Editing Safran SA (section)

=== Monte Carlo Simulations for Stock Price Predictions - Python Code ===
<syntaxhighlight lang="python">
"""
Created on Thu Jul 27 18:31:35 2023

@author: Jérémy Archier
"""

import csv
import matplotlib.pyplot as plt
import numpy as np
from datetime import datetime, timedelta
from scipy.stats import norm

# Step 1: Load data from the CSV file
csv_file_path = "SAF.PA.csv"

dates = []
opens = []

with open(csv_file_path, newline="") as csvfile:
    csv_reader = csv.DictReader(csvfile)

    for row in csv_reader:
        # Check if the 'Open' value is 'null' or empty or 0
        if row['Open'].strip() == '' or row['Open'].lower() == 'null' or float(row['Open']) == 0:
            # Skip the data point if there is no valid 'Open' value
            continue

        # Convert the 'Open' value to a float
        open_value = float(row['Open'])

        # Convert the 'Date' value to a datetime object
        date_value = datetime.strptime(row['Date'], '%Y-%m-%d')

        # Append the 'Date' and 'Open' values to their respective lists
        dates.append(date_value)
        opens.append(open_value)

# Step 2: Compute the logarithmic returns of the stock
log_returns = np.log(np.array(opens[1:]) / np.array(opens[:-1]))

# Plot the previous prices
plt.figure(figsize=(10, 5))
plt.plot(dates, opens, linestyle='-', color='b')
plt.xlabel('Date')
plt.ylabel('Open Value')
plt.title('Open Value as a Function of Dates')
plt.xticks(rotation=45)
plt.grid(True)
plt.tight_layout()
plt.show()

# Plot the log returns
plt.figure(figsize=(6, 5))
plt.hist(log_returns, bins=30, edgecolor='k')
plt.xlabel("Logarithmic Returns")
plt.ylabel("Frequency")
plt.title("Distribution of Logarithmic Returns")
plt.grid(True)
plt.tight_layout()
plt.show()

# Step 3: Compute the Drift
u = log_returns.mean()
var = log_returns.var()
drift = u - (0.5 * var)

# Step 4: Compute the Variance and Daily Returns
stdev = log_returns.std()
days = 50
trials = 1000000  # Reduce the number of trials to 100 for faster computation
Z = norm.ppf(np.random.rand(days, trials))  # days, trials
daily_returns = np.exp(drift + stdev * Z)

# Step 5: Calculating the stock price for every trial
price_paths = np.zeros_like(daily_returns)
price_paths[0] = opens[-1]
for t in range(1, days):
    price_paths[t] = price_paths[t - 1] * daily_returns[t]

# Step 6: Calculate the probability of reaching the target price
target_price = 150.0  # Replace this with the desired target price

# Count how many times the stock price reaches or exceeds the target price in each trial
success_count = np.sum(price_paths[-1, :] >= target_price)

# Calculate the probability as the ratio of successful trials to total trials
probability = success_count / trials

# Print the probability
print(f"The probability of the stock reaching or exceeding {target_price} is: {probability:.2%}")

# Step 7: Plot the histogram of final stock prices
final_prices = price_paths[-1, :]
plt.figure(figsize=(10, 6))
plt.hist(final_prices, bins=30, edgecolor='k', density=True)
plt.axvline(x=target_price, color='r', linestyle='dashed', linewidth=2, label=f'Target Price ({target_price})')
plt.xlabel('Stock Price')
plt.ylabel('Probability Density')
plt.title('Probability Distribution of Final Stock Prices')
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.show()

# Step 8: Plot the price paths for each trial
plt.figure(figsize=(10, 6))
for trial in range(trials):
    trial_dates = [dates[-1] + timedelta(days=i) for i in range(days)]
    plt.plot(trial_dates, price_paths[:, trial], linewidth=0.5)
plt.xlabel('Date')
plt.ylabel('Stock Price')
plt.title('Monte Carlo Simulation of Stock Price')
plt.tight_layout()
plt.grid(True)
plt.show()
</syntaxhighlight>