Stock Market Analysis and Buy Signal Design - Part 2

Abhishek Srivastava

2024-08-04

R Markdown

Introduction
In the stock market, identifying optimal buying opportunities is crucial for making
profitable investment decisions. This project aims to design a Buy Signal Filter using data
science techniques in R, leveraging key technical indicators to help investors pinpoint
potential buy signals.

Data Importing and Cleaning

How to import and clean my data
First, we need to import the necessary packages and collect historical stock price data for
Apple Inc. (AAPL), Microsoft Corp. (MSFT), and Amazon.com Inc. (AMZN) from Yahoo
Finance.

library(quantmod)
quantmod is a popular R package used for quantitative financial modeling and trading. It
provides a range of functions and tools for quantitative analysis of financial data,
particularly in the context of stock market analysis and trading strategies.

library(TTR)
The TTR (Technical Trading Rules) library is a package in R that provides a collection of
technical analysis functions for financial time series data. TTR is often used in conjunction
with other packages like quantmod for analyzing and visualizing financial data.
# Load necessary libraries
library(quantmod)

## Warning: package 'quantmod' was built under R version 4.4.1

## Loading required package: xts

## Warning: package 'xts' was built under R version 4.4.1

## Loading required package: zoo

## Warning: package 'zoo' was built under R version 4.4.1

##
## Attaching package: 'zoo'

## The following objects are masked from 'package:base':

##
## as.Date, as.Date.numeric

## Loading required package: TTR

## Warning: package 'TTR' was built under R version 4.4.1

## Registered S3 method overwritten by 'quantmod':

## method from
## as.zoo.data.frame zoo

library(TTR)
library(dplyr)

## Warning: package 'dplyr' was built under R version 4.4.1

##
## ######################### Warning from 'xts' package
##########################
## #
#
## # The dplyr lag() function breaks how base R's lag() function is supposed
to #
## # work, which breaks lag(my_xts). Calls to lag(my_xts) that you type or
#
## # source() into this session won't work correctly.
#
## #
#
## # Use stats::lag() to make sure you're not using dplyr::lag(), or you can
add #
## # conflictRules('dplyr', exclude = 'lag') to your .Rprofile to stop
#
## # dplyr from breaking base R's lag() function.
#
## #
#
## # Code in packages is not affected. It's protected by R's namespace
mechanism #
## # Set `options(xts.warn_dplyr_breaks_lag = FALSE)` to suppress this
warning. #
## #
#
##
#############################################################################
##
##
## Attaching package: 'dplyr'

## The following objects are masked from 'package:xts':

##
## first, last

## The following objects are masked from 'package:stats':

##
## filter, lag

## The following objects are masked from 'package:base':

##
## intersect, setdiff, setequal, union

# Set options to avoid scientific notation

options(scipen = 999)

# Download stock data

getSymbols(c("AAPL", "MSFT", "AMZN"), src = "yahoo", from = "2020-01-01", to
= "2024-07-26")

## [1] "AAPL" "MSFT" "AMZN"

# View the structure of the data

str(AAPL)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 6]
## Columns: AAPL.Open, AAPL.High, AAPL.Low, AAPL.Close, AAPL.Volume ...
with 1 more column
## Index: Date [1148] (TZ: "UTC")
## xts Attributes:
## $ src : chr "yahoo"
## $ updated: POSIXct[1:1], format: "2024-08-04 15:46:55"

str(MSFT)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 6]
## Columns: MSFT.Open, MSFT.High, MSFT.Low, MSFT.Close, MSFT.Volume ...
with 1 more column
## Index: Date [1148] (TZ: "UTC")
## xts Attributes:
## $ src : chr "yahoo"
## $ updated: POSIXct[1:1], format: "2024-08-04 15:46:56"

str(AMZN)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 6]
## Columns: AMZN.Open, AMZN.High, AMZN.Low, AMZN.Close, AMZN.Volume ...
with 1 more column
## Index: Date [1148] (TZ: "UTC")
## xts Attributes:
## $ src : chr "yahoo"
## $ updated: POSIXct[1:1], format: "2024-08-04 15:46:56"

Data Cleaning Steps

1. Remove NA Values: Ensure no missing values in the closing prices.
2. Calculate Indicators: Compute 50-day and 200-day SMA, RSI, and MACD for AAPL,
MSFT, and AMZN.
3. Create a Clean Data Frame: Combine all relevant indicators into a single clean data
frame for each stock.
library(dplyr)

