Shantanu retweeted
The next war may not begin with tanks or fighter jets. It may begin with silent swarms, bird-like drones, autonomous vehicles and AI systems. The battlefield of the future will be won by algorithms. Frightening, the times to come!
Shantanu retweeted
No one becomes a doctor or lawyer overnight. Then why expect to become a successful investor in a day?
"Treat investing as a career, not a casino."
#investors #investor #investing #investingstrategy #stock #stockmarket #stockmarketindia #stocks #share #sharemarket #weekend #weekendvibes #trendingreels #trending #investment #invest
Shantanu retweeted
My piece in @livemint today.
India faces 2 choices: Indian households blow up their savings, giving F2s minimal cost exits.
RBI blows up reserves, defending the Rupee, swapping SIP Rs into USD, giving F2s low cost exits.
Our projected import ( merch+ services) cover is just 6 months. Deep red zone.
Our projected BOP deficit is ~13% ( that's optimistic IMO) of current reserves. Even in 2008, it was just 8%.
So get the magnitude of the problem.
We can't fight this 2 front war. It's a ticket to the IMF.
Sensible choice: sacrifice the Stock Market. It never busts a nation.
FX crises always do.
Remember the 80s multiple Latam Tequila crises?
My reco:
RAISE taxation on MF investing.
Lower F2 new investment taxes.
( We used to have 0 tax on F2s and 20% on locals, in the 90s. We needed dollars then. We need dollars now. )
So:
Stock market crashes.
That's fine. F2s try selling but there are few willing buyers on the other side.
Selling into illiquidity always crashes prices.
But we don't lose USDs because quantum of selling absorbed will a fraction of today.
And at 30-40% lower prices, after a crash, the same selling F2s will become buyers.
I have seen this multiple times in my sell side life. " Market is cheap now": becomes the chatter in Manhattan after some champagne. " Let's go back in"
We saw $20 billion flow in 09-10.$ 10 billion in 13-14.
Each after a massive crash.
Crashes almost always trigger massive F2 inflows.
We must make this happen.
We simply can't give easy exits.
Exits must be made costly.
Even unviable.
That's the way it used to happen before.
I am prepared to endure the pain on my India holdings for a while.
Stock markets always come back in a year or two.
No currency ever regains the glory of its pre-crisis levels. Almost none.
Absorb this fact slowly.
Since March this year, I have put ~ Rs.200 cr into India.
I am fine to take pain on my holdings if it saves the country.
Are you, the Jain, the Gupta, the Patel, and millions others, prepared to do the same?
MF distributors? Asset Managers?
Are you prepared to let go your swarth for the country?
Because else, 12 months from now, we will be teetering at abyss' edge.
Shantanu retweeted
Claude Sonnet 4.5 as a Multi-Bot Orchestrator: Architecting an Autonomous AI Trading Command Center (Full Code Included)
building an ai trading agent that actually thinks like you is the only way to survive a market designed to hunt your emotions. if you’ve ever felt like the exchange is watching your every move and waiting for you to blink before it wicks you out then you know that human intuition is usually just a disguised version of fear or greed.
we are going into the mechanics of how to build a system that takes your specific logic and executes it with the cold precision of a machine that never sleeps. the real question is how you translate that gut feeling into a block of code that doesn’t panic when the red candles start printing
most people think they need a math degree from stanford or a team of developers in new york to build something like this. i used to think the same thing and spent hundreds of thousands of dollars on developers who couldn’t deliver because they didn’t understand the actual pain of a liquidation.
it turns out that code is the great equalizer because it doesn’t care about your background or how much money you’ve lost in the past. once you realize that an ai agent is just a reasoning engine paired with data you start to see that the barrier to entry has been smashed for the average person
the first loop we need to close is how an agent can actually replicate your thought process without the bias. if you’ve ever over traded because you were trying to revenge a loss you know that your brain is your own worst enemy in the pits.
an ai agent solves this by using a thought block which is a dedicated space in the code where the model has to explain its reasoning before it acts. this is the secret sauce that separates a simple bot from a true agent that can navigate complex market structures
you feed this agent the exact same data you look at every morning like vwap or rsi and the order book depth. then instead of you staring at the screen and getting sweaty palms the agent looks at that data and writes out a paragraph of why it thinks a long or a short is appropriate.
this creates a level of transparency that you can’t get with traditional black box algorithms. if the agent makes a mistake you can go back and read its thoughts to see exactly where its logic failed and then you just iterate on the prompt to fix it
this brings us to the second loop which is the struggle of coding if you aren't a computer science genius. i learned to code live on youtube because i was tired of being a slave to my own emotions and the slow pace of hired developers.
with the tools we have now you can use ai to write the majority of the boilerplate code while you focus on the high level strategy. you don’t need to know every nuance of python to build an openclaw style agent that can connect to an exchange api and manage your risk
when you start building these systems you have to understand that the first version will likely be trash. that is part of the process because with bots you must iterate to success and every failed agent is just a lesson in what not to do.
