in years. Waller’s mouth is harder than Powell’s: “If inflation picks back up, I will not hesitate to raise rates.”

The market has already repriced the probability of a September rate hike to 50-60%.

Rate-hike expectations are like a blade hanging over gold’s head. As long as this blade doesn’t drop, gold can’t climb comfortably.

Second thing: when the U.S.-Iran war breaks out, why can’t gold be lifted either?

In theory, when the cannons fire, gold soars. As the U.S.-Iran conflict escalates and the oil price rockets higher week over week, risk-off sentiment explodes—gold should fly.

But in reality—oil price spikes push up inflation, and once inflation rises, the Federal Reserve has to hike. When tightening tightens, gold has to die.

When the cannons fire, in rates—gold falls. Traditional safe-haven logic has been completely distorted in the high-interest-rate era. Gold is now a “hostage” to inflation; the higher oil prices go, the more panicked it gets.

Third thing: an extremely dangerous technical signal has appeared.

On July 27, gold hit 4,116 and fell back. After the FOMC yesterday, it rushed to 4,116 and fell back again. Today it’s once more approaching 4,110 and gets smashed. Three fake breakouts, three times slaughtering the longs.

This is a textbook “triple-top resistance”—once it’s confirmed, the impact is enormous.

Gold is currently a trapped beast under suppression from rate-hike expectations. If it wants to break out, it must truly hold above 4,116 on volume. Otherwise, it will likely pull back to 4,000 or even 3,960.

Bulls vs bears—you decide.

On one side:

Global central banks net bought 244 tons of gold in Q1, and they’re still continuing to buy after the quarter

Escalation in the U.S.-Iran conflict provides a safe-haven backstop

The dollar and U.S. Treasury yields weakened in the short term after the FOMC

If today’s GDP/PCE data is soft, it directly ignites the longs

On the other side:

4,116 was rejected three times, and the probability of triple-top resistance being confirmed has jumped

Three committee members voted for a rate hike, and hawkish forces are rising within the FOMC

Oil prices surge and push up inflation; September rate-hike expectations are still heating up

Gold is still down over 25% this year, and the trend hasn’t reversed yet

Key levels

Top resistance: 4110-4116 → 4130-4166 → 4200

Bottom support: 4065-4040 → 4000 → 3960-3959

Short-term traders:

① Pull back to 4040-4065 to stabilize and go long, stop loss below 4000, target 4110-4120; ② If it again tests 4110-4116 and prints a clear rejection candlestick (long upper wick, heavy-volume bearish candle), short directly, stop loss above 4125, target 4045-4040; if it breaks down, watch 4000.

Swing traders:

A big-volume hold above 4116-4120 and a 4H close confirming it → chase longs, target 4160-4200. A valid breakdown below 4000 with volume confirmation → switch to shorts, target 3960-3900. M

Long-term believers:

Dollar-cost average in batches below 4000. The global central-bank gold-buying logic stays the same. Even by year-end, institutional targets still look for 4500+. But remember—gold is for survival, not for betting your life.

Gold right now is like Bitcoin in October 2022—

Everyone was cursing that “a safe-haven asset isn’t acting like one,” and it turned out that was BTC’s final shakeout at 16,000.

4116 is the “Normandy” for gold longs.

If it breaks through, it’s 4,500. If it can’t, it’s 3,600.

Do you think 4116 can make it through this time?

Tonight’s GDP and PCE data—are you betting long or betting short?#美联储维持利率不变 #USD1持币生息最高8% #三星电子涨4.5%领涨韩股 $BTC $ETH $XAU

Alpha went for it on Mars, but didn’t make money, and didn’t dare to gamble. Still, the good news is that it can unlock the BSC maximum. If Mars goes to 30m, then every day there will be 3m of that day’s “golden puppy” to play with. If Mars goes to 50m, then every day there will be 5m to play with. Missing out is fine too—unlocking the BSC maximum is good for everyone. The village is distributing gold bars, so we can head back to the village.

