Powerpei
Powerpei
I will share some of my experiences here, welcome to follow me, I will also reply to build the OKX planet together
1.6KFollowing
1.4Kfollowers
Feed
Feed
Powerpei reposted

Compiled a list of popular stocks in various sectors of the US AI investment field for everyone.
From semiconductors, optical communications, to storage and power,
Which sectors have brought you substantial gains this year? And which ones are you currently focusing on?
For panoramic AI investment, come to Panorama Exchange!
Get all your favorite stocks in one place:

Early this morning, several old groups were all showing off their orders for Ledong Robot ()
Last night, the dark pool saw a surge of over 200 points, with screenshots of a single lot floating profit of more than 10,000 HKD flying everywhere
Seeing such an excited atmosphere in the group, I guess many people were envious
Everyone only focused on those lucky chosen ones feasting
But no one calculated how many people in that 512.7 billion HKD margin (financing) army
Not only didn’t get a single lot, but also had to endure several days of high financing interest
Retail investors are most easily brainwashed by headlines about oversubscription of 6700 times
Thinking this is an absolute no-brainer buy signal
If you calm down and do the math, you’ll understand
→→→
6700 times means the single lot winning rate was brutally compressed to a pitiful 0.8%
Those big players using tens of millions in capital to hit the top, after deducting capital costs, at best barely make a little profit from one lot
For those who got the lot, of course they’re happy, but for most retail investors, this is a meat grinder with outrageously low capital utilization
Looking at the current trend in Hong Kong stocks, the only thing giving ordinary people a chance is actually the mandatory clawback mechanism
Usually, the public market share might be just 10%, but because Ledong was oversubscribed over 100 times
Retail investors’ share was forcibly clawed back to 35%
Every time before I enter, I stare hard at this clawback threshold and calculate for a long time
➢ If this lifeline can’t be triggered, small funds going in are purely paying brokerage fees as cannon fodder, with no chance to even sit at the table
→→→
Now about the base position game
Many people subscribe to IPOs without even checking who the sponsors in the prospectus are, blindly following others (luck is indeed part of strength, haha)
Do you really think Ledong’s stability this time is all thanks to the AI concept?
Actually, it’s Haitong International, Guotai Junan and other big banks using real money as the anchor
If a new stock’s cornerstone investors are all unknown small funds
No matter how much hype the concept gets, the probability of opening below issue price is very high
The pricing power days before listing is never about faith, it’s all about the muscle of these sponsors’ factions (these small details must be taken seriously)
→→→
As for yesterday afternoon’s dark pool trading, that was the most truthful reveal
I’m used to opening Futu and Phillip’s dark pool windows on my computer simultaneously to watch
Once the dark pool trend weakens or discounts, it means big money is rushing to exit, and the first second of the next day’s open is all about clearing positions and stopping losses, no nonsense
Last night Ledong surged to 229% on Futu, the buy-side premium was a bit ridiculous, that’s the only reason I dared to hold my base position calmly today
Since subscribing to Xiaomi’s IPO in 2019 and surviving till now
I’ve stepped into the deep pit of biotech stock break-issues
And also enjoyed several waves of AI dividends
In this environment of severe liquidity differentiation, I don’t believe in luck, only in some of my reference standards: no strong narrative, no play if oversubscription is under 500 times, no overnight if dark pool doesn’t have a 50% safety cushion
Although I got this piece of meat today, the money in the market is endless, but the principal can definitely be lost
Don’t use your hard-earned money to fill those unpopular pits no one wants, understand the underlying rules before entering, it’s much more useful than any illusory screenshot of orders
(Personal real trading logic breakdown, absolutely not investment or account opening advice. DYOR.)
#HongKongIPO

Real ones remember the 2016 vibes
The industry was built in the trenches and late-night spots
not in boring corporate boardrooms
Let people enjoy the win while it lasts

Jen || OKX
tbh if you are judging people for going to e11, then i'm judging you. it's a pinnacle of miami nightlife, every major event that takes place in miami has after parties at e11.
and the problem isn't the venue, what happens there, and most certainly not the women who work there. a lot of ppl tweeting under the guise of respecting women have 0 idea what that means.
so anyways ya is it kinda funny ppl got there at 7 pm?
sure but props to the organizers. you never see crypto events getting this much press.
let people have fun. we need more of that sometimes.
anyways here's me, working while at e11 in 2016. guess we really are back ... only this time i was trenching 😭🤣

