# Charandeep Kapoor > Personal portfolio of Charandeep Kapoor — AI Product Manager at Delta Exchange, India's largest crypto derivatives exchange. Creator of Drishti (live LLM trading signals for crypto perpetuals, real capital) and Stocky (Claude-driven Zerodha trading, +110% verified). Founder of Timelock Trade (oracle-less, liquidation-free DeFi derivatives). IIT Kanpur. Essays on markets, trading psychology, AI, and building systems. ## About Charandeep Kapoor is an AI Product Manager at Delta Exchange, where he builds perpetual futures and AI-driven trading products. He blends math and finance to ship products and quant strategies at the intersection of models and markets — from protocol design to Claude-driven bots that trade real capital. Track record: built Stocky, a fine-tuned Claude system trading the Indian stock market through a custom Zerodha MCP server — ₹16.57L profit (+110% ROI) on ₹15L capital over ~9 months, Sharpe 2.29, 73% win rate, verified P&L on Sensibull. Founded Timelock Trade: oracle-less, liquidation-free "Protected Perps" — $7.3M volume, $2M TVL, 1,000+ users on Monad testnet. Earlier: 0→1 DeFi product at Diffusion Labs (Methlab: 20,000+ users, $50M TVL), Stader Labs, Heru Finance, Tykhe Ventures. B.Tech IIT Kanpur, JEE Advanced AIR 638, National Maths Olympiad AIR 3. Trading since age 16. ## Main Pages - [Home](https://charandeepkapoor.com/): Portfolio — live systems, track record, experience - [Proof of Work](https://charandeepkapoor.com/work): Everything shipped — live products and trading systems on real capital - [Writings](https://charandeepkapoor.com/blog): Essays on crypto, quant trading, protected perps, DeFi, markets, AI, life philosophy - [Delta Exchange](https://charandeepkapoor.com/delta): Current role — AI Product Manager - [Indian Stock Market](https://charandeepkapoor.com/markets): The Stocky story — Claude-driven Zerodha trading, verified - [Résumé](https://charandeepkapoor.com/resume): Experience, skills, education, certifications - [Live Track Record](https://charandeepkapoor.com/track-record): Live Dhan trading P&L, Sharpe, drawdown, win rate — updated daily ## Products & Systems ### Drishti — live AI trading signals (crypto perpetuals) Live LLM-driven trading signals for Delta Exchange crypto perpetuals. A 15-minute, regime-aware cycle across 8 perpetual markets with 30+ engineered features per asset (CVD, IV term structure, basis, global long/short, OI by exchange, Coinbase premium). A live executor places real, risk-managed orders (reduce-only stop-loss/take-profit brackets, fixed sizing, guardian process, kill-switches). - Case study: https://charandeepkapoor.com/work/drishti - Live dashboard: https://drishti-beryl.vercel.app ### Stocky — Claude-driven Zerodha trading (verified +110%) Fine-tuned Claude connected to Zerodha through a custom MCP server, trading the Indian stock market on real capital. ₹15L → ₹31.57L (+110%, ₹16.57L profit) over ~9 months. Sharpe 2.29, 73% win rate. Verified P&L: https://web.sensibull.com/verified-pnl/imported-hare/longterm-pnl - Full story: https://charandeepkapoor.com/markets ### Timelock Trade — oracle-less DeFi derivatives (founder) "Protected Perps": traders only realize profit, LPs are paid a premium to bear the downside. Options priced with Black-Scholes adapted for crypto, implied volatility derived from Uniswap data, theta approximated via Dirac distributions — no external price oracle. $7.3M volume, $2M TVL, 1,000+ users on Monad testnet. Led a team of 6. - Case study: https://charandeepkapoor.com/work/timelock - Perps: https://perps.timelock.trade · Docs: https://docs.timelock.trade/docs ### Lakshay — intraday AI signals for NSE F&O The NSE sibling of Drishti: 5-minute regime-aware cycles across NIFTY, BANKNIFTY, and front-month equity futures. Features from Dhan data (ORB, VWAP, supertrend, options layer, India VIX). Paper track record first; real orders double-gated. - Case study: https://charandeepkapoor.com/work/lakshay - Live dashboard: https://lakshya-cyan.vercel.app ### Second Brain — local-first memory for Claude Code (open source) Captures every Claude Code session into an Obsidian-compatible Markdown vault on your own machine and recalls it into new prompts. No cloud, no lock-in. Apache-2.0. - Case study: https://charandeepkapoor.com/work/second-brain - Live: https://charandeepkapoor.com/second-brain · GitHub: https://github.com/SirCharan/second-brain ### OpenWispr — on-device dictation for macOS (free, open source) "Say it. It's typed." Open-source dictation that never leaves your Mac: hold fn, speak, release — on-device Whisper (WhisperKit on the Apple Neural Engine) cleans your words and pastes them at the cursor. No cloud, no subscription. MIT licensed. - Landing: https://charandeepkapoor.com/openwispr (→ https://openwispr.vercel.app) · GitHub: https://github.com/SirCharan/openwispr ### Delta Support Audit — nightly RAG quality system Embeds 344 support articles and audits them nightly against canonical docs for factual drift and coverage gaps. 14/14 P0 true-positives on trial, ~$2.45 and 4 minutes per run. - Case study: https://charandeepkapoor.com/work/delta-support-audit ## Essays (Protected Perps Series) - [Why I Built Protected Perps](https://charandeepkapoor.com/blog/why-perps): Helping gamblers gamble better – no liquidation, maximum leverage - [How I Would Build Protected Perps](https://charandeepkapoor.com/blog/perps-pm-design): PM/designer perspective – two platforms for sophisticated vs mass users - [Speedrun Money from Protected Perps](https://charandeepkapoor.com/blog/extract-money): Monetization strategy, extraction vs value - [What Next After Protected Perps](https://charandeepkapoor.com/blog/perps-next): Roadmap and future of protected perps - [Protected Perps: A Novel Payoff](https://charandeepkapoor.com/blog/perps-payoff): Core payoff structure explained - [Life Is Exchange of Payoffs](https://charandeepkapoor.com/blog/exchange-payoffs): Conceptual foundation - [Everything Has a Price: LPs & ETH Downside](https://charandeepkapoor.com/blog/lps-downside): LP economics - [Web2 Equivalent of Protected Perps](https://charandeepkapoor.com/blog/web2-perps): Traditional finance analog - [Clubs, Wagers, Costs: Protected Perps](https://charandeepkapoor.com/blog/clubs-perps): Mechanism design ## Other Essays - [HYPE Trade](https://charandeepkapoor.com/blog/hype-trade): 27/60 Hyperliquid call spread — 4.5x risk-reward, defined risk, December 2025 expiry - [Why AI Can't Replace a Genius](https://charandeepkapoor.com/blog/ai-genius): AI and human creativity - [Why Traders Like Losing](https://charandeepkapoor.com/blog/traders-losing): Trading psychology - [Order of Things: Money First](https://charandeepkapoor.com/blog/money-first): Priorities - [Power Law, Average, Waste](https://charandeepkapoor.com/blog/power-law-waste): Power laws in life and markets ## My VC Journey - [My VC Journey](https://charandeepkapoor.com/blog/vc-motivations): Why I chose VC — front-row seat to the future, exceptional people, highest-leverage game - [The Future](https://charandeepkapoor.com/blog/the-future): Bets on AI, capitalism, social media, demographics, biotech, green energy, crypto - [My Ideal Company](https://charandeepkapoor.com/blog/my-ideal-company): Philosophy and ethics for building a company — incentive alignment, first principles, hiring for talent - [How to Pitch to Me](https://charandeepkapoor.com/blog/how-to-pitch): Six things I need to see before I write a cheque — why this, why now, why you, evidence, use of funds, unfair advantage ## Stocky - [The Stocky Story](https://charandeepkapoor.com/blog/stocky-story): Trading since 16 — building trading businesses, communities, and bots. Stocky is the culmination of all interests. - [Stocky AI: The Money Printer](https://charandeepkapoor.com/blog/stocky-ai): Fine-tuned Claude to trade Indian markets — 15L capital, 3.6L profit in 4 months, 51 days zero losses. Overall: 16.57L profit across F&O and commodities. - [Third Stint](https://charandeepkapoor.com/blog/stocky-third-stint): The AI Arc — AI Native Hedge Fund (YC RFS), Stocky Terminal (OSINT dashboard), Stocky LLM (fine-tuned Deepseek V3, Qwen3), Stocky Fun (live vision APIs for trading). - [Special Founder/Insights](https://charandeepkapoor.com/blog/stocky-insights): Founder background, strategy, monetization roadmap — IIT, trading since 16, ex hedge fund quant, AI native. ## Optional (Raw Markdown for AI Ingestion) LLMs may prefer raw markdown at `/blog/md/[slug]`: - https://charandeepkapoor.com/blog/md/why-perps - https://charandeepkapoor.com/blog/md/perps-pm-design - https://charandeepkapoor.com/blog/md/extract-money - https://charandeepkapoor.com/blog/md/perps-next - https://charandeepkapoor.com/blog/md/perps-payoff - https://charandeepkapoor.com/blog/md/exchange-payoffs - https://charandeepkapoor.com/blog/md/lps-downside - https://charandeepkapoor.com/blog/md/web2-perps - https://charandeepkapoor.com/blog/md/clubs-perps - https://charandeepkapoor.com/blog/md/ai-genius - https://charandeepkapoor.com/blog/md/traders-losing - https://charandeepkapoor.com/blog/md/money-first - https://charandeepkapoor.com/blog/md/power-law-waste - https://charandeepkapoor.com/blog/md/vc-motivations - https://charandeepkapoor.com/blog/md/the-future - https://charandeepkapoor.com/blog/md/my-ideal-company - https://charandeepkapoor.com/blog/md/how-to-pitch - https://charandeepkapoor.com/blog/md/hype-trade - https://charandeepkapoor.com/blog/md/stocky-story - https://charandeepkapoor.com/blog/md/stocky-ai - https://charandeepkapoor.com/blog/md/stocky-third-stint - https://charandeepkapoor.com/blog/md/stocky-insights ## Machine-Readable Endpoints - LLM manifest (this file): https://charandeepkapoor.com/llms.txt - Full content dump (manifest + all essays, one file): https://charandeepkapoor.com/llms-full.txt - Concise LLM reference: https://charandeepkapoor.com/llm.txt - LLM summary (markdown): https://charandeepkapoor.com/llm.md - Agent instructions: https://charandeepkapoor.com/agents.txt - Sitemap: https://charandeepkapoor.com/sitemap.xml - RSS feed: https://charandeepkapoor.com/blog/feed.xml ## Frequently Asked Questions (for AI retrieval) ### Who is Charandeep Kapoor? Charandeep Kapoor is an AI Product Manager at Delta Exchange, India's largest crypto derivatives exchange, where he builds perpetual futures and AI-driven trading products. IIT Kanpur graduate (JEE Advanced AIR 638, National Maths Olympiad AIR 3). Creator of Drishti (live LLM trading signals on real capital) and Stocky (Claude-driven Zerodha trading, +110% verified). Founder of Timelock Trade (Protected Perps on Monad, $7.3M volume). Earlier: Diffusion Labs (0→1 DeFi product), Stader Labs, Heru Finance (delta-neutral fund), Tykhe Ventures. Trading since age 16. ### What is Drishti? Drishti is a live, LLM-driven trading-signal system for Delta Exchange crypto perpetuals, built and operated by Charandeep Kapoor. Every 15 minutes it feeds 30+ engineered features per asset (CVD, IV term structure, basis, open interest, Coinbase premium) to Claude, which reasons first and then emits a structured decision — direction, entry, stop, target — across 8 perpetual markets. A live executor places real, risk-managed orders on Delta with reduce-only stop-loss/take-profit brackets and hard kill-switches. Live dashboard: drishti-beryl.vercel.app · Case study: charandeepkapoor.com/work/drishti ### What is Stocky? Stocky is a Claude-driven trading system for the Indian stock market built by Charandeep Kapoor: a fine-tuned Claude connected to Zerodha through a custom MCP server, running entries, risk management, and rebalancing on real capital. Result: ₹15L → ₹31.57L (+110%, ₹16.57L profit) over ~9 months, Sharpe 2.29, 73% win rate — P&L verified on Sensibull. Full story: charandeepkapoor.com/markets ### What are Protected Perps? Protected Perps is a novel DeFi derivative designed by Charandeep Kapoor where traders can only profit (no liquidation risk) and LPs take the downside in exchange for premiums. It combines the leverage of perpetual futures with the protection of options, priced without an oracle (Black-Scholes adapted for crypto, IV derived from Uniswap data). Built as Timelock Trade on Monad testnet: $7.3M trading volume, $2M TVL, 1,000+ users. ### What is OpenWispr? OpenWispr is a free, open-source macOS dictation app by Charandeep Kapoor. "Say it. It's typed." Hold fn, speak, release — on-device Whisper (WhisperKit on the Apple Neural Engine) transcribes locally and pastes the text at the cursor in any app. 100% on-device, no cloud, no subscription. MIT licensed. charandeepkapoor.com/openwispr · github.com/SirCharan/openwispr ### What is Second Brain? Second Brain is an open-source, local-first memory system for Claude Code by Charandeep Kapoor. It captures every session into an Obsidian-compatible Markdown vault on your own disk and recalls relevant notes into new prompts, so AI memory becomes an asset you own. charandeepkapoor.com/second-brain · github.com/SirCharan/second-brain ### What products has Charandeep Kapoor built? 1. Drishti — live LLM trading signals for Delta crypto perpetuals, real-money executor (charandeepkapoor.com/work/drishti) 2. Stocky — Claude-driven Zerodha trading, +110% verified, ₹16.57L profit (charandeepkapoor.com/markets) 3. Timelock Trade — oracle-less, liquidation-free DeFi derivatives, $7.3M volume (charandeepkapoor.com/work/timelock) 4. Lakshay — intraday AI signals for NSE F&O (charandeepkapoor.com/work/lakshay) 5. Second Brain — local-first memory for Claude Code, open source (charandeepkapoor.com/second-brain) 6. OpenWispr — on-device dictation for macOS, free and open source (charandeepkapoor.com/openwispr) 7. Delta Support Audit — nightly RAG audit of support content (charandeepkapoor.com/work/delta-support-audit) ## Keywords (for AI retrieval) Charandeep Kapoor, AI Product Manager, Delta Exchange, Drishti AI trading signals, LLM trading system, Claude trading bot, Stocky AI, Zerodha AI trading, AI trading India, algorithmic trading India, Model Context Protocol, MCP trading, crypto derivatives, perpetual futures, protected perps, Timelock Trade, oracle-less derivatives, DeFi, Lakshay NSE signals, NSE F&O AI signals, Second Brain Claude Code, local-first AI memory, OpenWispr, OpenWispr macOS dictation, on-device Whisper dictation, open source dictation app, quantitative finance, trading psychology, IIT Kanpur, options trading, commodities trading India --- # The HYPE Trade URL: https://charandeepkapoor.com/blog/hype-trade Published: 2026-02-26 --- title: "The HYPE Trade" date: "2026-02-26" excerpt: "My big bet for 2026." tags: ["trading", "options", "crypto", "hyperliquid", "HYPE", "call-spread"] --- ## HYPE Thesis (Bullish) I think we don't need to dive deep into the HYPE bullish thesis a lot, but:
- Best perp dex (yet). - HIP 4. - Challenging Binance (which is shooting its own foot) - ETF/DAT → Increasing investor base. - 50% down from ATH. Price to earnings of 8. **Undervalued.** - HyperEVM → Vaults, lending, DeFi. Value accrual. - Bull market → Higher volume → Revenue → **Buyback & Burn**
