Bollinger Bands

Q1 Load tidyquant and tidyverse packages.
Q2 Import Walmart for the last one year. Save the result under stock and print it.
Q3 Calculate 20-day moving averages and 20-day running standard deviation. Save the result under stock and print it.
Q4 Calculate the Bollinger Bands. Save the result under stock and print it.
Q5 Keep variables to build the Bollinger Bands. Save the result under stock_selected and print it.
Q6 Transform data to long form from wide form for graphing.Save the result under stock_long and print it.
Q7 Visualize data.
Q8 List both buying or selling points with dates and closing prices.
Q9 If you had invested $1 million on the day of the first buying point and followed the Bollinger Bands strategy throughout the considered period, would you have more or less than $1 million today?

Q1 Load tidyquant and tidyverse packages.

Hint: Use library().

library(tidyquant)
library(tidyverse)

Q2 Import Walmart for the last one year. Save the result under stock and print it.

Hint: Use tq_get() from the tidyquant package.


# Import data
from = today() - years(1)
stock <- tq_get("WMT", get = "stock.prices", from = from)
stock  
## # A tibble: 251 x 7
##    date        open  high   low close   volume adjusted
##    <date>     <dbl> <dbl> <dbl> <dbl>    <dbl>    <dbl>
##  1 2018-03-21  87.9  89.7  87.7  88.2 11836000     86.2
##  2 2018-03-22  87.5  88.3  87.0  87.1 12248200     85.2
##  3 2018-03-23  87.3  87.7  85.3  85.4 10125100     83.5
##  4 2018-03-26  86.2  87.7  86.1  87.5  9199800     85.5
##  5 2018-03-27  88.0  88.0  85.5  86.1  6545800     84.1
##  6 2018-03-28  86.3  88.2  86.2  87.8 11576300     85.8
##  7 2018-03-29  87.9  89.3  87.8  89.0  9274600     87.0
##  8 2018-04-02  88    88.2  84.8  85.6 19128900     83.6
##  9 2018-04-03  86.2  86.9  85.6  86.8  9695500     84.8
## 10 2018-04-04  85.6  87.5  85.4  87.2  6536600     85.2
## # ... with 241 more rows

Q3 Calculate 20-day moving averages and 20-day running standard deviation. Save the result under stock and print it.

Hint: Take stock, pipe it to tidyquant::tq_mutate to calculate 20-day moving averages, pipe it to tidyquant::tq_mutate to calculate 20-day running standard deviation, pipe it to rename() to rename value to SD, and assign the result to stock. To calculate the running standard deviation, use runSD in place of SMA. You can see all the available functions in tidyquant::tq_mutate using tq_mutate_fun_options().


stock <-
  stock %>%
    tq_mutate(select = close, mutate_fun = SMA, n = 20) %>%
    rename(SMA.20 = SMA) %>%
    tq_mutate(select = close, mutate_fun = runSD, n = 20) %>%
    rename(SD = value)

stock
## # A tibble: 251 x 9
##    date        open  high   low close   volume adjusted SMA.20    SD
##    <date>     <dbl> <dbl> <dbl> <dbl>    <dbl>    <dbl>  <dbl> <dbl>
##  1 2018-03-21  87.9  89.7  87.7  88.2 11836000     86.2     NA    NA
##  2 2018-03-22  87.5  88.3  87.0  87.1 12248200     85.2     NA    NA
##  3 2018-03-23  87.3  87.7  85.3  85.4 10125100     83.5     NA    NA
##  4 2018-03-26  86.2  87.7  86.1  87.5  9199800     85.5     NA    NA
##  5 2018-03-27  88.0  88.0  85.5  86.1  6545800     84.1     NA    NA
##  6 2018-03-28  86.3  88.2  86.2  87.8 11576300     85.8     NA    NA
##  7 2018-03-29  87.9  89.3  87.8  89.0  9274600     87.0     NA    NA
##  8 2018-04-02  88    88.2  84.8  85.6 19128900     83.6     NA    NA
##  9 2018-04-03  86.2  86.9  85.6  86.8  9695500     84.8     NA    NA
## 10 2018-04-04  85.6  87.5  85.4  87.2  6536600     85.2     NA    NA
## # ... with 241 more rows

Q4 Calculate the Bollinger Bands. Save the result under stock and print it.

