Apollo Investment Management 

The benefits of dollar cost averaging

The benefits of dollar cost averaging are twofold

(1) arithmetical:  see the numbers - this can dramatically increase your returns.

(2) psychological:  don’t you hate seeing an investment go down right after purchase?  Do you anguish, trying to get the timing right, while knowing you can’t?  Adopt “baby steps” – it lifts all the pressure!

The idea is that, by investing a fixed dollar amount per month, more shares will be purchased when they are cheap than when they are expensive, and the average unit price is acceptably low.  During 1998, many investors felt that shares in Asia were cheap, but worried about the timing, and the possibility of prices going lower in the short term.  In such a situation, “baby steps” may be a logical solution.

Moreover, I expect global financial markets to remain very volatile, and in such conditions the benefits of dollar cost averaging are considerable.  Nothing beats perfect timing, of course – but most of us will acknowledge the difficulty of market timing, and many of us consider it next to impossible.  Dollar cost averaging can achieve results which are almost as spectacular.  The best way of demonstrating this is with a personal example.
ChartObject Apollo 001 fund NAV

1998 was a roller coaster year in Asia.  Net asset value of the Apollo 001 Fund started at $10.00, then fell to $6.50, and rose as high as $13.65 before closing the year at $13.12.  In 1998, I personally was gradually liquidating other assets for investment in Asia, and accordingly purchasing units in the fund throughout the year (although not evenly).  Buying and holding at the beginning of the year would have resulted in a gain of 31%;  perfect timing on purchase (buying at the bottom) would have increased this to 101%;  my actual gain for the year was 70%.

(Annualising this figure would give a profit of 104%, but I think this would be misleading.  70% was the actual dollar profit on the average dollars invested.  For the record, the dollar profit compared to the year-end dollars invested was 46%, but some of these investments were made right in the closing days of December, so that figure clearly understates the performance.)

If you are in the fortunate position of having regular income, you can do this year after year.  If you are thinking of investing a fixed pool of funds, it is only relevant until the funds are invested. “The baby steps” principle may be relevant if you want to change your asset allocation, but if you are reasonably happy with your investment mix you will then settle down to the normal return on funds fully invested.  However, that first-year benefit can be tremendous.  Late December 1997 was the first time in years when I felt comfortable to invest aggressively in Asia:  prices might go up, they might go down, but I knew I was buying assets at prices which in the long run would prove bargains.  As a result of dollar cost averaging, I have 30% more units in the fund than if I had invested all that money at the outset.

Let me show you a more formal example. This is what would have happened if you invested a fixed dollar amount for twelve consecutive months at NAV. (I'm ignoring fees here; this is just an example to demonstrate the arithmetic of dollar-cost averaging. Please note that neither this nor any other Apollo funds are open to public investment. This website is for information only.)

Apollo 001 Fund - the 1998 roller-coaster tamed by dollar cost averaging

Valuation date  NAV, US$ Invest monthly Units bought Value at  % gain
31-Dec-97 10.0000  $ 1,000  100  $ 1,312  31.2%
31-Jan-98 7.1739  $ 1,000  139  $ 1,829  82.9%
28-Feb-98 12.3654  $ 1,000  81  $ 1,061  6.1%
31-Mar-98 13.6516  $ 1,000  73  $ 961  -3.9%
30-Apr-98 12.5589  $ 1,000  80  $ 1,045  4.5%
31-May-98 10.5358  $ 1,000  95  $ 1,245  24.5%
30-Jun-98 8.9751  $ 1,000  111  $ 1,462  46.2%
31-Jul-98 7.8998  $ 1,000  127  $ 1,661  66.1%
31-Aug-98 6.8523  $ 1,000  146  $ 1,915  91.5%
30-Sep-98 8.4540  $ 1,000  118  $ 1,552  55.2%
31-Oct-98 10.9307  $ 1,000  91  $ 1,200  20.0%
30-Nov-98 13.6684  $ 1,000  73  $ 960  -4.0%
At 31-Dec-98 13.1197  $ 12,000  1,235  $ 16,202  35.0%

The average entry cost would have been a comfortable $9.72, and the gain on the overall portfolio by year end would have been 35%. Had one been impeccably lucky with the timing, and invested the whole sum at end-August, one could have made 92% - but then again, it was possible to argue just as plausibly for a bottom eight months earlier, and make only one-third as much. Had one been unlucky enough to invest at the twin peaks, the outcome for the year would have been a small loss. The median of the returns on a lump-sum investment is 28% - that is, investments in six of the months would have been up more than 28% by year end, and investments in six of the months would have done worse. The arithmetic mean of these returns was 35%, and that is the figure obtained by the dollar-cost averager - who might also have slept better during the year, comforted during the dips by the knowledge that he still had cash to spend and was currently obtaining even more bang for the buck.

I was fortunate enough to make a good return in 1998, but the opportunities are just as interesting now as they were throughout last year - and we are by no means out of the choppy water yet, so it may well be that volatility will remain quite high, to the advantage of the new investor and of those using dollar cost averaging. 

Claire Barnes, 16 January, 1999

Companion report: Previous reports:
Investment philosophy
Reports & articles
Poetry & doggerel
Why Apollo?
Who's Claire Barnes?
What's new?