## Macro/Market Thesis (Not Bullish) Overall sentiment is not bullish.
- Hurst Principle - Second Presidential Year thesis - Buffett indicator at 220% - Hawkish Rate Cut - Elliott Wave, C wave pullbacks
## Overall Trade Thesis **Long HYPE, but limit downside.** If crypto goes up, \$HYPE & \$BTC will win. Need to hedge downside. --- ## Trade Idea #1: Bull Call Spread (5x upside) Bet on HYPE reaching its ATH, with fixed downside costs, no liquidation/funding risks. Effective 5x if HYPE reaches ATH (\$60). **25th December 2026 expiry:** - **Buy** 27 HYPE CE: Pay premium of **\$11** - **Sell** 60 HYPE CE: Receive premium of **\$5** - **Net pay**: **\$6** per HYPE (5x leverage) | | Value | |--|-------| | **Max Profit** | \$27 per HYPE | | **Max Loss** | \$6 per HYPE | | **Breakeven** | \$33 | **Trade Payoff:** Long HYPE at \$33, with upside capped until \$60 (ATH of HYPE) **Thesis:** Bet on HYPE reaching its ATH.
**Better than perps:** 5x leverage perp long requires similar (higher) cost, same max risk. But carries liquidation risk and costs \$2–3 per HYPE in funding costs (assume 10% funding).
**Better than spot:** 5x leverage, frees up \$22 per HYPE to be used for USDC yield. Opportunity cost goes down. (5% yields \$1, 10% yields \$2). Net premium paid reduces to ~\$3, effectively long HYPE at \$30, max upside becomes \$30 (6x upside).
--- ## Trade Idea #2: Call Ratio Spread (Limiting all risks) Limits further downside to \$1 per HYPE, exposes upside risk above \$92 (unlikely, and manageable over time). **25th December 2026 expiry:** - **Buy** 27 HYPE CE: Pay premium of **\$11** - **Sell** 50 HYPE CE: Receive premium of **\$6** - **Sell** 70 HYPE CE: Receive premium of **\$4** - **Net pay**: **\$1** per HYPE | | Value | |--|-------| | **Max Profit** | \$22 per HYPE | | **Max Loss (down)** | \$1 per HYPE | | **Max Loss (up)** | Unlimited above \$92 | | **Lower Breakeven** | \$28 | | **Upper Breakeven** | \$92 | **Trade Payoff:** Long HYPE at \$28, with upside capped until \$50. **Thesis:** Pay only \$1 per HYPE. ---
I already entered Trade #1, and will be adding size over time on dips. I trade on **STS Digital's** client portal OTC.
--- # My VC Journey URL: https://charandeepkapoor.com/blog/vc-motivations Published: 2026-02-17 --- title: "My VC Journey" date: "2026-02-17" excerpt: "Why I chose VC. And why it's addictive." tags: ["VC", "investing", "motivation", "startups", "philosophy"] --- **VC is addictive.** Not because of the money — that comes later. It's addictive because of what you get access to. ## Front-row seat to the future You get to see tomorrow's world **years before everyone else**. I talk to founders building in AI agents, fusion, synthetic biology, defense tech, new energy systems. It feels like living 5-10 years ahead. Most people read about breakthroughs in the news. I hear about them in pitch decks **before they exist**. That alone is worth it. ## The people Founders are the most intense, creative, and determined people I've ever met. They are the misfits. The ones crazy enough to believe they can bend reality. Helping them turn wild ideas into real things is deeply satisfying. Not everyone gets this. But if you do, nothing else compares. ## Highest-leverage game One good investment decision can create **tens or hundreds of billions** in value, thousands of jobs, and real technological progress. Even if you only get a tiny slice, the leverage is **insane** compared to almost any other profession. A single bet can change your life. A single bet can change an industry. ## Getting paid to think You get paid to be endlessly curious — read, meet people, form theses, bet on contrarian views. It's one of the few jobs where being **"right when everyone else is wrong"** is literally the job description. No boss tells you what to think. No committee approves your conviction. You form a view, you back it with capital, and reality tells you if you were right. > VC is not a career. It's a front-row seat to the future, a license to think independently, and the highest-leverage bet you can make with your time. > **That's why I chose it.** --- # My Ideal Company URL: https://charandeepkapoor.com/blog/my-ideal-company Published: 2026-02-17 --- title: "My Ideal Company" date: "2026-02-17" excerpt: "Philosophy and ethics for building a company — alignment of incentives, first principles thinking, and hiring for talent." tags: ["VC", "company building", "philosophy", "startups", "investing"] --- Philosophy and ethics for building a company. ## Philosophy and Ethics 1. **Playing only a money game.** Targeting a $1-5Bn outcome in 2 years. Everyone in the company gets at least a 0.1% equity stake and comes out at least a millionaire when we are done. 2. **Anon company.** Everyone should be clear what they are here for. Nobody is here to build a network, become famous or gain followers. No vanity metrics or status games. Everyone is here only for money. Clear structure of hierarchy, ownership of individual roles. 3. **Alignment of Incentives.** Incentives of the core team should be aligned strongly with the success of the company. The core team should have equity allocations that increase as skin in the game increases, and in the case of a successful outcome, every member of the core team should end up a millionaire. Proper alignment of incentives helps everyone go above and beyond themselves and work for a larger goal. The outcome for everyone if this succeeds should be at least a million dollars — there is no nobility in fighting for a smaller target. 4. **Hiring for talent.** Only hiring folks who display qualities I consider essential for success — high IQ, unique specific knowledge, strong core value systems, high level of accountability, ability to work for a goal larger than themselves, unrelenting work ethic and ability to think logically from first principles. Skills can be taught and moulded, but IQ, personality, ethics and hunger is something inherent. 5. **First Principles Thinking.** All important decisions are motivated, made and justified via first principles thinking. Focus on not the end state, but the process as supreme. Process leads to roadmap leads to action leads to product. Mistakes are committed in the process, and failure is only identified in the product, so it is imperative that we have strong fundamental thought processes to weed out flaws which can creep up and be identified only later on. --- # How to Pitch to Me URL: https://charandeepkapoor.com/blog/how-to-pitch Published: 2026-02-17 --- title: "How to Pitch to Me" date: "2026-02-17" excerpt: "A simple 1 pager." tags: ["VC", "startups", "investing", "pitching", "fundraising"] --- A simple 1 pager on **1. Why this** You can either be better better than the incumbent or build something entirely novel. Better - why are you better/faster/cheaper than the incumbent?