# Convert AAPL xts object to a data frame

AAPL_df <- data.frame(date = index(AAPL), coredata(AAPL))

# Perform operations using dplyr

AAPL_df <- AAPL_df %>%
mutate(AAPL.Return = (AAPL_df$AAPL.Close - AAPL_df$AAPL.Open) /
AAPL_df$AAPL.Open)

# Convert the modified data frame back to xts

AAPL_modified <- xts(AAPL_df[, -1], order.by = as.Date(AAPL_df$date))

# Convert MSFT xts object to a data frame

MSFT_df <- data.frame(date = index(MSFT), coredata(MSFT))

# Perform operations using dplyr

MSFT_df <- MSFT_df %>%
mutate(MSFT.Return = (MSFT_df$MSFT.Close - MSFT_df$MSFT.Open) /
MSFT_df$MSFT.Open)

# Convert the modified data frame back to xts

MSFT_modified <- xts(MSFT_df[, -1], order.by = as.Date(MSFT_df$date))

# Convert AMZN xts object to a data frame

AMZN_df <- data.frame(date = index(AMZN), coredata(AMZN))

# Perform operations using dplyr

AMZN_df <- AMZN_df %>%
mutate(AMZN.Return = (AMZN_df$AMZN.Close - AMZN_df$AMZN.Open) /
AMZN_df$AMZN.Open)

# Convert the modified data frame back to xts

AMZN_modified <- xts(AMZN_df[, -1], order.by = as.Date(AMZN_df$date))
# Indicator Calculation for AAPL_modified
AAPL_modified$SMA50 <- SMA(Cl(AAPL_modified), n = 50)
AAPL_modified$SMA200 <- SMA(Cl(AAPL_modified), n = 200)
AAPL_modified$RSI <- RSI(Cl(AAPL_modified), n = 14)

# Calculate MACD for AAPL_modified

macd_data_AAPL_modified <- MACD(Cl(AAPL_modified), nFast = 12, nSlow = 26,
nSig = 9, maType = EMA)
AAPL_modified$MACD <- macd_data_AAPL_modified[, "macd"]
AAPL_modified$SignalLine <- macd_data_AAPL_modified[, "signal"]

# Indicator Calculation for MSFT_modified

MSFT_modified$SMA50 <- SMA(Cl(MSFT_modified), n = 50)
MSFT_modified$SMA200 <- SMA(Cl(MSFT_modified), n = 200)
MSFT_modified$RSI <- RSI(Cl(MSFT_modified), n = 14)

# Calculate MACD for MSFT_modified

macd_data_MSFT_modified <- MACD(Cl(MSFT_modified), nFast = 12, nSlow = 26,
nSig = 9, maType = EMA)
MSFT_modified$MACD <- macd_data_MSFT_modified[, "macd"]
MSFT_modified$SignalLine <- macd_data_MSFT_modified[, "signal"]

# Indicator Calculation for AMZN_modified

AMZN_modified$SMA50 <- SMA(Cl(AMZN_modified), n = 50)
AMZN_modified$SMA200 <- SMA(Cl(AMZN_modified), n = 200)
AMZN_modified$RSI <- RSI(Cl(AMZN_modified), n = 14)

# Calculate MACD for AMZN_modified

macd_data_AMZN_modified <- MACD(Cl(AMZN_modified), nFast = 12, nSlow = 26,
nSig = 9, maType = EMA)
AMZN_modified$MACD <- macd_data_AMZN_modified[, "macd"]
AMZN_modified$SignalLine <- macd_data_AMZN_modified[, "signal"]

What does the final data set look like?

The final data set for each stock includes the closing price, 50-day and 200-day SMAs, RSI,
MACD, and Signal Line.
# Display the structure of the cleaned data
str(AAPL_modified)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 12]
## Columns: AAPL.Open, AAPL.High, AAPL.Low, AAPL.Close, AAPL.Volume ...
with 7 more columns
## Index: Date [1148] (TZ: "UTC")

str(MSFT_modified)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 12]
## Columns: MSFT.Open, MSFT.High, MSFT.Low, MSFT.Close, MSFT.Volume ...
with 7 more columns
## Index: Date [1148] (TZ: "UTC")

str(AMZN_modified)

