Introduction to Stock Market Analysis

suppressPackageStartupMessages({
  library(tidyverse)
  library(scales)
  library(glue)
  library(tidyquant)
  library(DT)
  library(lubridate)
  library(treemapify)
})

1 S&P 500 Index

In this project we will analyze stocks in S&P 500 index (symbol: ^GSPC) which measures the stock performance of 500 large companies listed on stock exchanges in the United States (Wikipedia).

We can get daily values of the index using tq_get() function of the tidyquant package

sp <- tq_get("^GSPC", from = "1950-01-01") 

sp %>% 
  select(-adjusted, -symbol) %>% 
  DT::datatable(
    caption = "S&P 500 Raw Data",
    options = list(
      dom = 'tip',
      pageLength = 10,
      autoWidth = TRUE
    ),
    rownames = FALSE) %>% 
  formatRound(columns = 2:5,digits = 2)

Let’s use ggplot to plot the close column vs date (fill in the missing code)

sp %>%  
  ggplot(aes(x = date, y = close))+
  geom_line(color= "steelblue")+
  scale_x_date(breaks = pretty_breaks(8))+
  scale_y_continuous(breaks = pretty_breaks(8), labels = comma_format(prefix = "$"))+
  labs(x = "", y="", title = "S&P 500 Index")+
  theme_bw()

A question: how much did the index S&P 500 gain since Jan 1, 1950?

sp %>% 
  summarise(gain = last(close)/first(close))

Annual returns can be calculated as the following

sp_annual_returns <- 
  sp %>%
  tq_transmute(select     = close, 
               mutate_fun = annualReturn,
               col_rename = "gain") %>% 
  mutate(year= year(date))

sp_annual_returns %>% 
  mutate(is_pos = gain > 0) %>% 
  ggplot(aes(x = year, y = gain))+
  geom_col(aes(fill = is_pos), width = 1,show.legend = FALSE)+
  scale_x_continuous(breaks = pretty_breaks(8))+
  scale_y_continuous(breaks = pretty_breaks(8), labels = percent_format(accuracy = 1))+
  labs(x = "", y="", title = "Annual Returns S&P 500 Index")+
  theme_bw()+
  geom_smooth(method = "lm", se = FALSE, lty = "dashed", size  = 0.5)

Average annual return

sp_annual_returns %>% 
  filter(year < 2020) %>% 
  summarise(average_gain = percent(mean(gain),accuracy = 0.01))

NA

2 Stocks comprising the S&P500 Index

df <- tq_index("SP500") %>% 
  select(symbol, company, weight, sector)

df %>% 
  DT::datatable(
    caption = "S&P 500 Stocks",
    filter = "top",
    options = list(
      dom = 'tip',
      pageLength = 10,
      autoWidth = TRUE
    ),
    rownames = FALSE) %>% 
  formatPercentage(columns = 3,digits = 2)

Let’s now find the composition of the index based on various sectors

df %>% 
  group_by(sector) %>% 
  summarise(N= n(),
            weight = sum(weight)) %>% 
  arrange(-weight) %>% 
  DT::datatable(
    caption = "S&P 500 Breakdown by Sector",
    options = list(
      dom = 't',
      pageLength = 11,
      autoWidth = TRUE
    ),
    rownames = FALSE) %>% 
  formatPercentage(columns = 3,digits = 2)

df %>% 
  ggplot(aes(area = weight, fill = fct_reorder(sector, weight, .fun = sum,.desc = TRUE),label = symbol, subgroup = sector))+
  geom_treemap(alpha = 0.5, start = "topleft")+
  geom_treemap_text(start = "topleft", size = 12)+
  scale_fill_viridis_d(name = "Sector")

Which sector performed best since the beginning of the month?

qt<- 
  df %>% 
  pull(symbol) %>% 
  tq_get(from="2020-06-01")

dfj <- 
  qt %>% 
  filter(date > today()-ddays(7)) %>% 
  group_by(symbol) %>% 
  summarise(gain = last(close)/first(close)-1) %>% 
  left_join(df, by = "symbol")

dfj %>% 
  group_by(sector) %>% 
  summarise(average_gain = sum(gain*weight,na.rm = TRUE)/sum(weight,na.rm = TRUE)) %>% 
  arrange(-average_gain) %>% 
  DT::datatable(
    caption = "Last Week Perfomance by Sector",
    width = "500px",
    options = list(
      dom = 't',
      pageLength = 11,
      autoWidth = TRUE
    ),
    rownames = FALSE) %>% 
  formatPercentage(columns = 2,digits = 2)

Which stocks had the highest gain in each sector?

dfj %>% 
  group_by(sector) %>% 
  top_n(1,gain) %>% 
  select(symbol, company, gain, sector) %>% 
  arrange(-gain) %>% 
  DT::datatable(
    caption = "Best Stocks by Sector in June",
    width = "500px",
    options = list(
      dom = 't',
      pageLength = 11,
      autoWidth = TRUE
    ),
    rownames = FALSE) %>% 
  formatPercentage(columns = 3,digits = 2)

NA

3 Individual Stocks Performance

3.1 Candlestick Plot

c("AAPL") %>% 
  tq_get(from = "2020-01-01") %>% 
  ggplot(aes(date, close))+
  geom_candlestick(aes(open = open, high = high, low = low, close = close)) +
  scale_y_continuous(breaks = breaks_extended(6), labels = comma_format(prefix = "$"))+
  scale_x_date(breaks = pretty_breaks(6))+
  facet_wrap(~symbol, scales = "free_y") +
  theme_tq()+
  labs(x= "",y= "")

NA