120% APY on USDC. 👀
That’s the kind of staking opportunity that gets attention.
Sl8’s event is live for a limited time if you’re holding USDC, worth checking out.
Join now :

$54 HYPE — do you cut your loss or buy the dip?

First, take a look at the surface: a dense barrage of bearish news is coming in waves, and the bulls are scattered.

June 16 ATH was 76.85; today it’s 54. In just over a month, it’s down 30%. Within 7 days, 6.93 million HYPE tokens have been de-locked from vesting, worth $415 million. Tomorrow alone, 3.3 million tokens will be released, approaching $200 million. HIP-3-related RWA suddenly crashed, and $57.4 million instantly evaporated. The daily chart breaks below all short-term moving averages, and the MACD keeps making a bearish crossover—either $52 finds a floor, or it directly collapses.

First thing: tomorrow there will be $200 million worth of de-locking, but you might get scared stiff.

Tomorrow’s 3.3 million HYPE de-lock will be worth nearly $200 million. Retail investors hear that and panic: “Run, the whales are about to dump!”

But look at the historical data—whenever HYPE has a large-scale unlock, the price actually rises. Because unlock ≠ dump; what matters is whether holders choose to sell.

Second thing: the RWA flash crash wipes out $57.4 million in liquidations, but it exposes an overlooked bullish factor.

HIP-3-related SK Hynix perpetual prices were abnormally priced, and $57.4 million got liquidated. Trade xyz stepped in and said there is “partial backstop,” and the market panicked and dumped.

RWA trading volume has already surpassed crypto perpetuals, becoming HYPE’s #1 source of income. The more diverse the products, the faster the flywheel turns—97% of protocol fees are used to buy back and burn HYPE, and the cumulative burned amount is already quite significant.

A Grayscale research report just came out: based on expected net profit of $1 billion in 2027, HYPE at $54 corresponds to a forward PE of 15–18x. Traditional finance-tech stocks usually trade at 30–50x—so who’s expensive and who’s cheap?

Third thing: a technical signal has appeared that you must watch closely.

$53–$54 is the most critical support zone, and the final line of defense of the uptrend line since January.

RSI is recovering slightly from the oversold zone, and the MACD histogram is starting to narrow. The direction is about to be chosen—either a breakout surge or a sudden crash.

Bull vs bear — you decide.

On one side:

$200 million unlocked tomorrow, but the historical pattern is “down before unlock, up after unlock”

Grayscale valuation at 15x PE, far below traditional finance-tech

97% of revenue used to buy back and burn — a hard-core deflation mechanism

RWA + prediction markets + spot, and the product lineup keeps expanding

On the other side:

3.3 million tokens de-locked in a single day tomorrow, and the sell-pressure expectation is maxed out

Institutions transferring to exchanges hurts short-term confidence

Daily chart breaks below all moving averages; the technicals are bearish

The broader market is ranging, and alts overall are under pressure

Key levels

Resistance: 55–56 → 58–60 → 62–65 → 70+

Support: 53–54 → 50–52 → 48

For short-term traders:

Volatility will be extremely high around tomorrow’s unlock. If it holds $54 and recaptures 55+ on increased volume → go long with a small position, stop-loss at 52.5, target 58–60. If it breaks below 53 with heavy volume → avoid longs, stay on the sidelines, and reconsider at 50–52.

For swing traders:

Wait for a sell-off-stabilization signal in the 53–50 range, then build positions in batches, stop-loss at 48, target 60–65 → 70+.

For long-term believers:

DCA with eyes closed below 53. The three big logics—income flywheel + buyback and burn + product expansion—haven’t changed. The 2027 target is still valuation re-rating. But remember—if you can’t even handle the panic of tomorrow’s unlock, don’t call it a long-term hold.

Right now, HYPE is like SOL in 2023—

Everyone is cursing “unlock dumping” and “FDV is too high,” and the price still went from 20 to 200.

At $54, is it a meat grinder or a money-printing machine?

The answer isn’t in the candlestick chart—it’s in your mindset and position management.

