The Dividend Investing Resource Center

Letting Dividends Do Their Work

Blind Dollar Cost Averaging Variations Part 2

George L Smyth

Last week I decided upon a means to measure the risk of selecting different timeframes for making regular investments. I determined that measuring the high and low purchase price during the period would serve to tell us what might have been. For instance, during the month of December 1999, Coke's low for the month was $58 1/16 and the high was $69. The latter price being 18.8% higher than the former could be used for risk comparison purposes when looking at other timeframes.

To measure this risk, I selected four stocks that have received a lot of attention on the boards:Coca-Cola (NYSE: KO), Enron (NYSE: ENE), Intel (Nasdaq: INTC), and Pfizer (NYSE: PFE). I decided to start with the silly notion of making a purchase every trading day between January 3, 1995 and December 31, 1999. This would serve as a minimum risk level.

The four companies averaged a price swing of 2.62% when examined on a daily basis over the past five calendar years. That's fairly small risk. Of course, nobody is going to invest that often, so I moved on to see how things looked on a weekly basis.

As expected, the volatility grew. The average price variation more than doubled to 5.6%. Of course, increasing the timeline by a factor of seven might have made one think that the change would have been greater.

However, I would think that if there is a standard, it should be monthly investments, as this is how most of us make our purchases. Volatility examined this way measured just under 10%, at 9.9%. Again, a period increase of 4.3 only doubled the risk as measured by volatility.

With a standard of 10% established, I was particularly interested to see how the numbers changed when we examined a bimonthly timeframe. This was of particular concern to me, as I purchase half of my Drip portfolio one month, the other half the following month. (Out of necessity, the Fool's Drip Port invests in its companies in an irregular fashion, too.) This time, a mere doubling of the time period, from one month to two, increased the risk over 1.5 times to 15.9%, up from 9.9%.

Some companies only offer quarterly purchases. Also, in an attempt to lessen the impact of fees, a number of people select this purchase period. However, by only tripling the period over our monthly standard, the risk more than doubled to 21%.

I dont believe that many investors make semiannual or annual purchases, but I calculated the volatility risk on these time measures as 35.5% and 60.1%, respectively.

Below is a chart detailing the calculations made for each company, with a monthly investment being the benchmark to measure against. The information is stored in an Excel 97 spreadsheet and is rather large, so prepare yourself if you have a slow modem.

Volatility as Measured During 1995-1999
  Daily Weekly Monthly Bimonthly Quarterly Semiannual Annual
Coke 2.25% 4.92% 9.28% 13.78% 21.14% 30.94% 46.86%
Enron 2.35% 4.85% 7.28% 11.87% 14.80% 25.25% 37.23%
Intel 3.22% 7.06% 13.70% 22.32% 29.25% 50.04% 90.85%
Pfizer 2.67% 5.71% 9.50% 15.75% 20.58% 35.66% 65.33%
Average 2.62% 5.63% 9.94% 15.93% 21.44% 35.47% 60.07%
Diff -73.64% -43.36% 0% 60.26% 115.69% 256.84% 504.32%

As we can see from the numbers, a bimonthly schedule increases volatility risk by 60%, and quarterly purchases almost double the monthly risk. This means that people who are setting funds aside for quarterly purchases to reduce fees may be making a larger tradeoff than they had realized.

Of course, it is all about knowledge and comfort. If you know your investment situation and you can place yourself at the proper comfort level, you will certainly be able to sleep better at night.

This website is maintained by George L Smyth