i remember the days of getting liquidated because i thought i was smarter than the market and refused to take a stop loss. an automated system doesn't have an ego so it will cut the trade at the exact price point you told it to without hesitation
the third loop we are opening is the institutional edge and how we can actually compete with the big guys. wall street has had these systems for decades and they have used them to harvest liquidity from retail traders who are still trying to draw lines on a chart.
by building an ai agent that uses models like claude 3.5 sonnet you are giving yourself a reasoning engine that is more capable than many junior analysts at major firms. the difference is that your analyst costs about twenty dollars a month and works twenty four hours a day without complaining
the way you structure the prompt for your agent is where the true power lies. you have to talk to the ai as if it were a professional trader and give it a strict set of rules about position sizing and market conditions.
if you tell the agent to be conservative it will wait for the perfect confluence of indicators before it puts a single dollar at risk. this removes the fomo that usually leads to buying the top of a parabolic move just because everyone on social media is screaming about it
the fourth loop involves the actual integration of real time data and how to prevent the agent from hallucinating. if you give an ai outdated or messy data it will give you a messy trade signal that results in a loss.
you need to build a robust data pipeline that fetches the latest candles and the current open interest from the exchange. then you bundle this information into a clean json format that the ai can easily parse and reason over during its thought cycle
once the agent has the data and the prompt it returns a signal that your execution script can understand. this script is the bridge between the ai's brain and the exchange's api which actually carries out the orders.
this is where you have to be extremely careful with your error handling because a simple typo in your code could lead to unintended consequences. i’ve spent countless hours debugging these bridges but it is a small price to pay for the freedom of not having to watch the charts myself
the final loop to close is the transition from a manual gambler to a systematic developer. this isn't just about making money but about regaining your time and your sanity from a market that is designed to be addictive.
code is the great equalizer because it allows a regular person to build a fortress around their capital. i decided to share this journey live because i wanted to show that anyone can do this if they are willing to put in the work and iterate through the failures
now that we are here with fully automated systems trading for me i can look back at those liquidations as the fuel that drove me to learn. you don't need a massive team or a huge bankroll to start building your own trading agents today.
you just need the persistence to keep refining your systems until they reflect your best self on your best trading day. the future of the markets is automated and the only question is whether you will be the one writing the code or the one providing the liquidity for those who did
Shantanu retweeted
BIG: German Rheinmetall CEO Armin Papperger on Iran War:
If it lasts another month, I think we will nearly have no missiles available.
Shantanu retweeted
There are different ways to deal with problems.
teapawt.org/ChrisEcxel102/status/2…
Readers added context they thought people might want to know
This is AI
gijn.org/resource/guide…
Shantanu retweeted
Jo leke gaye baith, wohi to hai kamaye. ✌️
#investors #investor #investing #investingstrategy #stock #stockmarket #stockmarketindia #stocks #share #sharemarket #weekend #weekendvibes #trendingreels #trending #investment #invest
Shantanu retweeted
An Airbus 380 is on its way across the Atlantic. It flies consistently at 800 km/h at 30,000 feet, when suddenly a Eurofighter with a Tempo Mach 2 appears.
The pilot of the fighter jet slows down, flies alongside the Airbus and greets the pilot of the passenger plane by radio: "Airbus, boring flight isn’t it? Now have a look here!"
He rolls his jet on its back, accelerates, breaks through the sound barrier, rises rapidly to a dizzying height, and then swoops down almost to sea level in a breathtaking dive. He loops back next to the Airbus and asks: "Well, how was that?"
The Airbus pilot answers: "Very impressive, but watch this!"
The jet pilot watches the Airbus, but nothing happens. It continues to fly straight, at the same speed. After 15 minutes, the Airbus pilot radios, "Well, how was that?
Confused, the jet pilot asks, "What did you do?"
The AirBus pilot laughs and says: "I got up, stretched my legs, walked to the back of the aircraft to use the washroom, then got a cup of coffee and a chocolate fudge pastry."
Shantanu retweeted
To achieve amazing results trading stocks, you must do a few things:
1. Be willing to sit out in cash for extended periods and never let the averages pressure you into forcing trades that don't meet your buy criteria.
2. Keep ALL LOSSES relatively small in relation to reward.
3. When trades are working, trade aggressively.
4. When trades are not working, back off significantly.
5. Never average down a loss.
6. Never let a big gain turn into a loss.
6. Always respect risk and put discipline over ego.
minervini.com
Shantanu retweeted
Thank you, Ananya jain, good wishes, bete.
Replying to @VijayKedia1
@VijayKedia1 sir आज से एक साल पहले आपने मेरी anniversary k din mujhe follow Kiya tha 18 Feb 2023 ko,Aaj meri daughter ne apka song gaaya hai
Shantanu retweeted
This story of the legendary Stanley Druckenmiller illustrates the advantage that well prepared and risk managed youngsters have in bull markets


