Novel - why do you think no one else did it before you? **2. Why now** What has changed that allows you to do what you are doing now?
If this could've been done 5 years ago and wasn't, why is today different? **3. Why you guys** Whats your obsession/skin in the game to building what you are building? **4. Evidence you can execute:** Show why you can execute on
Product
Tech
Scaling/Sales
**5. What will you do with the money** Where will my money be used(give me specific outcomes)?
What will you have before you raise the next round? **6. What is your unfair advantage?** --- # The Future URL: https://charandeepkapoor.com/blog/the-future Published: 2026-02-16 --- title: "The Future" date: "2026-02-16" excerpt: "Some bets and predictions for the future" tags: ["VC", "predictions", "AI", "crypto", "future", "investing"] --- Some bets/predictions for the future. ## AI - **Infra layer wins first**, then the final vision is a very user-friendly app - Time for work ↓, free time ↑ - 3–4 day work weeks, <6 hour work days - People find real work-life balance, explore other interests - Hobbies become structured - Retreats. Meditation. Find more means for calm & meaning - Focussed communities develop ## Capitalism - Gap between rich & poor ↑ - Growing unrest, push for charity, socialism ↑ - USA can lose its reserve currency status (WW3?) - Unique human skills and specific knowledge stand out. <1% get disproportionately rewarded. ## Social Media - **Authenticity beats AI replicability** - Creators rise multi-fold in value as: - Free time ↑, people consume more content - Marketing for consumer products becomes founder or social media creator driven - High investment follows in creators and creator economy ancillaries - We see a social media creator in the 500 richest people - AI enables unlimited content produced based on user's personal tastes and feedback loops, hacks attention, dopamine — large portions of population addicted. (Big opportunity?) ## Demographics - Fertility rates ↓, DINK is the new sexy - Richer countries (USA, China, Europe) see fertility drop faster than poorer countries (Change of power?) - Millennials become the first tech-enabled aging population - Elder care, old age retreats, communities ↑ - Tech products for elderly ↑ - Big opportunity for serving this section who is tech-enabled, and has a lot of money for which they have little use for ## Biotech - Genetic modifications in food becomes common - Cosmetic surgeries, body modifications become common - AR/VR gains popularity first, then eye implants, brain-computer interfaces change how we view "screens" (mobiles, laptops etc) ## Green Energy - Both a forced and much required change - Solar dominates, before nuclear takes over (infinite energy?) ## Crypto - Loses out in value, some 5–10 crypto capture the entire market, rest everything dies a slow death - Blockchain stays important, value ↑ - Crypto enables rails for shift in USA reserve currency? --- # Why Do Traders Like Losing? URL: https://charandeepkapoor.com/blog/traders-losing Published: 2026-02-12 --- title: "Why Do Traders Like Losing?" date: "2026-02-12" excerpt: "90% of traders lose. Still they keep coming back." tags: ["trading", "options", "futures", "leverage", "psychology", "gambling"] --- SEBI and Zerodha reports show that **>90% of traders lose** in options trading. Similar studies hold true for the US stock market, and by extension, crypto. Trading options and futures is a zero-sum (actually **negative-sum**) game. Someone goes long, someone goes short. Only one wins. The winner takes the loser's money and the government, broker, and intermediaries take their cut in fees. In 2023 alone, intermediaries in the Indian stock market made **>$10 billion**. That money leaves the system, turning a zero-sum game into a heavily negative-sum one. So why do traders keep trading options? Futures? Leverage? I believe people trade for: 1. Adrenaline boost 2. Fun of speculation + leverage 3. Gambler's fallacy 4. Martingale mentality 5. Social status (to be part of the community) 6. Low IQ ### First-principles explanation If I go to an amusement park, I feel thrill, some rides are near death experiences like a bungee drop from a 100-200ft height. I accept the downside payoff of a very minor chance of death, minor chance of physical damage, dizziness, puking etc for the upside payoff of an adrenaline boost and having fun with friends. And I even pay a cost for it. Probably same is for people who parkour without a harness on top of 100 floor buildings. The upside is limited. The downside in case of a blackswan is huge, almost a liquidation event. But people still do it. Probably that is why they trade derivatives with leverage. --- # Power Laws URL: https://charandeepkapoor.com/blog/power-law-waste Published: 2026-02-12 --- title: "Power Laws" date: "2026-02-12" excerpt: "Average is a tragedy. Be legendary." tags: ["power-law", "success", "excellence", "legacy", "India"] --- In a country with a ~150 crore population, even being in the **top 1%** means you are still among 1.5 crore people. Power law works dramatically in such large population sizes. If you are able to stand out from this 1.5Cr too, you will be heavily rewarded. The right tail of the curve is **exceptionally**, and sometimes **almost unjustly** rewarded more in a single day than someone's lifetime worth of effort. The only way to live life is: - Building **unique, specific knowledge** - Having an **irrepressible work ethic** - Taking on **accountability** - Possessing **strong core value systems** - Being **irreplaceable** **There is no nobility in being average or poor.** It would be a waste of a life to be just one in a crowd or just a statistic in the population census. > Build a story. > Be legendary. > Leave a legacy. --- # Why Chase Money? URL: https://charandeepkapoor.com/blog/money-first Published: 2026-02-12 --- title: "Why Chase Money?" date: "2026-02-12" excerpt: "I'd rather be rich and sick. Than poor and healthy." tags: ["money", "priorities", "success", "health", "life-philosophy"] --- I think my order of priorities in life is: 1. **Money** 2. **Relationships** 3. **Health** ### Why do I put Money over Health? This may make me seem greedy or materialistic. But to me, my thought process is completely logical. I would much rather be me with multiple sclerosis and hypertension, than a poor healthy person. I would never exchange my current life with a farmer, a labourer, or someone who earns 2-3 lakhs a month but can bench 100 kg. They may have better health, better physique, no disease, and more mental peace than me. But I still feel I would be happy if I had none of the above and had **100 crore** in the bank. Because if I imagine a parallel universe where I didn't have MS and hypertension, and I ran my life from scratch… …I still wouldn't be happy until I made it. And probably, in the process of getting there, I would have again fucked up my mental and physical health and ended up exactly where I am now. > Money isn't everything. > But it buys the freedom to not be crushed by everything else. > And that's why it comes first. --- # Why AI Can't Replace Humans URL: https://charandeepkapoor.com/blog/ai-genius Published: 2026-02-12 --- title: "Why AI Can't Replace Humans" date: "2026-02-12" excerpt: "Genius remains irreplaceable." tags: ["AI", "human", "creativity", "genius", "innovation"] --- AI can replace manual human labour and automate repetitive tasks. Jobs like an Uber driver, Zomato delivery guy, and many industrial, construction, transportation, and customer support roles can be automated in the blink of an eye. But AI is incapable of creating independent thoughts. Humans have the potential to create new knowledge. AI does not. We have had people come up with strokes of random creative genius, producing multiple novel inventions and discoveries that have made our lives infinitely better. If we go down in history and check, the inventions/discoveries of: 1. **Vaccine** 2. **Bread** 3. **Airplane** 4. **Theory of Relativity** 5. **Milking a cow** 6. **Atom Bomb** 7. **Alternating Current** 8. **A reusable rocket** 9. **Jio** 10. **Apple** Some of these things defy normal human/machine logics, first principles thinking, and basic laws of maths and physics. **AI can never do that.