Slept until almost noon on Saturday
Habitually checked my phone to glance at my brokerage account
Last night’s plunge in the US stock market near the close
Brutally wiped out the profits I had just made earlier in the night
Watching the market while reading the news really messes with your head
Oil tankers got bombed in the Middle East
The Strait of Hormuz got tense
Risk-off sentiment dragged the whole market down hard
But if you got scared by this macro panic and sold everything
Chances are you missed the most blatant capital migration on last night’s market
➢➢➢
I reviewed yesterday’s individual stocks and found the main logic in the AI sector has completely shifted
A few months ago everyone blindly bought computing power and chips
But last night Arm dropped 10%
Broadcom and AMD also collectively pulled back
Instead, cloud security company Datadog, after releasing some good earnings news, surged nearly 40% in a revenge rally
Wall Street money is more opportunistic than anyone
This is actually a very clear signal of a market shift:
Big money is totally fed up with the basic infrastructure stories that just pile up computing power and build hardware lumps
Now whoever can turn AI into real money on the software side is where the cash flows
Switching from hardware hype to application implementation
That’s the core reason behind last night’s counter-trend surge
➢➢➢
The Hong Kong market is basically the same story
Semiconductor leaders like SMIC and Hua Hong are pulling back and diverging
But when Duan Yongping said he bought Pop Mart
The toy stocks immediately heated up
My personal view is that this is not a broad bull market at all
It’s all existing funds desperately searching for short-term safe havens amid extreme anxiety
In this crazy market with geopolitical chaos and rapid sector rotation
Trying to guess the market bottom is just making yourself miserable.
I plan to use the weekend market closure to clear out most of the stocks still selling AI hardware stories based on PPT slides
>Tencent and Alibaba earnings reports are coming next week, so tech stock volatility will definitely be pulled around
These days, trading by following where the money wants cash and realized profits is way more reliable than watching Middle East news to guess oil prices.
Wash up, go downstairs for brunch, force myself offline for the weekend, no trading talk.
#USStocks #HKStocks #AI

Powerpei
AMD's Q1 earnings report last night was like throwing a fire into a haystack
Revenue broke through $10.3 billion, with data center business surging 57%
This extremely violent liquidity not only forcibly pushed NVIDIA back to a $5 trillion market cap
The cross-market spillover effect is astonishingly fast
Today, Hong Kong stocks like SMIC and Hua Hong Semiconductor directly soared
Southbound funds swept over 910 billion in buying volume
Confirming that capital is frantically searching for semiconductor undervalued spots across the entire network
Seeing the screen full of surges, many people emotionally shout that the bull market is back
If we strip away this noise and look at the underlying data
The foundation of this rally is actually very solid.
Macroscopically, the ceasefire news in the Middle East directly hammered oil prices down by more than ten percent
The shadow of inflation rebound has been temporarily forcibly dispelled
Looking at the micro industry side, capital expenditure in the AI inference era has become an open card
Now the big players are competing not just on GPUs
Severe capacity mismatch between CPUs and HBM memory is triggering real supply-demand panic in the supply chain.
Capital is pouring down this hardest industrial logic chain, leaving no room for shorts.
But the purer this logic is, and when everyone thinks "buying means making money"
The risk level on the market is even higher
When fundamental benefits are quickly overdrawn in a very short time
Crowded trading channels inevitably plant the seeds for a sharp high-level shakeout
A trend of this magnitude is indeed worth using a small position to ride the premium
But heavy buying at this point is purely like putting your neck on the chopping block
My current move is to only keep the core leaders' base positions
Firmly hold over 30% cash exposure
Let the frenzy money be earned by others
When this wave of sentiment is exhausted and a real structural correction is hammered out
The cash in hand will be the only proof to hold chips at the bottom.
#HongKongStocks #USStocks #AIsemiconductors

I just read a paper from Silicon Valley
It said that the iteration of large models is about to hit an invisible wall
The reason is quite funny: it's not a shortage of Nvidia GPUs, but that the standard text data available on the internet for crawling has basically been exhausted by giants like OpenAI and Google
What these large models are most envious of now
Are those non-standard, accented, and even mixed with local slang "live human voices" from the physical world
Without this grassroots data
No matter how powerful the AI is, when it comes to Asia or emerging markets
It’s still a useless tool that can’t understand human speech
Following this line of thought, I suddenly understood what @psdnai’s recently launched Numo project is up to
They are collecting voice data in small languages like Bengali, Hindi, and Tamil
On the surface, this work is dirty and tiring, extremely "grassroots"
But the commercial intuition is extremely sharp
Big companies want to sell AI globally but are struggling to buy this high-quality local data
Numo is basically inserting a straw right into the biggest anxiety gap of these giants.
For ordinary players like us, it’s meaningless to understand the underlying code of large models
But using this data hunger to earn some profit makes sense
Record some audio, contribute some data, and get rewards
If you happen to be a core user of the old Poseidon App, you can even get a free multiplier on your earnings
Instead of taking over those overvalued shell AI coins in the secondary market
It’s better to do this most fundamental water-selling business
After all, big companies will always lack real data, and the buyers can’t escape paying for it.
Note: Not a promotion, not investment advice, be cautious of risks! DROY