## An xts object on 2020-01-02 / 2024-07-25 containing:

## Data: double [1148, 12]
## Columns: AMZN.Open, AMZN.High, AMZN.Low, AMZN.Close, AMZN.Volume ...
with 7 more columns
## Index: Date [1148] (TZ: "UTC")

# Display a sample of the cleaned data

head(AAPL_modified)

## AAPL.Open AAPL.High AAPL.Low AAPL.Close AAPL.Volume

AAPL.Adjusted
## 2020-01-02 74.0600 75.1500 73.7975 75.0875 135480400
72.96046
## 2020-01-03 74.2875 75.1450 74.1250 74.3575 146322800
72.25114
## 2020-01-06 73.4475 74.9900 73.1875 74.9500 118387200
72.82685
## 2020-01-07 74.9600 75.2250 74.3700 74.5975 108872000
72.48433
## 2020-01-08 74.2900 76.1100 74.2900 75.7975 132079200
73.65033
## 2020-01-09 76.8100 77.6075 76.5500 77.4075 170108400
75.21474
## AAPL.Return SMA50 SMA200 RSI MACD SignalLine
## 2020-01-02 0.013873940 NA NA NA NA NA
## 2020-01-03 0.000942281 NA NA NA NA NA
## 2020-01-06 0.020456718 NA NA NA NA NA
## 2020-01-07 -0.004835851 NA NA NA NA NA
## 2020-01-08 0.020292094 NA NA NA NA NA
## 2020-01-09 0.007778983 NA NA NA NA NA

head(MSFT_modified)

## MSFT.Open MSFT.High MSFT.Low MSFT.Close MSFT.Volume

MSFT.Adjusted
## 2020-01-02 158.78 160.73 158.33 160.62 22622100
154.2157
## 2020-01-03 158.32 159.95 158.06 158.62 21116200
152.2954
## 2020-01-06 157.08 159.10 156.51 159.03 20813700
152.6891
## 2020-01-07 159.32 159.67 157.32 157.58 21634100
151.2969
## 2020-01-08 158.93 160.80 157.95 160.09 27746500
153.7068
## 2020-01-09 161.84 162.22 161.03 162.09 21385000
155.6270
## MSFT.Return SMA50 SMA200 RSI MACD SignalLine
## 2020-01-02 0.011588338 NA NA NA NA NA
## 2020-01-03 0.001894819 NA NA NA NA NA
## 2020-01-06 0.012414037 NA NA NA NA NA
## 2020-01-07 -0.010921450 NA NA NA NA NA
## 2020-01-08 0.007298834 NA NA NA NA NA
## 2020-01-09 0.001544736 NA NA NA NA NA

head(AMZN_modified)

## AMZN.Open AMZN.High AMZN.Low AMZN.Close AMZN.Volume

AMZN.Adjusted
## 2020-01-02 93.7500 94.9005 93.2075 94.9005 80580000
94.9005
## 2020-01-03 93.2250 94.3100 93.2250 93.7485 75288000
93.7485
## 2020-01-06 93.0000 95.1845 93.0000 95.1440 81236000
95.1440
## 2020-01-07 95.2250 95.6945 94.6020 95.3430 80898000
95.3430
## 2020-01-08 94.9020 95.5500 94.3220 94.5985 70160000
94.5985
## 2020-01-09 95.4945 95.8910 94.7900 95.0525 63346000
95.0525
## AMZN.Return SMA50 SMA200 RSI MACD SignalLine
## 2020-01-02 0.012271973 NA NA NA NA NA
## 2020-01-03 0.005615431 NA NA NA NA NA
## 2020-01-06 0.023053733 NA NA NA NA NA
## 2020-01-07 0.001239211 NA NA NA NA NA
## 2020-01-08 -0.003198008 NA NA NA NA NA
## 2020-01-09 -0.004628553 NA NA NA NA NA

Questions for Future Steps

What information is not self-evident?
• The specific thresholds for triggering buy signals using the calculated indicators.
• The performance impact of different combinations of indicators.

What are different ways you could look at this data?

• Indicator Crossovers: Identify points where short-term moving averages cross
long-term moving averages.
• RSI Thresholds: Determine buy signals when RSI falls below a certain level (e.g.,
30).
• MACD Crossovers: Identify buy signals when the MACD line crosses above the
Signal Line.
How do you plan to slice and dice the data?
• Filter by Date: Analyze specific time periods to identify trends.
• Indicator-Based Slicing: Focus on periods where indicators suggest potential buy
signals.

