Dollar cost averaging can be performed a number of ways. From a recent poll on the Drip message board, about a quarter of the people indicated that every month they send the same dollar amount to make a purchase. This is what I refer to as straight dollar cost averaging. About a third always alter the amount they send, in an attempt to get more bang for their buck.
In the February 25, 2000 Drip Portfolio article, Jeff wonders about mistakes that may have been made. After about two-and-a-half years, it is proper to consider the right and wrong things that have been done, and learn from the past.
Early in the portfolio's existence, the decision against straight dollar cost averaging was made. Soon after its selection, Intel (Nasdaq: INTC) lost favor with the market and its price dropped from $96 in early September 1997 to $69 by the following December. One nice thing about Drips is that they don't necessarily give one a chance to panic.
Although Johnson & Johnson (NYSE: JNJ) had been selected in November of 1997, Intel's drop was seen as a buying opportunity and the decision was made to purchase larger amounts. By the time that Campbell Soup (NYSE: CPB) was purchased in April 1998, the amount sent to Intel was more than two-and-a-half times greater than was sent to Johnson & Johnson.
What would have happened if straight dollar cost averaging had been applied to the Drip Port's purchases instead? For one thing, this would not have been possible because the minimum purchase requirement of Mellon Financial Corporation (NYSE: MEL) requires the entire month's funds, or $100. However, for argument's sake, we will assume that it was possible to purchase each company (from its inception) each month, sending the same amount each time. Let's see how this would have worked out.
First, we look at the average cost of the shares that the Fool's Drip Portfolio paid for Intel, Johnson & Johnson, and Mellon Bank (I am excluding Campbell Soup from the comparison because regular purchases have not been made):
Comparing the February 25, 2000 price of each company, we have the following:
|Company||Feb 25, 2000||Gain/Loss|
To make this determination, I obtained the closing price of each company on the date that it would most probably have been purchased. Since Mellon purchases every week, I selected the 7th of the month as the date to obtain the price of these purchases (you can see a table with full prices at this page):
As we can see by comparing the tables, there is little difference with Johnson & Johnson and Mellon, but a large one with Intel. This is not surprising at all.
The reason that one might have guessed there to be a difference with Intel and not the others has to do with the price swing during the life of the purchases. In this timeframe, Intel's lowest price was a split-adjusted $33.19 and highest was $101.41. Dividing the latter by the former gives us a value of 3, which is quite a range.
When we look at the price swings of Johnson & Johnson and Mellon, we see respective lows and highs of $62.125 and $103.78, and $23.67 and $37.53. Using the same division gives us the values of 1.7 and 1.6. Adjusting the purchase amount to take advantage of the current price has a greater effect on volatile stocks than relatively calm stocks.
A final note might be to wonder what would have happened if outright one-time purchases had been made, as opposed to dollar cost averaging. Again, we can look at a table for this:
Intel would not have fared as well, as the decision to load up when the stock was under pressure made a big difference. Also, remember that the outright purchaser would have had to weather a nearly 30% drop in price very soon after purchase. For the investor not committed to a long-term strategy, this certainly must have caused numerous sleepless nights.