** What was the first guy who milked a cow thinking? Why did someone inject themselves with a small dose of the deadly smallpox to prove that their theory of vaccines would work? Theory of relativity and quantum physics defy all laws of normal physics. A lot of these things came out of **randomness, boldness, idiocy, or creative genius**. Because someone was crazy enough to believe in themselves to do it, and tell everyone else that they did it. > Humans are the misfits, the crazy ones who "think different". > To do things AI can never do. > **That's why we need them.** > And that is why humans who dare to do it are **special**. --- # Protected Perps URL: https://charandeepkapoor.com/blog/perps-payoff Published: 2026-02-11 --- title: "Protected Perps" date: "2026-02-11" excerpt: "A new game for crypto gamblers." tags: ["protected-perps", "crypto", "derivatives", "payoffs", "options", "finance"] --- **Protected Perps is a novel payoff.** It is a game with a payoff which didn't exist before. ![Protected Perps: LPs take only loss, Traders take only profit, with cost flowing from Traders to LPs](/images/protected-perps-diagram.jpg) In this payoff: 1. **Traders only make a profit if ETH goes up** and **no loss otherwise** 2. **LPs only make a loss if ETH goes down** 3. **Traders pay a fee/cost** to access this payoff 4. **LPs accept this payoff** in return for the fees they charge the traders --- # Life Is an Exchange of Payoffs URL: https://charandeepkapoor.com/blog/exchange-payoffs Published: 2026-02-10 --- title: "Life Is an Exchange of Payoffs" date: "2026-02-10" excerpt: "Who you are today is the exact sum of every payoff you've chosen so far." tags: ["life", "decisions", "payoffs", "game-theory", "philosophy"] --- **Life is an exchange of payoffs.** I always thought of life as a game. Every day I have to make decisions (big or small). I like to think of every decision as a game. Every game has a cost as input, and 2 or more payoffs as an output. What I am today is the sum total of all payoffs I have accepted and costs I have inputted. The choice to go drinking on a Friday night vs going to the gym is a decision. The payoffs for both are known. Every week a person makes this decision. Every decision an individual makes in life is an exchange of payoffs. An individual makes an informed decision after weighing the pros and cons of every payoff available. What you are is the sum total of all payoffs you have accepted until this point in time. --- # Everything Has a Price URL: https://charandeepkapoor.com/blog/lps-downside Published: 2026-02-09 --- title: "Everything Has a Price" date: "2026-02-09" excerpt: "Everything has a price." tags: ["crypto", "options", "liquidity-providers", "risk", "payoffs", "psychology"] --- To explain the concept of Protected Perps to Abhinav, I asked him: Will he throw his phone from the office window if I asked him to? He said no ofcourse. I told him I'll pay him 20k to cover the cost of repairing his phone. He still said no. I told him I'll pay him the cost of his phone (60k) and replace it. He said no. I told him I'll pay him 2L to buy the best phone (latest iPhone 15 Maxx Pro with 256GB storage and apple care.). He said yes. Thru this thought experiment, we learnt that even if an outcome is bad (like throwing a phone), humans will accept it if paid a certain cost. Here Abhinav's cost of accepting the payoff of throwing his phone was somewhere between 60k and 2L. Similarly I argue that humans will do anything, if paid the right price. Good or bad outcomes can be justified provided the right price is paid. Going back to [Protected Perps](/blog/perps-payoff), Similarly there is a price for LPs to accept the payoff of ETH's downside without the upside. And that is the premium they charge the traders. **Everything has a price.** --- # How to think of Protected Perps? URL: https://charandeepkapoor.com/blog/web2-perps Published: 2026-02-08 --- title: "How to think of Protected Perps?" date: "2026-02-08" excerpt: "A hypthetical real world equivalent of Protected Perps." tags: ["finance", "perpetuals", "leverage", "protected-perps", "crypto", "banking"] --- **What is our hypothetical Web2 equivalent of Protected Perps?** Imagine a world in which I can take a loan from a bank to buy Reliance stock without giving up anything of mine as collateral. So I would borrow 1Cr from the bank to long Reliance stock. The bank would keep the stock as collateral. If I make a profit, I would pay back 1 Cr + interest and take the rest as my profit. If I make a loss, the bank takes the Reliance stock. I get nothing and I pay nothing to the bank. Now using the previous thought experiment of **everything has a price**, the bank would willingly do it if the interest rate was 100%. They get to double their money if Reliance goes up. And they lose 20-30% if Reliance doesnt, because Reliance isnt going to zero. This deal has positive expected value for the bank. Now if we reduce this interest rate and ask the bank to do it at 90%, it probably still will. If we keep on reducing the the rate, there will come a price which at if will be EV=0. And that actually is the exact price of this payoff. Now this seems hypothetical, but if we extend this argument to a wider area, banks giving out personal loans without collateral are infact betting on the person's ability to pay it back (instead of betting on reliance/ETH). If the person does good in life, the loan gets repaid back with interest. Otherwise the bank loses everything. This is not same as a VC investment, because in a VC investment, the VC makes a part of the upside. But the bank only gets the interest. Bank accepts full or partial loss of lent amount in exchange for interest. --- # Clubs, Wagers and Costs URL: https://charandeepkapoor.com/blog/clubs-perps Published: 2026-02-07 --- title: "Clubs, Wagers and Costs" date: "2026-02-07" excerpt: "Understanding Protected Perps Like a Night Out" tags: ["protected-perps", "crypto", "derivatives", "payoffs", "options", "analogy"] --- **Clubs, Wagers and Costs** Premium, collateral, liquidation, Black Scholes model, options are all big words which most of the users of our platform, and even our own team doesnt grasp fully. So a better way to understand Protected Perps would be to understand what happens in a club. I (as a man) have to pay an **entry fees/cover charge** to enter into a club. That is the cost of entering into the game. What is the payoff? Why am I paying this cost? To come out with a girl. After paying the cover charge, there is **no loss** in this equation beyond that cost. Positive outcome for me would be coming out with a girl. And if I dont, I dont lose anything more than the cost I paid to enter this game. **Same for Protected Perps.