Creating Strategies

Learn how to effectively communicate your trading ideas to Lona’s AI and get the strategies you want.

The Interview Process

When you start a new chat, you’ll interact with Lona’s Interview Agent - an AI specialized in understanding trading strategies. Think of it as having a conversation with an experienced trader who asks the right questions.

How the Interview Works

1. You describe your initial idea in natural language
2. Lona asks clarifying questions to understand the details
3. You provide more information about indicators, rules, and logic
4. Lona confirms understanding by summarizing what it learned
5. You click “Extract Requirements” when ready to proceed

Key Insight: The Interview Agent focuses ONLY on understanding your strategy logic. It won’t write code or discuss implementation details - that comes later.

What to Tell Lona

Essential Information

Every trading strategy needs these core elements:

1. Strategy Type/Concept

Describe the overall approach or philosophy:

Examples:

• “A trend-following strategy”
• “A mean reversion strategy”
• “A breakout strategy”
• “A momentum-based approach”
• “A volatility trading strategy”

2. Technical Indicators

Specify which indicators you want to use:

Examples:

• “Use a 20-period simple moving average”
• “Calculate RSI with a 14-period lookback”
• “Compare fast (12) and slow (26) exponential moving averages”
• “Use Bollinger Bands with 2 standard deviations”
• “Apply MACD with standard settings”

3. Entry Conditions

Define when to open a position:

Examples:

• “Buy when price crosses above the moving average”
• “Enter long when RSI drops below 30 and then rises above it”
• “Open a position when fast MA crosses above slow MA”
• “Buy when price breaks out above the upper Bollinger Band”
• “Enter when MACD line crosses above the signal line”

4. Exit Conditions

Define when to close a position:

Examples:

• “Sell when price crosses below the moving average”
• “Exit when RSI reaches 70”
• “Close position when fast MA crosses below slow MA”
• “Exit when price returns inside the Bollinger Bands”
• “Close when MACD line crosses below the signal line”

Optional Information

You can also specify (but don’t have to):

• Risk Management: Stop losses, take profits, position sizing
• Multiple Conditions: “AND/OR” logic for entry/exit
• Time-based Rules: Only trade during certain periods
• Custom Logic: Any specific rules unique to your strategy

Important: Lona creates market-agnostic strategies that work on any timeframe. You don’t need to specify the market (stocks, forex, crypto) or timeframe (daily, hourly) - these are configured during execution.

Communication Tips

✅ Good Strategy Descriptions

Example 1: Clear and Specific

"I want to create a moving average crossover strategy. Use a 10-period fast 
moving average and a 30-period slow moving average. Buy when the fast MA 
crosses above the slow MA. Sell when the fast MA crosses below the slow MA."

Why it works: Specific periods, clear entry/exit rules, unambiguous logic.

Example 2: Detailed RSI Strategy

"Create a mean reversion strategy using RSI. When RSI falls below 30, 
that signals oversold conditions. Buy when RSI crosses back above 30. 
Exit the position when RSI reaches 70 (overbought). Use a 14-period RSI."

Why it works: Explains the reasoning, defines clear thresholds, specifies the indicator period.

Example 3: Multi-Indicator Approach

"I want a trend-following strategy that combines two indicators. Use a 
50-period moving average to identify the trend direction. Use MACD (12,26,9) 
for entry timing. Only take long positions when price is above the MA AND 
MACD line crosses above the signal line. Exit when MACD crosses back below 
the signal line."

Why it works: Clear combination logic, specific parameters, well-defined entry/exit.

❌ Unclear Strategy Descriptions

Example 1: Too Vague

"Make a trading strategy with moving averages that makes money."

Problem: No specific periods, no entry/exit rules, no clear logic.

How to improve: Specify MA periods, define when to buy/sell, explain the rules.

Example 2: Missing Critical Details

"Use RSI to buy low and sell high."

Problem: No RSI period, no specific thresholds, no clear entry/exit triggers.

How to improve: Specify RSI period (e.g., 14), define thresholds (e.g., below 30 = buy signal), explain exact entry/exit conditions.

Example 3: Implementation Details

"Write a Python script that uses pandas to calculate a 20-day SMA and 
plots it on a chart."

Problem: Focuses on implementation (code, libraries) rather than strategy logic.

How to improve: Describe what the strategy should DO, not HOW it should be implemented technically.

Sample Conversations

Here are realistic examples of effective conversations with Lona:

Example 1: Simple Moving Average

You: “I want to create a simple moving average crossover strategy”

Lona: “Great! A moving average crossover is a classic strategy. To help me understand your approach better:

• What periods would you like for your moving averages? (e.g., 10 and 30, or 50 and 200)
• Should I use simple moving averages (SMA) or exponential moving averages (EMA)?
• What are your entry and exit rules?”

You: “Use simple moving averages with periods of 20 and 50. Buy when the 20 crosses above the 50. Sell when the 20 crosses below the 50.”

Lona: “Perfect! Here’s what I understand:

• Fast SMA: 20 periods
• Slow SMA: 50 periods
• Entry Signal: 20 SMA crosses above 50 SMA (bullish crossover)
• Exit Signal: 20 SMA crosses below 50 SMA (bearish crossover)

This is a straightforward trend-following strategy. Would you like to add any additional conditions, or shall we extract the requirements?”

