Michael Fritzell (Asian Century Stocks)
RT @KerrisdaleCap: $ACMR has 4x'd since our writeup a year ago. Only 6x more to go to fulfill our report title "Fabrication of a 10-bagger". Shoutout to @TheBenSchmark from whom we got the idea and @Citrini7 for adding the rocket fuel. Teamwork makes the dream work
tweet
RT @KerrisdaleCap: $ACMR has 4x'd since our writeup a year ago. Only 6x more to go to fulfill our report title "Fabrication of a 10-bagger". Shoutout to @TheBenSchmark from whom we got the idea and @Citrini7 for adding the rocket fuel. Teamwork makes the dream work
We're long $ACMR, our largest position. Report at https://t.co/z79uQd9L89 & we explain why $ACMR could be 10-bagger. The leading wafer-cleaning toolmaker in the fast-growing Chinese semicap sector, $ACMR is a Trump trade & a massive beneficiary of the US-China chip war 1/8 tweet
X (formerly Twitter)
Kerrisdale Capital (@KerrisdaleCap) on X
We're long $ACMR, our largest position. Report at https://t.co/z79uQd9L89 & we explain why $ACMR could be 10-bagger. The leading wafer-cleaning toolmaker in the fast-growing Chinese semicap sector, $ACMR is a Trump trade & a massive beneficiary of the US…
Offshore
Photo
Michael Fritzell (Asian Century Stocks)
Barry Callebaut, Daikin and Olympus pitched here
https://t.co/f3jeZyJdRD
tweet
Barry Callebaut, Daikin and Olympus pitched here
https://t.co/f3jeZyJdRD
tweet
Offshore
Photo
Javier Blas
According to the @FT, the Trump administration now has realized its tariffs are increasing the price of everyday items (read: beer and soda cans) and it's planning to scale back some tariffs on steel and aluminium goods.
Link: https://t.co/YZh67QCaa4
tweet
According to the @FT, the Trump administration now has realized its tariffs are increasing the price of everyday items (read: beer and soda cans) and it's planning to scale back some tariffs on steel and aluminium goods.
Link: https://t.co/YZh67QCaa4
CHART OF THE DAY: Due to the impact of Trump's tariffs, the all-in cost of aluminum in America has surged to a record high well above $5,000 per metric ton (LME price + US Midwest premium).
The impact would be soon felt on anything made with aluminum -- including beer cans. https://t.co/hR2zsA0Khq tweet
Javier Blas
RT @ChristopherJM: The next round of trilateral peace talks between the US, Russia and Ukraine will be held in Geneva on February 17-18, says Kremlin spokesman Dmitry Peskov. He says the Russian delegation will be headed by Vladimir Medinsky, who annoyed the Ukrainians in Istanbul talks last spring by using his time to deliver a long-winded revisionist history lesson. President Zelenskyy just told reporters at a briefing last Friday that the Russian delegations in Abu Dhabi talks were more serious and productive, specifically because they weren't doing what Medinsky has notorious done when he's been involved. His inclusion in Geneva isn't a positive sign.
tweet
RT @ChristopherJM: The next round of trilateral peace talks between the US, Russia and Ukraine will be held in Geneva on February 17-18, says Kremlin spokesman Dmitry Peskov. He says the Russian delegation will be headed by Vladimir Medinsky, who annoyed the Ukrainians in Istanbul talks last spring by using his time to deliver a long-winded revisionist history lesson. President Zelenskyy just told reporters at a briefing last Friday that the Russian delegations in Abu Dhabi talks were more serious and productive, specifically because they weren't doing what Medinsky has notorious done when he's been involved. His inclusion in Geneva isn't a positive sign.
tweet
Offshore
Photo
NecoKronos
Why split your attention across 10 different charts to get the full picture?
From delta and liquidations to OB imbalances and CVD, this Bar Statistics indicator packs everything you need into a perfectly clean layout.
It gives you the complete narrative behind every single candle on any timeframe.
#BTC
tweet
Why split your attention across 10 different charts to get the full picture?