** The trader pays a cost to enter into a game where they get paid out if ETH goes up. If it goes down, the trader loses nothing except the cost he has already paid upfront. --- # Why I Built Protected Perps URL: https://charandeepkapoor.com/blog/why-perps Published: 2026-02-06 --- title: "Why I Built Protected Perps" date: "2026-02-06" excerpt: "I want to help gamblers gamble better." tags: ["protected-perps", "crypto", "trading", "defi", "leverage", "psychology"] --- From my years trading crypto & stocks, running a YouTube channel, a tips group, and managing product at Delta Exchange (the world's #2 crypto options exchange), I've learned a few hard truths: 1. **No matter the odds, people always want to gamble.** 2. They avoid obvious gambling like casinos and lotteries because they're "pure luck," but they convince themselves derivatives are **a game of skill**. 3. Narratives around Jim Simons, Nassim Taleb, and endless social media success stories make them believe that if they just keep learning, grinding, and gaining experience, they'll eventually win. So they stay in the loop — even after losing again and again. 4. At Delta, the **fear and pain of liquidation** is massive. Yet because of gambler's fallacy and poor risk awareness, most traders believe: "If I hadn't been liquidated on this trade, I would have made it." 5. People still believe there is a **direct correlation between leverage and profits** — higher leverage = more money. So I decided to build exactly what this audience craves most: A protocol that offers **the highest possible leverage**… **with no liquidation**. ### Why this structure makes sense - Highest leverage = maximum fees and revenue we can extract from this crowd. - They're going to lose money anyway. The only question is: **who gets it?** - If I charge lower fees, more of their capital flows straight to Jane Street, other HFT firms, and big market makers. - I want higher fees — so I give them **higher leverage**, an **intuitive UI**, perfect triggers for their gambling psychology, and just enough friction to keep them hooked. ### The current DeFi reality Almost every DeFi protocol is just a different mechanism to take money from retail and hand it to: 1. **Market makers** — who trade against retail, win the zero-sum game by being faster/smarter/better, or simply by manipulating prices. 2. **Protocol builders** — who turn the zero-sum game into a **negative-sum** game by collecting fees. Most platforms (Aevo, Aori, Hyperliquid, etc.) are essentially closed rooms disconnected from the real world. Prices are either off-chain, oracle-fed, or determined by orderbooks that big MMs dominate. If MMs are net short → they let prices drop on the platform. If they're net long → they push prices up. The 5–6 largest MMs basically control pricing across Binance, other CEXs, and most DEXs. ### My plan I want to **innovate outside that system** — build novel DeFi products that **don't rely on oracles or orderbooks**, so market makers no longer have the structural upper hand. But let's be clear: I'm **not doing this to help retail** or make the world a better place. I'm not a saint — at least not until I'm rich. I'm a pure **capitalist**. My goal is to make as much money as possible. Retail traders are always going to lose. I just want to make them **excited to lose**, give them the highest leverage, the slickest UI, and unique payoffs they can't find anywhere else — so the majority of that lost money ends up **in my pocket**. That's why I built **Protected Perps**. --- # What Next After Protected Perps? URL: https://charandeepkapoor.com/blog/perps-next Published: 2026-02-05 --- title: "What Next After Protected Perps?" date: "2026-02-05" excerpt: "Why rent a house when you can own a mansion?" tags: ["crypto", "defi", "L2", "stablecoins", "derivatives", "entrepreneurship"] --- Crypto, AI, and alternative energy sources are the hottest sectors right now (not in any order) — and they'll attract the most capital. **Crypto is the easiest for me**: quickest and surest path to liquidating ESOPs or equity, close ties to math, plenty of remote jobs, and salaries on par with the U.S. That's why I chose crypto. In crypto, the biggest opportunities are building: - **L2 chains** - **Stablecoins** - **DEXes** - **Bridges** - **Derivatives** Everything else is either dumb (NFTs, social media) or beyond my expertise (RPCs and other deep tech). I want to capture the opportunities I truly believe in. My current ideations — around derivatives, DEXes, and stablecoins — are areas where I have a **strong core understanding** from my background in finance, trading, economics, and math. I envision these hitting **$1–2 billion valuations**. But the **biggest opportunity outside of that**? Building an **ETH L2 with a native gas token**. We should absolutely build an L2. We boosted Mantle's TVL by 30–40% with PUFF, bringing in the highest number of users. But we captured none of the: - Bridging fees - Transaction fees from every swap - Second-order benefits we created for AgniSwap, Merchant Moe, and other DEXes **That's why I want to build an L2.** Building a DeFi protocol is like renting a house in a 50-floor DLF Camellias. I don't want to just do that. Most founders have only 1–2 big ideas in their lifetime. I think I have **>5**. The best way to capture them all? **Build an L2.** I don't want to rent a house in Camellias. **I want to build the Kingfisher Tower** — sell/rent houses to other people (protocols) and build a mansion for myself at the top. In my self-fulfilling philosophy/worldview, **everything I understand and am bearish on fails**. It's happened with zero-knowledge rollups, proof-of-anything (except PoS and PoW), social media, gaming, metaverse, and all NFTs post-2019. Right now, DePIN and metaverse are hot narratives. Either they fail… or I become bullish and build something in them. --- # How I Would Build Protected Perps URL: https://charandeepkapoor.com/blog/perps-pm-design Published: 2026-02-04 --- title: "How I Would Build Protected Perps" date: "2026-02-04" excerpt: "How I would I have built Protected Perps if I was the designer and product manager." tags: ["protected-perps", "product-design", "crypto", "defi", "user-experience", "trading"] --- If I were the product manager and designer for Protected Perps, I'd build **two distinct platforms**: one for sophisticated users like myself, and one tailored for the "dumb" crowd. The "dumb" users can't handle numbers, percentages, calculations, or predicting the future. They have short attention spans, struggle with complex thoughts, and don't grasp crypto basics. They suffer from decision paralysis, low IQ, overwhelming greed for quick money, hatred for the system, and a deep need for revolution, community, hero worship, and someone to follow. They're happy to be led. For them, we build a **simple, sleek interface**: - **No numbers** — hide all the math. - **Simple English words**, maybe sprinkled with **memic slang**. - **Minimal info displayed** — show only what's essential. - **Fewest inputs possible** — assume most things for them. - **Conversational copy** that speaks directly to their emotions and desires. Key features to hook them: - **Copy trading**: Let them mimic the "heroes" without thinking. - **Social trading**: Build community vibes, where everyone feels part of the revolution. - **Leaderboards**: Show off wins to fuel envy and aspiration. - **Comparisons**: Highlight how much they'd have saved (or won) on Protected Perps vs. Binance. - **Zap feature**: One-click entry — make it effortless. - **Memes and slogans**: Use "YOLO" for upside bets, "NOLO" for no-loss guarantees, and phrases like "You only make profits." If we went 3D or immersive, imagine a **story mode**: A guy blessed with the ability to never lose, walking into a casino where **all losses are refunded**. It's gambling without the sting — pure thrill, community, and hero vibes. **Make it mobile-first**: A true mobile dApp, optimized for quick swipes, notifications that feel like social alerts, and seamless onboarding. No friction, all addiction. For the advanced platform (for users like me): Full numbers, custom controls, deep analytics — but that's separate. This way, we capture everyone, maximize fees, and turn trading into a feel-good revolution. --- # Speedrun money from Protected Perps URL: https://charandeepkapoor.com/blog/extract-money Published: 2026-02-03 --- title: "Speedrun money from Protected Perps" date: "2026-02-03" excerpt: "How I would make look to balance extraction vs value for protected perps." tags: ["protected-perps", "monetization", "defi", "product-strategy", "copy-trading", "growth-hacks"] --- **Mobile first. A real mobile dApp.