Part 3 | How To Improve Your Trading With ChatGPT
Want Professional Level Trading Answers From Chat GPT? Ask Professional Level Questions!
Before we dive into today’s trading lesson, I’m holding a free trading webinar on Tuesday, February 25, on How To Accomplish In 20 Minutes With ChatGPT What Takes A Team Of 25-40 Professionals At A MultiBillion-Dollar Hedge Fund 4-6 Weeks To Do.
As traders, we’re always looking for professional-level answers to our trading questions.
Whether it’s for trading ideas, trading strategy development, portfolio optimization, and even something as simple as which options strike price and expiration to select when we buy or sell options, we’d prefer our answer to come from someone who has dozens of years of trading experience to answer us, rather than someone who’s only been trading part-time for a few months.
This is making the obvious, obvious.
Yes, when it comes to using AI for trading, especially with ChatGPT, the overwhelming number of traders who use it ask questions that will deliver pedestrian, very basic answers instead of what AI has inside of it…literally quant hedge fund level knowledge.
In Lesson One, we learned: Tell ChatGPT who it is.
In Lesson Two, we learned: Tell ChatGPT who you are.
In today’s Lesson, I’ll teach you step 3; how to prompt ChatGPT with questions that will give you far better and higher-level answers simply by asking a better question.
The following examples will provide you with this knowledge. As you can see, basic questions, the ones often written by the majority of traders when it comes to using ChatGPT, lead to very basic answers.
Now, compare them to the professional-level questions. The answers are far deeper, and you gain far more knowledge and insight.
Why? The questions contain much more “context.” They are precise, and they consistently lead to answers that are more professional.
Here are a handful of examples:
1. How to Identify the Best Day Trades?
Basic Prompt:
“What are the best stocks to day trade?”
Weak Answer:
• Look for high-volume stocks with strong momentum.
• Stocks with news catalysts tend to be good for trading.
• Tech and small-cap stocks often provide opportunities.
Better Prompt (Rated 10):
“Using an institutional-level approach, identify the three highest-probability day trading setups based on order flow, liquidity absorption, and volatility regimes. Provide a tactical execution plan for each.”
Why It’s Better:
• Introduces order flow (used by hedge funds & HFT firms).
• Adds liquidity absorption (critical for spotting smart money).
• Requires specific execution tactics (not just ideas).
Pro-Level Answer From ChatGPT:
Setup 1: Opening Drive Momentum Trap
• Criteria: Stock gaps up >2% but shows early large ask-side liquidity blocks on the order book.
• Execution: Wait for first 1-minute fake breakout above premarket highs, then short into the liquidity wall with stop-loss at new highs.
• Exit: Cover at VWAP or when bid-side liquidity increases.
Setup 2: VWAP Reclaim + Hidden Buyer Accumulation
• Criteria: Stock sells off but stalls near VWAP with high-lot iceberg orders appearing on the bid.
• Execution: Enter long on bid-side absorption with tight stop below VWAP.
• Exit: Take profits at previous session’s high with scaling.
Setup 3: Late-Day Liquidity Sweep into Reversal
• Criteria: At 3 PM, stock makes a fake breakout, triggering stops before reversing.
• Execution: Use Level 2 tape reading to confirm aggressive buying after the sweep.
• Exit: Ride trend into the close, trailing stop behind higher lows.
Note: You can see it’s suggesting professional-level set-ups compared to the first answer which was pedestrian.
The above requires you to integrate your data into ChatGPT. If you don’t have the ability to do so you can ask it to create strategies that don’t require data. The strategy ideas will likely be less professional in nature – most professional intra-day trading firms do use this type of data and more, so know who you are competing against.
The bottom line is by asking a more professional question, which included using the term “institutional level approach” in your prompt, it provided you with an institutional level answer.
2. How to Manage Risk?
Basic Prompt:
“How do I manage risk when day trading?”
Weak Answer:
• Use stop-loss orders.
• Risk only 1-2% per trade.
• Stick to a trading plan.
Better Prompt (Rated 10):
“Compare volatility-adjusted position sizing with Kelly Criterion for a $250K professional trading account. Include real-time market volatility considerations and execution constraints.”
Why It’s Better:
• Volatility-adjusted sizing ensures stable % risk across assets.
• Kelly Criterion adds mathematical risk optimization.
• Real-time market conditions force contextual adaptability.
ChatGPT Pro-Level Answer:
1. Volatility-Adjusted Position Sizing
• Calculate 10-day ATR for each asset.
• Position size = (Max risk per trade) ÷ (ATR * multiplier).
• Reduces overexposure in high-volatility environments.
2. Kelly Criterion for High-Win Rate Strategies
• Kelly Formula: f* = Edge / Variance.
• If a strategy wins 65% of trades with a 2:1 reward-to-risk, optimal bet sizing = 0.13 (or 13% of capital per trade).
• This method scales dynamically with performance.
3. Execution Constraints
• If spreads widen by 2x ATR, reduce size by 50%.
• If implied vol jumps >30% (e.g., news event), cut exposure by 75%.