From delta and liquidations to OB imbalances and CVD, this Bar Statistics indicator packs everything you need into a perfectly clean layout.
It gives you the complete narrative behind every single candle on any timeframe.
#BTC
tweet
Offshore
Video
Moon Dev
I Spent $100K on Developers Before Learning This: Building a 24/7 Polymarket Bot With Claude
most people think an ai bot trading polymarket 24/7 while they sleep is a dream until they see the actual p and l data that proves they are just burning money on fees. you see the flashy titles about automated wealth but the truth is usually found in a messy csv file that most traders are too scared to look at because it reveals their edge is actually non existent. i spent hundreds of thousands of dollars on developers in the past because i thought i was not smart enough to code and that mistake cost me more than just the cash. it cost me the ability to iterate fast enough to find a winning system which is the only way you actually survive in these markets.
building a bot that places trades is the easy part and anyone can do that with a bit of help from claude or gpt. the real challenge is building the tools that tell you when to stop trading a specific strategy before it liquidates your entire account. i have been running this stink bid bot on polymarket lately and while it feels like it is working the data shows i am up exactly eight cents after a full day of activity. that might sound like a joke but being break even in a new market is actually a massive win when you consider what most manual traders are paying in hidden costs. i will show you exactly why manual trading is basically a donation to the exchange and how a simple python filter can expose your biggest losers.
the core of my system right now is a p and l tracker that i built to slice and dice my trading history based on specific keywords. i realized that just looking at a total balance change is useless because it hides the specific behaviors that are draining your wallet. for example i can filter my history for any market that includes the word bitcoin or any market that asks if a price will resolve above a certain level. by doing this i found out that my trades on market cap predictions were actually bleeding money while my other positions were keeping me afloat.
this is where the loop starts to close on why most people fail because they treat trading like a single game instead of a series of mini games. if you do not have a way to isolate your performance in sports vs crypto vs political markets you are just gambling in the dark. i found that i was down over eight dollars on just three trades involving market cap keywords which told me to immediately strip that logic out of my automated systems. most traders would have just seen a slight dip in their total balance and assumed the whole bot was broken instead of just one specific niche.
there is a massive secret about fees that nobody on wall street wants you to understand because their entire business model depends on your ignorance. when you trade by hand you are often paying three times more in fees than you would if you just automated your entries and exits through the api. i was looking at my data and realized i had fifty two thousand in volume versus seventeen thousand in a different category and the fee discrepancy was staggering. if you are not careful you are essentially walking into the apple store and offering to pay three thousand dollars for a one thousand dollar phone just because you like doing things manually.
one of my colleagues recently pointed out that using specific routes like astradex can make you eight times more profitable just by optimizing the execution layer. imagine doing the exact same trades with the exact same strategy but making eight times more money simply because you changed how the orders are sent to the blockchain. this is why i believe code is the great equalizer because it allows a solo trader sitting at home to have the same execution efficiency as a massive hedge fund. through losing money with liquidations and over trading i learned that the only way to win is to remove the human element as much as possible.
the p and l calculator i built uses python to download the portfolio csv and[...]
I Spent $100K on Developers Before Learning This: Building a 24/7 Polymarket Bot With Claude
most people think an ai bot trading polymarket 24/7 while they sleep is a dream until they see the actual p and l data that proves they are just burning money on fees. you see the flashy titles about automated wealth but the truth is usually found in a messy csv file that most traders are too scared to look at because it reveals their edge is actually non existent. i spent hundreds of thousands of dollars on developers in the past because i thought i was not smart enough to code and that mistake cost me more than just the cash. it cost me the ability to iterate fast enough to find a winning system which is the only way you actually survive in these markets.
building a bot that places trades is the easy part and anyone can do that with a bit of help from claude or gpt. the real challenge is building the tools that tell you when to stop trading a specific strategy before it liquidates your entire account. i have been running this stink bid bot on polymarket lately and while it feels like it is working the data shows i am up exactly eight cents after a full day of activity. that might sound like a joke but being break even in a new market is actually a massive win when you consider what most manual traders are paying in hidden costs. i will show you exactly why manual trading is basically a donation to the exchange and how a simple python filter can expose your biggest losers.
