Learn Challenge: Implement a Volatility-Based Stop-Loss | Building and Evaluating Trading Strategies

In trading, volatility-based stop-loss strategies provide a dynamic way to manage risk by adjusting stop-loss levels according to recent price fluctuations. This approach helps you avoid being stopped out by normal market noise while still protecting your capital during periods of increased volatility. You will now learn how to implement such a strategy using a hardcoded set of closing prices, calculate the rolling standard deviation as a measure of volatility, and determine stop-loss levels that adapt to changing market conditions.


              12345678910111213141516171819202122232425262728
            
import pandas as pd
import numpy as np

# Hardcoded closing prices for demonstration
data = {
    "close": [100, 102, 101, 99, 98, 100, 103, 105, 104, 102, 101, 99, 97, 98, 100]
}
df = pd.DataFrame(data)

# Calculate 3-day rolling standard deviation (volatility)
df["volatility"] = df["close"].rolling(window=3).std()

# Set stop-loss at 2 times rolling volatility below current close
df["stop_loss"] = df["close"] - 2 * df["volatility"]

# Simulate next day's low prices (for demonstration, use close minus random noise)
np.random.seed(42)
df["low"] = df["close"] - np.abs(np.random.normal(0.5, 0.7, size=len(df)))

# Shift stop-loss by one day to compare with next day's low
df["stop_loss_prev"] = df["stop_loss"].shift(1)

# Identify if stop-loss would have been triggered (next day's low below previous day's stop-loss)
df["stop_triggered"] = df["low"] < df["stop_loss_prev"]

# Show relevant columns
result = df[["close", "volatility", "stop_loss", "low", "stop_loss_prev", "stop_triggered"]]
print(result)

This code calculates the rolling volatility and sets a volatility-based stop-loss for each day. The stop-loss is placed at a distance of two times the recent 3-day rolling standard deviation below the closing price, making it responsive to market conditions. By comparing the next day's low to the previous day's stop-loss, you can see on which dates your stop-loss would have been triggered, helping you evaluate the effectiveness of this risk management technique.

Task

Swipe to start coding

Implement a volatility-based stop-loss strategy as described:

Create a DataFrame with a hardcoded list of closing prices.
Calculate the 3-day rolling standard deviation ("volatility") of the closing prices.
For each day, compute a stop-loss price set at 2 times the rolling volatility below the closing price.
Simulate next day's low prices by subtracting random positive noise from each close.
For each day, compare the next day's low to the previous day's stop-loss price, and create a column stop_triggered that is True if the stop-loss would have been hit.
Output a DataFrame showing the close, volatility, stop_loss, low, stop_loss_prev, and stop_triggered columns.

Your solution will be checked for correct calculation of rolling volatility, stop-loss levels, and trigger identification.

Solution

Everything was clear?

Thanks for your feedback!

Section 3. Chapter 5

single

Ask AI

Ask anything or try one of the suggested questions to begin our chat

Suggested prompts:

Can you explain how the rolling standard deviation is calculated in this context?

What does it mean when the stop-loss is triggered in this strategy?

How can I adjust the sensitivity of the stop-loss to market volatility?

Swipe to show menu


              12345678910111213141516171819202122232425262728
            
import pandas as pd
import numpy as np

# Hardcoded closing prices for demonstration
data = {
    "close": [100, 102, 101, 99, 98, 100, 103, 105, 104, 102, 101, 99, 97, 98, 100]
}
df = pd.DataFrame(data)

# Calculate 3-day rolling standard deviation (volatility)
df["volatility"] = df["close"].rolling(window=3).std()

# Set stop-loss at 2 times rolling volatility below current close
df["stop_loss"] = df["close"] - 2 * df["volatility"]

# Simulate next day's low prices (for demonstration, use close minus random noise)
np.random.seed(42)
df["low"] = df["close"] - np.abs(np.random.normal(0.5, 0.7, size=len(df)))

# Shift stop-loss by one day to compare with next day's low
df["stop_loss_prev"] = df["stop_loss"].shift(1)

# Identify if stop-loss would have been triggered (next day's low below previous day's stop-loss)
df["stop_triggered"] = df["low"] < df["stop_loss_prev"]

# Show relevant columns
result = df[["close", "volatility", "stop_loss", "low", "stop_loss_prev", "stop_triggered"]]
print(result)