Exponential Moving Average in Python

Learn how to calculate one of the most popular technical analysis measurements in python! Similar to the Simple Moving Average (SMA), the Exponential Moving Average (EMA) is a rolling average value that adds some spice! The EMA has additional weight on values closer to the present day to increase the importance of most recent data. This blog post will show you how to perform Exponential Moving Averages in Python!

Import Needed Libraries

import yfinance as yf 

import pandas as pd 

import matplotlib.pyplot as plt 

from datetime import timedelta, datetime, date

Create a Custom Yahoo Finance GET Data Function

def GetYahooFinanceData(ticker, start, end):

    formatDate = "%Y-%m-%d"

    # Add 1 day to end date

    end = datetime.strptime(end, formatDate)

    end = end + timedelta(days = 1) #add one day to end date to get desired result

    end = end.strftime(formatDate) #converts back to a string type date 

    df = yf.download(ticker, start, end)

    return df 

symbol = 'AAPL'

stock_data = GetYahooFinanceData(symbol,'2022-01-01','2024-02-15')

stock_data

[*********************100%%**********************] 1 of 1 completed

Out[1]:

	Open	High	Low	Close	Adj Close	Volume
Date
2022-01-03	177.830002	182.880005	177.710007	182.009995	179.724564	104487900
2022-01-04	182.630005	182.940002	179.119995	179.699997	177.443573	99310400
2022-01-05	179.610001	180.169998	174.639999	174.919998	172.723587	94537600
2022-01-06	172.699997	175.300003	171.639999	172.000000	169.840240	96904000
2022-01-07	172.889999	174.139999	171.029999	172.169998	170.008118	86709100
...	...	...	...	...	...	...
2024-02-09	188.649994	189.990005	188.000000	188.850006	188.850006	45155200
2024-02-12	188.419998	188.669998	186.789993	187.149994	187.149994	41781900
2024-02-13	185.770004	186.210007	183.509995	185.039993	185.039993	56529500
2024-02-14	185.320007	185.529999	182.440002	184.149994	184.149994	54630500
2024-02-15	183.550003	184.490005	181.350006	183.860001	183.860001	65434500

533 rows × 6 columns

Now Calculate the EMA!

# Calculate EMA12, EMA128 EMA200 

stock_data['EMA12'] = stock_data['Close'].ewm(span=12, min_periods=0, adjust=False).mean() 

stock_data['EMA128'] = stock_data['Close'].ewm(span=128, min_periods=0, adjust=False).mean() 

stock_data['EMA200'] = stock_data['Close'].ewm(span=200, min_periods=0, adjust=False).mean() 

stock_data

	Open	High	Low	Close	Adj Close	Volume	EMA12	EMA128	EMA200
Date
2022-01-03	177.830002	182.880005	177.710007	182.009995	179.724564	104487900	182.009995	182.009995	182.009995
2022-01-04	182.630005	182.940002	179.119995	179.699997	177.443573	99310400	181.654610	181.974181	181.987009
2022-01-05	179.610001	180.169998	174.639999	174.919998	172.723587	94537600	180.618516	181.864813	181.916691
2022-01-06	172.699997	175.300003	171.639999	172.000000	169.840240	96904000	179.292591	181.711871	181.818017
2022-01-07	172.889999	174.139999	171.029999	172.169998	170.008118	86709100	178.196807	181.563935	181.722017
...	...	...	...	...	...	...	...	...	...
2024-02-09	188.649994	189.990005	188.000000	188.850006	188.850006	45155200	188.726492	184.599864	180.690855
2024-02-12	188.419998	188.669998	186.789993	187.149994	187.149994	41781900	188.483954	184.639401	180.755125
2024-02-13	185.770004	186.210007	183.509995	185.039993	185.039993	56529500	187.954114	184.645612	180.797760
2024-02-14	185.320007	185.529999	182.440002	184.149994	184.149994	54630500	187.368865	184.637928	180.831116
2024-02-15	183.550003	184.490005	181.350006	183.860001	183.860001	65434500	186.829039	184.625867	180.861254

533 rows × 9 columns

Why Stop Here? Automate Buy and Sell Signals!

# Initialize buy and sell columns 

stock_data['Buy_Signal'] = 0 

stock_data['Sell_Signal'] = 0

#iterate through the dataframe using a for loop 

for i in range(1,len(stock_data)):

    if (stock_data['EMA12'][i] > stock_data['EMA128'][i]) and (stock_data['EMA12'][i-1] <= stock_data['EMA128'][i-1]):

        stock_data['Buy_Signal'][i] = 1 

    elif (stock_data['EMA12'][i] < stock_data['EMA128'][i]) and (stock_data['EMA12'][i-1] >= stock_data['EMA128'][i-1]):

        stock_data['Sell_Signal'][i] = -1

Visualize Your New Trading Algorithm!

plt.figure(figsize=(20,8))

#Lines - Close and EMA Lines

plt.plot(stock_data.index, stock_data['Close'], label='Close Price', color = 'black')

plt.plot(stock_data.index, stock_data['EMA12'], label='EMA12', color = 'blue')

plt.plot(stock_data.index, stock_data['EMA128'], label='EMA128', color = 'red')

plt.plot(stock_data.index, stock_data['EMA200'], label='EMA200', color = 'green')

# Plot buy and sell signals 

buy_signals = stock_data[stock_data['Buy_Signal'] == 1]

plt.scatter(buy_signals.index, buy_signals['Close'], marker='^', color='g', label='Buy Signal', s=100) 

sell_signals = stock_data[stock_data['Sell_Signal'] == -1]

plt.scatter(sell_signals.index, sell_signals['Close'], marker='v', color='r', label='Sell Signal', s=100)

plt.title(f'{symbol} Price with EMA and Trading Signals')

plt.xlabel('Date')

plt.ylabel(f'{symbol} Price')

plt.legend() 

plt.show()

EMA Lines for AAPL Stock

But How Does the Algorithm Perform?

initial_capital = 10000

position = 0

capital = initial_capital

bought_and_held_capital = capital 

bought_and_held_shares = capital / stock_data['Close'][0]

#Simulate trading 

for i in range(len(stock_data)):

    if stock_data['Buy_Signal'][i] == 1:

        #Buy 10,000 worth of the asset 

        shares_to_buy = capital / stock_data['Close'][i] 

        position += shares_to_buy 

        capital -= shares_to_buy * stock_data['Close'][i]

    elif stock_data['Sell_Signal'][i] == -1:

        #Sell all shares 

        capital += position * stock_data['Close'][i]

        position = 0 

    # Update bought and held strategy 

    bought_and_held_capital = bought_and_held_shares * stock_data['Close'][i] 

#Calculate final value 

algorithm_final_value = capital + position * stock_data['Close'][-1]

bought_and_held_final_value = bought_and_held_capital

algorithm_final_percent = (algorithm_final_value - initial_capital)/initial_capital

bought_and_held_final_percent = (bought_and_held_final_value - initial_capital)/initial_capital

# Print results

print("Algorithm Final Value: $", algorithm_final_value, "%: ", round(algorithm_final_percent*100,1))

print("Bought and Held Final Value: $", bought_and_held_final_value, "%: ", round(bought_and_held_final_percent*100,1))

Algorithm Final Value: $ 10217.106243115186 %: 2.2 Bought and Held Final Value: $ 10101.643105288162 %: 1.0

We see the algorithm slightly performing better on AAPL stock compared to a buy and held position with the algorithm up 2.2% while AAPL itself is up 1%. The algorithm is not perfect and not expected to outperform in each scenario. I hope you enjoy the EMA algorithm and the code! Happy trading!

*Intrendias is not financial advice.