the core of my system right now is a p and l tracker that i built to slice and dice my trading history based on specific keywords. i realized that just looking at a total balance change is useless because it hides the specific behaviors that are draining your wallet. for example i can filter my history for any market that includes the word bitcoin or any market that asks if a price will resolve above a certain level. by doing this i found out that my trades on market cap predictions were actually bleeding money while my other positions were keeping me afloat.
this is where the loop starts to close on why most people fail because they treat trading like a single game instead of a series of mini games. if you do not have a way to isolate your performance in sports vs crypto vs political markets you are just gambling in the dark. i found that i was down over eight dollars on just three trades involving market cap keywords which told me to immediately strip that logic out of my automated systems. most traders would have just seen a slight dip in their total balance and assumed the whole bot was broken instead of just one specific niche.
there is a massive secret about fees that nobody on wall street wants you to understand because their entire business model depends on your ignorance. when you trade by hand you are often paying three times more in fees than you would if you just automated your entries and exits through the api. i was looking at my data and realized i had fifty two thousand in volume versus seventeen thousand in a different category and the fee discrepancy was staggering. if you are not careful you are essentially walking into the apple store and offering to pay three thousand dollars for a one thousand dollar phone just because you like doing things manually.
one of my colleagues recently pointed out that using specific routes like astradex can make you eight times more profitable just by optimizing the execution layer. imagine doing the exact same trades with the exact same strategy but making eight times more money simply because you changed how the orders are sent to the blockchain. this is why i believe code is the great equalizer because it allows a solo trader sitting at home to have the same execution efficiency as a massive hedge fund. through losing money with liquidations and over trading i learned that the only way to win is to remove the human element as much as possible.
the p and l calculator i built uses python to download the portfolio csv and[...]
Offshore
Moon Dev I Spent $100K on Developers Before Learning This: Building a 24/7 Polymarket Bot With Claude most people think an ai bot trading polymarket 24/7 while they sleep is a dream until they see the actual p and l data that proves they are just burning…
then runs a pre processor to filter out the noise. it can look back at the last forty eight or seventy two hours and tell me exactly which keywords are generating profit and which ones are just farming airdrops. it even handles the complex parts like checking the current bid and ask price for markets that are still open so you get a real time view of your equity. most people just look at their realized gains but you need to know the value of your open positions if you want to avoid a nasty surprise when the market closes.
i am sharing all of this code because i have an abundance mindset and i know that if i give you these tools you might find an edge that i missed. wall street thrives on scarcity and keeping their algorithms behind closed doors but that only slows down the progress for everyone else. i want you to take this python script and run it against your own polymarket data to see where you are actually losing money. you might find that your sports betting bot is actually a genius but your crypto trend follower is a disaster and that one insight could save you thousands.
to build this tool i used a specific logic flow where i gave claude the structure of the polymarket csv and asked it to create a keyword based filter. we had to iterate through a few versions because the initial export from the exchange was not giving me the full transaction history. i had to go back and manually export the data to make sure every single win and loss was accounted for including the tiny sub cent fluctuations. once the data was clean the bot could finally tell me the truth about my performance without any emotional bias.
if you are still trading by hand you are basically playing a game with one hand tied behind your back while your opponents are using high speed machinery. even if you do not want to automate the entire decision making process you should at least automate the execution to save on those 3x fee premiums. the goal is to iterate to success by trying a hundred different small ideas and using the p and l tracker to kill the ninety nine that do not work. i will keep building these tools and sharing the logic live because that is how we all get better at this game of automated finance.
tweet
i am sharing all of this code because i have an abundance mindset and i know that if i give you these tools you might find an edge that i missed. wall street thrives on scarcity and keeping their algorithms behind closed doors but that only slows down the progress for everyone else. i want you to take this python script and run it against your own polymarket data to see where you are actually losing money. you might find that your sports betting bot is actually a genius but your crypto trend follower is a disaster and that one insight could save you thousands.