What’s your HYPE cost basis?

After tomorrow’s unlock, will you buy the dip or cut your loss?#美联储维持利率不变 #USD1持币生息最高8% #三星电子涨4.5%领涨韩股 $ETH $HYPE $BTC

Password’s here
I’ll continue doing duo gold $XAU
and get the tonight’s PCE data release
The prior expectation is that the PCE data will be below or in line with expectations, which is bullish for the market
It’s best to enter in batches before the data
See whether gold can break through 4120 in one go tonight

Password is here
Going long on gold $XAU
Get tonight’s PCE data release
The expected ahead of time is good news.
It’s best to enter in batches before the data.
Let’s see whether tonight can break above 4120 in one move
{future}(XAUUSDT)

@furan86999 One UTA gets it all done—lazy people rejoice!

@Cryptoxq8 Try un-staking TRX and see how it goes with your wave-stacking.

Micron Plunges 11% in a Single Day: Is This the “GPU Before the Mining Crash,” or “the Last Panic Before Getting In”?

Brothers, it’s another night of rivers of blood.

The Philadelphia Semiconductor Index (SOX) has slid from a high of over 14,600 points a month ago to just over 10,400 points now. On July 17, it officially fell below the 20% threshold, confirming entry into a technical bear market.

Micron Technology closed down 10.9% on Wednesday, marking its biggest single-day drop in recent times. Starting from its historical closing high of $1,213.56 on June 25, it has cumulatively fallen by about 30%, and its market cap has evaporated by about $414.95 billion.

SanDisk has been cut in half over a month, and SK Hynix’s ADR breaks the previous level. The entire memory sector looks no different from the crypto market’s big-cap plunge.

Now everyone is asking the same question:

Is this the “GPU before the mining crash,” or “the last panic before getting in”?

Don’t rush to pick a side. Let’s see how the bulls and bears tear into each other—

🟢 Bull logic: This is a bad overreaction—a deep V after SOL blew up following the FTX collapse

First, Micron’s fundamentals haven’t broken down; if anything, they’re insanely strong.

Micron’s FY2026 Q3 revenue was $41.46 billion, up 346% year over year, setting a record high. Net profit was $28.24 billion, up 14x year over year. Gross margin was 84.9%—higher than Nvidia’s 75% and Meta’s 81.9%, the highest among large-cap tech companies in the U.S.

The company signed 16 strategic customer agreements covering $100 billion in long-term investment. Management expects supply constraints to last through 2027. Q4 guidance calls for revenue of $49 billion to $51 billion, while the market expects only $43.2 billion.

Earnings aren’t just strong—they’re historically strong.

Second, the demand logic for HBM from AI hasn’t broken.

HBM’s value share in AI servers has surged from 7% in 2021 to more than 25% of GPU costs. HBM production in 2026 is expected to grow 103% year over year, and then rise another 77% in 2027. Samsung expects server DRAM, enterprise SSDs, and HBM demand to accelerate in the second half of the year, and the overall market will still be in short supply.

AI’s hunger for storage is like miners’ hunger for compute—logic hasn’t broken; it’s just market panic.

Third, analysts are still going crazy bullish.

The median target price from 54 analysts is $1,600. KeyBanc reiterated an “overweight” position with a $1,750 target. The average target price is $1,507, implying more than 100% upside from the current share price.

The bulls say: this sharp drop is emotion being purged, not a collapse in fundamentals. If you don’t buy the dip now, are you waiting to chase after it bounces back?

🔴 Bear logic: Be careful! This is the first big bearish candle after LUNA’s collapse

First, memory demand could be peaking.

Morgan Stanley issued a warning on July 21: the AI storage frenzy is nearing a turning point, and memory contract prices are expected to top out in the fourth quarter. The earnings-upgrade momentum has fallen from a 92% peak to 77%. Year-over-year growth in DRAM contract prices has already come off from the cyclical highs.

If prices can’t keep rising, earnings growth will slow down. What the market fears most is not earnings declining, but earnings growth declining.