Note – The initial prompt is weak. It’s too basic and it’s pedestrian. So it leads to a weak, basic, pedestrian response.
The professional prompt leads to a far better answer that even uses Kelly Criterion to calculate your position sizing, just as the professionals do.
3. How to Trade the Market Open?
Basic Prompt:
“What’s the best way to trade the open?”
Weak Answer:
• Trade high-volume stocks.
• Look for breakouts.
• Watch premarket price action.
Better Prompt (Rated 10):
“Break down the institutional trading dynamics in the first 15 minutes of the market open, incorporating dark pool activity, high-frequency trading (HFT) impact, and opening range psychology.”
Why It’s Better:
• Introduces dark pool influence on price action.
• Accounts for HFT behavior, crucial for fast-moving traders.
• Explores psychology of the first 15 minutes.
ChatGPT Pro-Level Answer:
1. Market Open Institutional Mechanics
• Dark pools clear premarket imbalance at 9:28-9:30 AM, often causing misleading prints.
• HFT firms deploy latency arbitrage strategies, exploiting bid/ask imbalances before retail reacts.
• Large funds use iceberg orders to manipulate liquidity zones.
• Scalp Long: Identify “fake flushes” where large bid absorption occurs before a sharp reversal.
• Fade Gaps: If a stock gaps up >5% on no significant news, sell short into mean-reversion liquidity zones.
Note – The professional prompt leads to the professional-level answer.
4. How to Trade Breakouts?
Basic Prompt:
“How do I trade breakouts?”
Weak Answer:
• Buy above resistance.
• Wait for confirmation.
• Use stop-loss orders.
Better Prompt (Rated 10):
“Develop a high-probability breakout model incorporating liquidity voids, hidden buyer activity, and algorithmic stop-hunting patterns. Provide a real-market execution example.”
Why It’s Better:
• Liquidity voids increase breakout success rate.
• Hidden buyer activity helps avoid fake breakouts.
• Algo stop-hunting exposes manipulation risks.
ChatGPT Pro-Level Answer:
1. Spotting Institutional Breakouts
• Liquidity voids occur when volume dries up in a tight range. This means big money is not actively selling.
• Watch for hidden buyers stacking bids below the breakout level.
2. Execution Tactics
• Entry: Use a stop-limit order just above resistance.
• Exit: Scale out in thirds (33% at 1R, 33% at 2R, 33% on trailing stop).
• TSLA trading at $197 resistance with clear liquidity void above.
• Hidden bids at $196.80-197.00 indicate a buyer is absorbing sellers.
• Breakout occurs, pushing price to $205 within 10 minutes.
Note – The better prompt with more professional context is leading to a better suggestion.
Also, you can have it suggest multiple strategies for you.
Please note that just because ChatGPT puts a strategy in front of you, it does not mean it will make money.
It’s making higher-level suggestions, and from there, you should backtest the ideas to see how they have historically, hypothetically performed.
Also note that just because a strategy back-tested well, it absolutely doesn’t mean it will perform that way in live trading.
For this lesson, our focus is not on building strategies – it’s on moving ChatGPT from “basic mode” where it starts, to a high-level mode where the greater knowledge is within it.
Note – The better prompt with more professional context is leading to a better suggestion.
5. How to Exit a Trade?
Basic Prompt:
“When should I exit a trade?”
Weak Answer:
• Exit when you hit the profit target.
• Use a trailing stop.
• Close the trade before the market closes.
Better Prompt (Rated 10):
“Optimize trade exits using AI-driven momentum models that adjust for market volatility shifts, bid-ask depth, and trend exhaustion signals.”
Why It’s Better:
• AI-driven models remove emotion.
• Market volatility shifts force dynamic exit adjustments.
• Trend exhaustion signals prevent holding too long.
ChatGPT Pro-Level Answer:
• Uses real-time IV skew data to measure whether to hold or exit.
• Exit early if volatility mean-reverts significantly.
• If liquidity starts thinning on the bid side, exit immediately.
• If iceberg orders stack up against your trade, reduce size.
Final Thoughts
These “advanced level prompts” force ChatGPT to act like a professional quant researcher, institutional trader, or hedge fund analyst, ensuring maximum insight extraction.
Learning to prompt like this is a skill. It doesn’t come overnight, but it’s easy to learn with repeated practice.
Taking the time to learn this skill will have an exponential impact on your trading knowledge, education, and ideally on your trading profits.
In my next Lesson, we’ll continue to grow your trading knowledge with ChatGPT.
Special Note – Register below if you’d like to attend a free webinar I’m teaching on Tuesday, February 25 at 1 pm ET:
In the webinar, I’ll show you how in just 20 minutes, using ChatGPT, I achieved the same results, including building a professional-grade long/short strategy, that would have taken a team of 25 to 40 hedge fund professionals weeks to accomplish just a few years ago.
Please note that Zoom handles 500 people and the webinar is nearly filled. If you’re interested in attending, please register here now.
For Every Issue of Larry Connors’ Trading Lesson Of The Day, Click Here.