to build this tool i used a specific logic flow where i gave claude the structure of the polymarket csv and asked it to create a keyword based filter. we had to iterate through a few versions because the initial export from the exchange was not giving me the full transaction history. i had to go back and manually export the data to make sure every single win and loss was accounted for including the tiny sub cent fluctuations. once the data was clean the bot could finally tell me the truth about my performance without any emotional bias.
if you are still trading by hand you are basically playing a game with one hand tied behind your back while your opponents are using high speed machinery. even if you do not want to automate the entire decision making process you should at least automate the execution to save on those 3x fee premiums. the goal is to iterate to success by trying a hundred different small ideas and using the p and l tracker to kill the ninety nine that do not work. i will keep building these tools and sharing the logic live because that is how we all get better at this game of automated finance.
tweet
Offshore
Video
Brady Long
RT @thisguyknowsai: The real AI revolution is in the group chat not the search bar.
tweet
RT @thisguyknowsai: The real AI revolution is in the group chat not the search bar.
Introducing Teamily AI: the world's first human + AI social network, built for better connection and collaboration.
This is the first time our founders @AidenChaoyangHe and @avestime speaking on camera in this way — because they genuinely wanted to share Teamily AI with you personally.
Come see what a Human + AI symbiotic social network feels like. We'd love for you to be part of this journey.
Try Teamily AI → https://t.co/7DwqxR9xND tweet
Jukan
GFHK noted in their Q&A following their February report that ARM-based CPUs have relatively weak momentum in AI servers, attributing this to lower GPU scheduling efficiency compared to x86.
They indicated that companies including NVIDIA plan to develop x86 CPU solutions in response. (It remains unclear whether this means licensing x86 architecture or co-developing x86 with Intel.)
$INTC $AMD
tweet
GFHK noted in their Q&A following their February report that ARM-based CPUs have relatively weak momentum in AI servers, attributing this to lower GPU scheduling efficiency compared to x86.
They indicated that companies including NVIDIA plan to develop x86 CPU solutions in response. (It remains unclear whether this means licensing x86 architecture or co-developing x86 with Intel.)
$INTC $AMD
tweet
Offshore
Photo
Moon Dev
openclaw with 6 claude codes
today i am giving openclaw 6 claude codes because i truly believe she can do anything we can do but better
this will all be shared inside the private zoom
see if you can still grab a ticket
join here https://t.co/JbJdIbW2p9
moondev
tweet
openclaw with 6 claude codes
today i am giving openclaw 6 claude codes because i truly believe she can do anything we can do but better
this will all be shared inside the private zoom
see if you can still grab a ticket
join here https://t.co/JbJdIbW2p9
moondev
tweet
Jukan
For me, this is pretty hard to believe. Hasn’t Google already decided to massively increase production of next-generation ARM CPUs?
tweet
For me, this is pretty hard to believe. Hasn’t Google already decided to massively increase production of next-generation ARM CPUs?
GFHK noted in their Q&A following their February report that ARM-based CPUs have relatively weak momentum in AI servers, attributing this to lower GPU scheduling efficiency compared to x86.
They indicated that companies including NVIDIA plan to develop x86 CPU solutions in response. (It remains unclear whether this means licensing x86 architecture or co-developing x86 with Intel.)
$INTC $AMD tweet
X (formerly Twitter)
Jukan (@jukan05) on X
GFHK noted in their Q&A following their February report that ARM-based CPUs have relatively weak momentum in AI servers, attributing this to lower GPU scheduling efficiency compared to x86.
They indicated that companies including NVIDIA plan to develop x86…
They indicated that companies including NVIDIA plan to develop x86…
Offshore
Photo
Benjamin Hernandez😎
Advance Auto swings back to profit as store closures streamline operations. Turnaround momentum could reshape sentiment across auto parts retailers.
$AAP $ORLY $AZO https://t.co/aG2DjDsWD3
tweet
Advance Auto swings back to profit as store closures streamline operations. Turnaround momentum could reshape sentiment across auto parts retailers.
$AAP $ORLY $AZO https://t.co/aG2DjDsWD3
tweet