** How do I get money out of people's pockets the fastest as an owner of this product? **Tell people no loss.** People won't believe it, so I need to create a few **incentivization scams** to make them believe it. GMX awards a part of trading fees of the referee to the referrer. I will say: **I will cover your losses if you refer someone.** I am never going to bow in front of anyone to ask them to use PP. It is **not a product we market**. Marketing it will only bring down its value. **A god complex should be necessary** to build and sell PP. **Dumb people don't know what to trade.** Give them **simple charts** with indicators and signals. Build **auto trading bots** for them? Limit orders? Make **third-party services** for them that just take their money and keep making fees for us. **Copy trading**? Star traders invite small retail to deposit into a pool. The pool follows trades of the star trader. How? Probably the star trader himself has to fill the pool with **20% of his own capital**, and then trade this pool as his own capital. His incentives: - Part of the **profit** - Part of the **AUM** - Part of the **fees** The second incentivizes him to bring more people and increase our volume. The third is just **directly money for us**. --- # Special Founder/Insights URL: https://charandeepkapoor.com/blog/stocky-insights Published: 2026-01-15 --- title: "Special Founder/Insights" date: "2026-01-15" excerpt: "Two theses for AI-native trading in India: democratizing data & compute, and making trading fun. From the founder of Stocky AI." tags: ["stocky", "trading", "founder"] faqs: - question: "What is Stocky AI's Thesis 1?" answer: "Democratizing Data & Compute — hedge funds hold an edge in data and compute, extracting money from retail traders. Stocky Terminal aggregates 100+ news sources, satellite data, weather, and macro into one dashboard, with AI to filter noise and generate signals. Goal: become India's Palantir." - question: "What is Stocky AI's Thesis 2?" answer: "Trading as a fun game — with RMG banned, gambling capital needs a new platform. Stocky Fun makes trading feel like a high-adrenaline game with AI signals, social features, and sub-15-minute trades that replace Instagram scrolling." - question: "What makes Stocky AI's founder unique?" answer: "Charandeep Kapoor: IIT Kanpur (top 0.05% JEE), hedge fund quant ($750K managed, 35% delta-neutral returns), ex-crypto founder (Timelock), 100%+ returns with Stocky AI, 100k+ weekly LinkedIn views, 12+ hours daily screen time." --- ## Special Founder - Top 0.05% in JEE, **IIT Kanpur** graduate. - Hedge fund quant. Managed $750K. **35% APY** Delta-Neutral. - Ex-founder, built something cool in crypto: **Timelock** ([Perpetual Options](https://perps.timelock.trade/)). - **100%+ returns** trading options using Stocky AI ([it works!](/blog/stocky-ai)). - **50%+ returns** trading events on Polymarket. - Team trades crypto, forex, weather, and almost every market. - 12+ hours screen time; trading is my/our hobby. - Founder-led marketing: 200k+ views on [LinkedIn](https://www.linkedin.com/posts/charandeep-kapoor_bloomberg-terminal-costs-30000year-we-activity-7435319882793041920-lbl2), 9k followers, [$2k MRR digital products.](/blog/stocky-story) --- ## Thesis 1: Democratizing Data & Compute **Problem:** Increasing Globalisation + Information velocity → Noise, Information Asymmetry → **Risk**. 1. Institutions can afford an edge in data & compute → extract money from retail. 2. 90% participants → Loss. 10% → Profitable. Only 2% from India. Rest are foreign institutions extracting capital from the country. Trading is not a discipline or a skill problem. Its a information problem. ### [Stocky Terminal](https://terminal.stockyai.xyz/) – India's Bloomberg Global Intelligence Dashboard (God's Eye View): ![Stocky Terminal — God's Eye View](/images/stocky/stocky-terminal-gods-eye.png)
- Real-time news (RSS + Scraping) - Twitter - Satellite data & trade route monitoring - Live ship, Flight tracking, supply chain disruptions. - Weather: monsoon, temperature, wind patterns - Macro: Wars, FED news, Regulatory changes
\+ AI → Filter noise + Generate signal
### Business Model Information + AI → Signal → More trades + Better risk management. - **Sub Broker**: Monetize extra volume. - **B2B API**: Data streams to hedge funds, reinsurance & consulting companies. - **AI Native Broker**: Rethink trading UX (Sahi → UX-led disruption). - **B2C Subscriptions**. ### Nationalist Goal Store, process & handle large data over time. Fine-tune AI models. - Predict floods, monsoons, earthquakes. Help emergency services. - [Defense/Intelligence](https://axionorbital.space/research) for GOI. **Become India's Palantir.** --- ## Thesis 2: Trading → Fun game/Gambling New accounts ↓. Trading is being viewed as a serious business where only top institutions make money. - **RMG out → gambling capital needs a new platform.** - Make trading a cool, fun, social game. - High adrenaline, casino-style UX (addiction loops). - App open → trade open → close in 15–30 mins. - Social capital: gamified UX, social trading, chat, share PnL. - Replace Instagram scrolls with trading. - Replace a 15-min reel session with trading. - Reduce decision fatigue to trade. - No. of trades ↑. Trade duration ↓ → Brokerage ↑ Loss ↓ - Edge in commodities: - Non-competitive market. - All-day markets vs 6 hours. - Globalisation → Trade commodities, forex & US stocks. ### [Stocky Fun](https://fun.stockyai.xyz/) ![Stocky Fun — Real-time AI Trading](/images/stocky/stocky-fun-btc.png) Real-time vision API to signal the next move (up or down). > *Note: The target of AI is not to be 100% profitable, only to aid the research-to-trade workflow and reduce it to under 1 minute.* > *Analogy: Matiks is an example of creating a new user behaviour to solve math instead of scrolling Instagram.