marketDataRetrieval.py

#Prediction model using an instance of the Monte Carlo simulation and Brownian Motion equation


#import of libraries
import numpy as np
import pandas as pd
from pandas_datareader import data as wb
import matplotlib.pyplot as plt
from scipy.stats import norm

#ticker selection
def mainFunction(tradingSymbol):
    data = pd.DataFrame()
    data[tradingSymbol] = wb.DataReader(tradingSymbol, data_source='yahoo', start='2019-1-1')['Adj Close']
    #percent change of asset price
    log_returns = np.log(1+ data.pct_change())


    #graph showing growth over time beginning from 2015
    data.plot(figsize = (10,6));
    plt.show()
    #graph of log returns of input ticker
    #returns are normally distributed and have a consistent mean
    log_returns.plot(figsize = (10,6))
    plt.show()


    #calculations
    averageDailyReturn = log_returns.mean()
    variance = log_returns.var()
    drift = averageDailyReturn-(variance/2)
    standardDeviation = log_returns.std()


#Brownian Motion equation
#r = drift + standardDeviation * (e^r)

#prediction of future stock price based on simulation below using numpy for storing data into array

    np.array(drift)
    drift.values
    standardDeviation.values

    #Brownian motion variable correlating to the distance between the mean and the number of standard deviation
    norm.ppf(0.95)

    #10 x 2 Matrix
    x = np.random.rand(10,2)
    norm.ppf(x)

    #stores distances from the mean value, 0, into the 10 x 2 matrix
    Z = norm.ppf(np.random.rand(10,2))



    #time interval for the stock price forecast
    timeInterval = 365
    iterations = 5

    #r = drift + standardDeviation * (e^r)
    #10 sets of 365 random future stock prices of the ticker symbol
    dailyReturns = np.exp(drift.values + standardDeviation.values * norm.ppf(np.random.rand(timeInterval,iterations)))


    #returns into price points
    presentPrice = data.iloc[-1]
    priceList = np.zeros_like(dailyReturns)
    priceList[0] = presentPrice
    #iteration for the time interavl of 365
    for t in range(1, timeInterval):
        priceList[t] = priceList[t-1] * dailyReturns[t]

    #showcases 10 paths of the future stock price
    plt.figure(figsize =(10,6))
    plt.plot(priceList)
    plt.show()