How could you summarize your data to answer key questions?

• Descriptive Statistics: Summarize the mean, median, and standard deviation of
indicators.
• Signal Count: Count the number of buy signals generated by each indicator.

What types of plots and tables will help you to illustrate the findings to your questions?
• Line Plots: Show the time series of stock prices and indicators.
• Crossover Points: Highlight buy signals on the plots.
• Summary Tables: Provide a concise view of signal counts and performance metrics.
library(ggplot2)
library(dplyr)

# Create line plot for AAPL stock prices and indicators

ggplot(data = AAPL_modified, aes(x = index(AAPL_modified), y = AAPL.Close,
color = "Stock Price")) +
geom_line() +
geom_line(aes(y = AAPL.Return, color = "Daily Return"), linetype =
"dashed") +
labs(title = "AAPL Stock Prices and Daily Returns", x = "Date", y =
"Price/Return") +
scale_color_manual(values = c("Stock Price" = "blue", "Daily Return" =
"red"))
library(xts)
library(ggplot2)

# Assuming AAPL_modified is your xts object with Buy_Signal column

# Filter the data to get only the rows where Buy_Signal is 1

buy_signals <- AAPL_modified[AAPL_modified$Buy_Signal == 1]

# Plot the data with buy signals highlighted

ggplot(data = AAPL_modified, aes(x = index(AAPL_modified), y =
AAPL_modified$AAPL.Close, color = "Stock Price")) +
geom_line() +
geom_point(data = buy_signals, aes(x = index(buy_signals), y =
buy_signals$AAPL.Close), color = "green", size = 3) +
labs(title = "AAPL Stock Prices with Buy Signals", x = "Date", y = "Price")
+
scale_color_manual(values = c("Stock Price" = "blue"))

## Don't know how to automatically pick scale for object of type <xts/zoo>.
## Defaulting to continuous.
library(xts)

# Assuming AAPL_modified is your xts object with Buy_Signal and AAPL.Return

columns

# Calculate Buy Signal Count

Buy_Signal_Count <- sum(AAPL_modified$Buy_Signal)

# Calculate Average Return and Total Return

Avg_Return <- mean(coredata(AAPL_modified$AAPL.Return))
Total_Return <- sum(coredata(AAPL_modified$AAPL.Return))

# Create a data frame for the summary statistics

summary_data <- data.frame(
Buy_Signal_Count = Buy_Signal_Count,
Avg_Return = Avg_Return,
Total_Return = Total_Return
)

# Print the summary data

print(summary_data)

## Buy_Signal_Count Avg_Return Total_Return

## 1 0 0.00105415 1.210165
Do you plan on incorporating any machine learning techniques to answer your research
questions?
When it comes to analyzing buy signals in the stock market, machine learning techniques
can be very valuable. Here are some ways machine learning can be used to analyze buy
signals:
1. Classification Models: Classification algorithms can be used to predict whether a
buy signal will result in a profitable trade or not. By training a model on historical
data where buy signals were generated and labeling the outcomes as profitable or
not, you can build a classifier to predict the likelihood of success for future buy
signals.
2. Time Series Analysis: Time series analysis techniques can be used to model the
patterns and trends in stock price movements over time. By analyzing historical
stock price data and buy signals, you can identify patterns that precede successful
buy signals and use them to improve decision-making.

3. Feature Engineering: Feature engineering is crucial for developing effective

machine learning models for buy signal analysis. You can create features based on
technical indicators, moving averages, volume, volatility, and other market-related
metrics to provide meaningful input to your models.

4. Ensemble Methods: Ensemble methods like Random Forest, Gradient Boosting, or

XGBoost can be used to combine the predictions of multiple models to improve
overall accuracy and robustness.

5. Reinforcement Learning: Reinforcement learning techniques can be used to

optimize trading strategies based on feedback from the market. By training a
reinforcement learning agent to make buy/sell decisions and maximize a reward
function, you can develop adaptive trading strategies.
Before applying machine learning techniques to analyze buy signals in the stock market, it’s
important to ensure that we have a well-defined research question, a suitable dataset that
includes historical buy signals and corresponding outcomes, and a thorough understanding
of the underlying principles of stock market analysis. Additionally, rigorous testing and
validation of your models on unseen data are essential to ensure their effectiveness in real-
world trading scenarios.