You: “That looks good, let’s extract the requirements.”

Example 2: RSI Mean Reversion

You: “I want to trade based on RSI when it shows oversold and overbought conditions”

Lona: “An RSI mean reversion strategy - excellent choice! To design this properly, I need to understand:

• What RSI period would you like to use? (14 is standard)
• What do you consider oversold and overbought levels? (commonly 30 and 70)
• Do you want to enter immediately when RSI crosses these levels, or wait for confirmation?”

You: “Use a 14-period RSI. I consider oversold as RSI below 30 and overbought as RSI above 70. Buy when RSI rises above 30 after being below it. Sell when RSI falls below 70 after being above it.”

Lona: “Excellent detail! So your strategy is:

• Indicator: RSI with 14-period lookback
• Oversold Level: 30
• Overbought Level: 70
• Entry: Long when RSI crosses above 30 (confirming oversold recovery)
• Exit: Close position when RSI crosses below 70 (leaving overbought territory)

This waits for RSI to recover from extremes rather than entering immediately. Is this correct?”

You: “Yes, exactly!”

Lona: “Great! I have all the details I need. Please click ‘Extract Requirements’ to proceed to code generation.”

Example 3: Multi-Indicator Strategy

You: “Create a strategy that uses moving averages for trend and RSI for entry timing”

Lona: “Interesting! A multi-indicator approach combining trend identification and momentum timing. Let me understand your logic:

• For the moving average trend filter: What period and type? What defines an uptrend vs downtrend?
• For RSI timing: What period and what signals trigger entries?
• How do these work together? For example: Only buy when in uptrend AND RSI gives a signal?”

You: “Use a 50-period SMA for trend. If price is above the SMA, we’re in an uptrend. Use 14-period RSI for timing. Only take long positions when price is above the SMA AND RSI crosses above 30. Exit when price falls below the SMA OR RSI crosses above 70.”

Lona: “Perfect multi-condition strategy! Let me confirm:

Trend Filter:

• 50 SMA determines trend direction
• Price above SMA = uptrend

Entry Conditions (ALL must be true):

• Price above 50 SMA (uptrend confirmed)
• RSI crosses above 30 (oversold recovery)

Exit Conditions (ANY can trigger exit):

• Price crosses below 50 SMA (trend change)
• RSI crosses above 70 (overbought)

This combines trend-following with mean reversion timing. Sound good?”

You: “Perfect! Let’s proceed.”

Strategy Types You Can Create

Lona can help you create many types of strategies:

Trend-Following

• Moving average crossovers
• Breakout strategies
• Momentum strategies
• Channel breakouts

Mean Reversion

• RSI oversold/overbought
• Bollinger Band reversals
• Support/resistance bounces
• Statistical mean reversion

Momentum

• MACD crossovers
• Rate of change strategies
• Relative strength strategies

Volatility-Based

• Bollinger Band squeezes
• ATR-based strategies
• Volatility breakouts

Multi-Indicator

• Combinations of any indicators
• Complex entry/exit logic
• Multiple confirmation signals

Common Indicators Explained

If you’re not sure which indicators to use, here’s a quick reference:

Tip: Start with one or two indicators. You can always create variations with more complexity later.

Framework Selection

When you click “Extract Requirements,” Lona automatically selects a trading framework for your strategy.

What’s a framework? It’s the underlying system that executes your strategy logic - you don’t need to understand it! Lona handles this automatically.

Note: In development mode, you might see the framework name displayed (e.g., “EventDriven v3.0”). This is for testing purposes and doesn’t affect how you use Lona.

Moving to Code Generation

Once Lona understands your strategy, you’ll see the Extract Requirements button appear.

When you click it:

1. Lona processes your entire conversation
2. Creates a structured list of requirements
3. Displays requirements in the Canvas panel on the right
4. Shows buttons for code generation: Generate Fast and Generate Advanced

You’re now ready to move to the next phase! Continue to Requirements & Code Generation.

Tips for Effective Interviews

✅ Be Conversational: Talk naturally, as if explaining to a colleague

✅ Think Step-by-Step: Entry → Exit → Special conditions

✅ Use Examples: “Like when price bounces off support” helps clarify

✅ Ask for Suggestions: If unsure, ask “What indicators work well for trend following?”

✅ Iterate: You can always create new versions of your strategy

❌ Don’t worry about: Markets, timeframes, data sources, coding, implementation

❌ Don’t include: Code syntax, library names, programming concepts

❌ Don’t rush: Take time to think through your logic

What If Lona Doesn’t Understand?

If Lona misunderstands your strategy:

1. Rephrase: Try explaining differently or more simply
2. Break It Down: Explain one part at a time
3. Use Analogies: “Similar to a moving average crossover, but with RSI”
4. Be Explicit: State exactly when to buy and when to sell
5. Provide Examples: “For example, if RSI is 25 and rises to 32, that’s a buy signal”

Remember: Lona is specialized in trading and understands industry terminology, so don’t hesitate to use proper trading language!

Next Steps: Learn about Requirements & Code Generation to understand what happens after the interview.