Hint: Take stock, pipe it to mutate(sd2up = SMA + 2 * SD, sd2down = SMA - 2 * SD), and assign the result to stock.

stock <-
  stock %>%
    mutate(sd2up = SMA.20 + 2 * SD, sd2down = SMA.20 - 2 * SD)

stock
## # A tibble: 251 x 11
##    date        open  high   low close volume adjusted SMA.20    SD sd2up
##    <date>     <dbl> <dbl> <dbl> <dbl>  <dbl>    <dbl>  <dbl> <dbl> <dbl>
##  1 2018-03-21  87.9  89.7  87.7  88.2 1.18e7     86.2     NA    NA    NA
##  2 2018-03-22  87.5  88.3  87.0  87.1 1.22e7     85.2     NA    NA    NA
##  3 2018-03-23  87.3  87.7  85.3  85.4 1.01e7     83.5     NA    NA    NA
##  4 2018-03-26  86.2  87.7  86.1  87.5 9.20e6     85.5     NA    NA    NA
##  5 2018-03-27  88.0  88.0  85.5  86.1 6.55e6     84.1     NA    NA    NA
##  6 2018-03-28  86.3  88.2  86.2  87.8 1.16e7     85.8     NA    NA    NA
##  7 2018-03-29  87.9  89.3  87.8  89.0 9.27e6     87.0     NA    NA    NA
##  8 2018-04-02  88    88.2  84.8  85.6 1.91e7     83.6     NA    NA    NA
##  9 2018-04-03  86.2  86.9  85.6  86.8 9.70e6     84.8     NA    NA    NA
## 10 2018-04-04  85.6  87.5  85.4  87.2 6.54e6     85.2     NA    NA    NA
## # ... with 241 more rows, and 1 more variable: sd2down <dbl>

Q5 Keep variables to build the Bollinger Bands. Save the result under stock_selected and print it.

Hint: Take stock, pipe it to dplyr::select to keep date, close, SMA, sd2up, and sd2down, and assign the result to stock_selected.

stock_selected <-
  stock %>%
    dplyr::select(date, close, SMA.20, sd2up, sd2down)

stock_selected
## # A tibble: 251 x 5
##    date       close SMA.20 sd2up sd2down
##    <date>     <dbl>  <dbl> <dbl>   <dbl>
##  1 2018-03-21  88.2     NA    NA      NA
##  2 2018-03-22  87.1     NA    NA      NA
##  3 2018-03-23  85.4     NA    NA      NA
##  4 2018-03-26  87.5     NA    NA      NA
##  5 2018-03-27  86.1     NA    NA      NA
##  6 2018-03-28  87.8     NA    NA      NA
##  7 2018-03-29  89.0     NA    NA      NA
##  8 2018-04-02  85.6     NA    NA      NA
##  9 2018-04-03  86.8     NA    NA      NA
## 10 2018-04-04  87.2     NA    NA      NA
## # ... with 241 more rows

Q6 Transform data to long form from wide form for graphing.Save the result under stock_long and print it.

Hint: Take stock_selected, pipe it to gather(key = type, value = price, close:sd2down), and assign the result to stock_long.


stock_long <-
  stock_selected %>%
    gather(key = type, value = price, close:sd2down)

stock_long
## # A tibble: 1,004 x 3
##    date       type  price
##    <date>     <chr> <dbl>
##  1 2018-03-21 close  88.2
##  2 2018-03-22 close  87.1
##  3 2018-03-23 close  85.4
##  4 2018-03-26 close  87.5
##  5 2018-03-27 close  86.1
##  6 2018-03-28 close  87.8
##  7 2018-03-29 close  89.0
##  8 2018-04-02 close  85.6
##  9 2018-04-03 close  86.8
## 10 2018-04-04 close  87.2
## # ... with 994 more rows

Q7 Visualize data.

Hint: Take stock_selected and pipe it to ggplot(). Map date to the x-axis, price to the y-axis, and type to color in the line chart.


stock_long %>%
  ggplot(aes(x = date, y = price, color = type)) +
  geom_line()

Q8 List both buying or selling points with dates and closing prices.

If the stock price shoots above the upper band, the stock may be considered overbought.
If the stock price falls below the lower band, the stock may be considered oversold.
When the bands narrow, it may be used as an indication that volatility is about to rise.

Buying points happen at July of 2018 and again in Deember of 2018 Selling points happen around mid August of 2018 and again mid November 2018.

Q9 If you had invested $1 million on the day of the first buying point and followed the Bollinger Bands strategy throughout the considered period, would you have more or less than $1 million today?

You would have aboout $20 more per share of the stock meaning you would have a lot more than that of $1million today.