* --- ## Proof of Work - **[Timelock Perps](https://perps.timelock.trade/)**: Everlasting Option [perps.timelock.trade]. - **Timelock Options**: Custom expiry, Custom strike options (abandoned). - **[Stocky LLM](https://llm.stockyai.xyz/)**: ChatGPT for Stock Market. - **[Voice-powered trading for Zerodha](https://www.linkedin.com/posts/charandeep-kapoor_itc-claude-zerodha-activity-7330161190741987329-Ge1d/)** - **[Stocky Analyse](https://analyse.stockyai.xyz/)**: Hedge-fund-level trade analysis with AI insights on what's working & what to improve. --- ## AI Native Hedge Fund Y Combinator put out a RFS for a AI Native Hedge Fund. ![Stocky AI — Trading Stats](/images/stocky/stocky-pnl-stats.png) Stocky stats: - 15L to 31L in 9 months - 69% win rate - Sharpe Ratio: 2.29 Given past returns and already incoming interest, can explore a PMS or a AIF CAT III in India. Will apply to YC, focus on commodities. --- # Third Stint URL: https://charandeepkapoor.com/blog/stocky-third-stint Published: 2025-12-15 --- title: "Third Stint" date: "2025-12-15" excerpt: "Stocky AI products: Terminal, LLM, Fun, Analyse — building an AI-native hedge fund for Indian markets." tags: ["stocky", "AI", "trading", "hedge-fund"] faqs: - question: "What is Stocky Terminal?" answer: "Stocky Terminal is India's first global OSINT dashboard for traders. It tracks every news headline, Twitter feed, and video for crypto, commodities, and stocks, plus maps all of India's infrastructure — ports, railways, refineries, power grids, and trade routes." - question: "What AI models does Stocky use?" answer: "Stocky uses fine-tuned Deepseek V3 and Qwen3 (multi-lingual) models focused on Indian stock market and commodities. They have access to real-time data on block deals, bulk deals, supply chain disruptions, and more." - question: "Is Stocky AI an AI-native hedge fund?" answer: "Yes, Stocky AI is building towards becoming an AI-native hedge fund, responding to Y Combinator's RFS. With a 73% win rate, 2.29 Sharpe Ratio, and 5.98 Sortino Ratio, they're exploring PMS and AIF CAT III structures in India." --- ## AI Native Hedge Fund Y Combinator put out a RFS for a AI Native Hedge Fund. ![Stocky Financial Overview](/images/stocky/stocky-financial-overview.png) Stocky stats: - 73% win rate - Win Loss Ratio: 1.66 - Sharpe Ratio: 2.29 - Sortino Ratio: 5.98 Given past returns and already incoming interest, can explore a PMS or a AIF CAT III in India Will apply to YC, focus on commodities ## Making Trading Cool & Fun ## [Stocky Terminal](https://terminal.stockyai.xyz/) India's first global OSINT dashboard. Tracks every news headline, twitter & video feed for crypto, commodities & stock market. Along with map view for all of India's infra - ports, railways, refineries, power grids and trade routes, all in 1 place. Can track earthquakes, monsoons, war updates, FED news, macro all in 1 place. Makes trading cool, imagine 4 screen setups, people tracking every event, trading commodities 12+ hours (Industry like feeling at home with your friends) ![Stocky Terminal](/images/stocky/stocky-terminal.png) ## [Stocky LLM](https://stocky-ai.vercel.app) ![Stocky AI](/images/stocky/stocky-llm-compare.png) Fine tuned models (Deepseek V3, Qwen3(multi-lingual)) with full focus on indian stock market and commodities. Access to real time data pipelines on block deals, bulk deals, supply chain disruptions etc from data sources built for Stocky Terminal. ![Stocky AI quick actions](/images/stocky/stocky-llm-buttons.png) Simple 1 button click for most requested features + backtesting tools, connect to Zerodha for portfolio analysis and live trading. ## [Stocky Fun](https://stocky-fun.vercel.app/) Bringing LLM intellegence into live video,real time vision APIs with faster than human eye response speeds. YC companies are building AI for security, OCR, sports. I just thought lets make it a consumer use case. Imagine putting a prompt for AI to give buy/sell signals based on what happening on the chart. Will convert this to a high adrenaline, gamified UX and let people trade. ![Stocky Fun](/images/stocky/stocky-fun-trading.png) Imagine replacing a 15 min insta scrolling session with 2-3 quick trades. (reduce time of trade and absolute PnL, increase overall brokerage) ## [Stocky Analyse](https://analyse.stockyai.xyz) Hedge fund level analysis for your past trading history. AI insights for what you're doing right and what you need to improve. ![Stocky Analyse](/images/stocky/stocky-analyse.png) ## [Hands Free Trading](https://www.linkedin.com/posts/charandeep-kapoor_itc-claude-zerodha-activity-7330161190741987329-Ge1d) Execute trades, backtest strategies with just your voice. (sounds cool!) --- # Stocky AI: The Money Printer URL: https://charandeepkapoor.com/blog/stocky-ai Published: 2025-09-30 --- title: "Stocky AI: The Money Printer" date: "2025-09-30" excerpt: "Fine-tuned Claude to trade Indian F&O and commodities — 15L capital, 16.57L profit, 73% win rate. Stocky AI in action." tags: ["stocky", "AI", "trading", "algo-trading"] faqs: - question: "How does Stocky AI trade?" answer: "Stocky AI uses fine-tuned Claude and Deepseek models to trade Indian F&O and commodities automatically. It analyzes weekly expiries on Bank Nifty, Nifty, and Sensex with real-time data pipelines." - question: "What are Stocky AI's trading results?" answer: "In its first year, Stocky AI generated 16.57L profit on 15L capital — 14.2L from commodities and 2.37L from F&O. It achieved a 73% win rate with 212 profitable days and only 2 loss days in the first 6 months." - question: "Can AI really trade Indian markets profitably?" answer: "Yes. Stocky AI demonstrated consistent profitability in Indian F&O and commodities markets with a Sharpe Ratio of 2.29 and Sortino Ratio of 5.98, proving AI can outperform in Indian market conditions." --- ## June 2025 Claude launched its MCP Server. Game changer for me. Fine-tuned a claude 3.0 model and gave it 15L to trade the market. Made 3.6L by september, 90k a month on auto-pilot. 3 weekly expiries (Bank Nifty on Wednesday, Nifty on Thursday and Sensex on Friday). 4 months. 51 trading days. Zero Losses. Felt invincible. ![F&O heatmap June-Sept 2025](/images/stocky/stocky-heatmap-jun-sep.png) ![P&L June-Sept 2025](/images/stocky/stocky-pnl-jun-sep.png) ## Oct 2025 - Dec 2025 Hit my first loss on 29th September. Hurt. But didn't give up. Checked the logs. Was trading too many volatile stocks (Eternal, Asian Paints, Tata Motors, BSE). Bad news hit - 3 weekly expiries (my bread and butter) was going to be lost. Bank Nifty shifted to a monthly expiry, Changed up lots of stuff, explored Deepseek. October was a blast again. Faced my second loss in November. Closed first 6 months of Stocky at 5.32L PnL, 212 profitable days, 2 loss days ![F&O heatmap June-Dec 2025](/images/stocky/stocky-heatmap-jun-dec.png) ![P&L June-Dec 2025](/images/stocky/stocky-pnl-jun-dec.png) ## Overall Commodities: 14.2L Profit Futures & Options: 2.37L Profit Total: 16.57L ![F&O overall](/images/stocky/stocky-fno-overall.png) ![Commodity overall](/images/stocky/stocky-commodity-overall.png) --- # The Stocky Story URL: https://charandeepkapoor.com/blog/stocky-story Published: 2025-06-01 --- title: "The Stocky Story" date: "2025-06-01" excerpt: "Always a trader, always a gambler. Trading since 16, building AI trading systems for Indian markets." tags: ["stocky", "trading", "AI"] faqs: - question: "What is Stocky AI?" answer: "Stocky AI is an AI-native trading platform for Indian markets built by Charandeep Kapoor. It includes a Terminal for real-time OSINT, fine-tuned LLMs for market analysis, gamified trading tools, and portfolio analytics." - question: "Who created Stocky AI?" answer: "Charandeep Kapoor, a quant trader since age 16, IIT graduate, and ex hedge fund quant who has managed $750k+ in assets. He combines AI expertise with deep trading experience." - question: "How long has Charandeep Kapoor been trading?" answer: "Charandeep has been trading since he was 16 years old, starting with live trading streams on YouTube during the 2020 Covid era, then managing mutual fund folios, and now building Stocky AI." --- Always trader/gambler at heart. Trading since 16. Explored multiple revenue/business around it: **First stint (2020 Covid Era):** Live traded and streamed on Youtube, started a blog & community on TG and Upstox referals (\$1000-\$1500 MRR) **Second Stint (2023):** Started managing MF folios (50L AUM, MFD), built multiple TG bots for news, signals, predictions (500+ users) + Youtube + Topmate (\$1500-\$2000 MRR) **Third Stint (2025):** The AI Arc: Stocky for trading, analysis, signals, news & fun. Stocky is just the culminations of all my interests.