Second, China’s DRAM giants are coming.

On July 27, Chinese ChangXin Memory Technologies (CXMT) listed on A-shares. It surged more than 460% on the first day, and its market cap broke above 3 trillion. The market is worried that after ChangXin Memory Technologies receives massive financing, it will expand production capacity, ramp up technology R&D, and even challenge the high-end HBM market.

The global DRAM supply landscape could change fundamentally. How long can Micron, SK Hynix, and Samsung’s oligopoly position hold?

Third, deleveraging in the Korean market is far from over.

Since the beginning of the year, Korean retail investors have been aggressively levering up to bet on Samsung and SK Hynix. The balance of credit financing surged from 27.4 trillion Korean won in early January to a peak of 38.6 trillion Korean won on June 24.

Then the trend reversed.

Leveraged ETF mechanisms force rebalancing—“add when it rises, cut when it falls.” When markets rise, it acts like a booster; when markets fall, it becomes a meat grinder—margin calls → forced liquidation → further declines → more margin calls. Once a negative-feedback spiral starts, it doesn’t shift based on anyone’s will.

Credit financing balances fell to 32.7 trillion Korean won on July 23, but the decline wasn’t nearly enough. Forced liquidations are still continuing.

The bears say: a sharp drop doesn’t necessarily mean the bottom. Highly leveraged funds haven’t been fully liquidated yet. If you buy now, be careful not to catch it halfway up the mountain.

🤔 My take:

This round of storage selloff is the result of three forces stacking together.

First, profit-taking. Micron has risen more than 700% over the past year, and SK Hynix and Samsung have risen even more wildly. Any asset that’s up 7x—once even a little wind blows, money will run.

Second, cycle fear. The memory industry is always a strong cyclical sector. When it rises, it rises crazily; when it falls, it falls crazily. Morgan Stanley says prices will top in Q4, and the market chooses to run early.

Third, deleveraging stampede. Korean retail investors’ leveraged positions are being forced to clear. This isn’t pricing based on fundamentals—it’s passive selling. What’s the characteristic of passive selling? It doesn’t care about reason or valuation; it only wants to close positions.

So is it the bottom, or just halfway up the mountain?

My answer is: in the short term, there may still be volatility, but in the medium to long term, the bulls’ logic is stronger.

Why?

Because HBM isn’t ordinary DRAM—it’s the “compute-and-memory” for the AI era. Nvidia’s GPUs can’t do without HBM, just like mining rigs can’t do without graphics cards. As long as AI training and inference continue, HBM demand won’t stop.

Samsung expects supply gaps to expand quarter by quarter in the second half of 2026, and new capacity releases are generally delayed until after 2027. If the supply-demand gap is still widening, how can prices collapse?

In the short term, ChangXin Memory Technologies entering the high-end HBM market still faces high technical barriers, and the structure of the global AI storage market is unlikely to change fundamentally in the near term.

But the problem is—right now the market doesn’t talk logic, it only talks sentiment.

Korean deleveraging hasn’t finished, and forced liquidations are still going on. Until this “bleeding” stops, any dip-buying could be catching falling knives.

📌 So my strategy is:

Watch three signals—

Signal one: Korean credit financing balances stabilize (currently 32.7 trillion; the peak was 38.6 trillion—still needs to keep falling)

Signal two: the amount of forced liquidations falls back to normal levels (currently still at the several-hundred-billion-KRW level)

Signal three: the SOX index stops falling and stabilizes (currently 10,400 points—getting farther and farther from the threshold)

Until these three signals appear, keep your hands off.

Once the signals appear—

Micron’s current share price still corresponds to analysts’ target prices with more than a 100% upside. A company with an 85% gross margin, 346% revenue growth, and $100 billion in long-term agreements—after it has dropped 30%—do you think it’s expensive or cheap?#美联储维持利率不变 #USD1持币生息最高8% #三星电子涨4.5%领涨韩股 $MU $SKHY $SNDK