Task

Compare the income (or loss) that comes from buying and selling a stock certificate to the income that comes from dividends over the same period. Answer these questions:

• Which source of income is bigger?
  
• Is there any correlation between the income from dividends and the income (or loss) from buying and selling?

Getting Price Data

Sites such as finance.yahoo.com collect and distribute information about individual companies. You can use the readStockPrices() function (in the DataComputing package) to read such data directly from Yahoo into R.

For example, here are some automotive stocks and their daily prices from 2010 to 2015.

companies <- c("F", "TM", "HMC")
Prices <- 
  read_stock_prices(companies, what="daily",
                  start_year=2000, end_year=2015) 

• Choose a few companies of interest to you. You can find stock company symbols at finance.yahoo.com. (Suggestion: pick a sector of the economy, e.g. energy, high-tech, consumer products, etc. and use companies from that sector.)

• Plot out the “closing price” (Close) versus date to get a graphic like Figure 1.

Buy/Sell Profit

⊕  action       date
1    buy 2006-01-03
2   sell 2014-12-30
The Actions table constructed by the previous command.

Pick a buy date and a sell date. You can use a command like this to create your data table:

Combine Prices and Actions to produce a table like this which shows the profits from buying at the start of 2006 and selling at the end of 2014.11 The dollar amount is profit per share.

Hints: (1) What kind of join should you use so that you get only those cases that match one of the dates in the Actions table? (2) Wide-vs-long techniques will be useful. (See Data Computing Chapter 11.)

From the data table with buy and sell prices, calculate the dollar amount of profit (or loss) and the percentage change, as shown here.

⊕  company profit percent
1       F   7.67   98.00
2     HMC   0.24    0.82
3      TM  19.10   18.00
Profit and percentage change in price.

Indexing Prices

Since stock prices vary markedly from one company to another, a common practice is to “index” the price to a particular date as in Figure 3. (Question: In the graph, roughly which date was used for the reference?)

• Pick a single date of your choice and extract the stock price information for each company on that date. In the result, there should be one case for each company. Select just the date, company, and close variables, renaming close as standard. Call the resulting data frame Reference.

ref_date <- ymd("2005-01-03")
Reference <-
  Prices %>% 
  filter(date==ref_date) %>%
  select(company, standard=close)

Figure 3. Indexed stock prices for Ford, Honda, and Toyota

• You now need to combine the Reference with each day’s price data for that company. You’ll find the standardized price on each day by creating a new variable which is the ratio of the day-to-day price (use Close) to the standard for that company. Before you can do this, you’ll need to combine the Prices and Reference data tables. You’ll use a join verb to do this. In order to check your results, sketch out what you think the result should be before you do the join.

Dividends

You can read in dividend data like this:

Dividends <- 
  read_stock_prices(companies, what="dividends") 

Once you have the dividend data, extract out the dividends for all dates between your buy and sell dates. (Hint: Join Dividends to Actions using company to match.)

• The dividend amount is actually a rate: the dividend paid (in dollars) divided by the stock price. Find the dollar amount of each dividend payment for one share of stock rather than one dollar of stock. This involves multiplying the dividend rate by the stock price on that date.

⊕  company total_dividend
1       F          1.750
2     HMC          1.885
3      TM          6.259
Table 1. Dividends (in dollars per share) paid out by each company over the interval 2005 through 2014.

• Find the total amount of dividends for each company during the period of interest. Compare this amount to the profit (or loss) from buying and selling the stock certificates. For the car companies, the result for the period 2005-01-01 though 2014-12-31 is shown in Table 1 .