Updated on 09.15.14

# Dollar Cost Averaging: How Does It Work?

For quite a while, I’ve been interested in how dollar cost averaging works for a regular, steady investment, so I spent some time and used the S&P 500 as a model to see how it worked in the real world.

Wait! What’s dollar cost averaging?
Dollar cost averaging is an investment philosophy in which you buy a particular investment regularly over a period of time with an equal amount of cash each time. When the investment’s value is high, you don’t get as many shares for your dollar; when the value is low, you get more shares for the dollar. In essence, this is what you do when you automatically contribute to your 401(k) out of your paycheck.

Here’s a very simple example:

Let’s look at a year in which the value of a stock starts at 100, goes up 10 a month until June (the peak), then stays steady for the rest of the year. You paid \$134.94 per share with dollar cost averaging. If you instead bought in at the start of the year with your complete investment, then you paid only \$100 per share.

On the other hand, let’s look at the reverse market: the stock starts at 100, goes down 10 a month until June, then stays steady the rest of the year. If you invested it all right off the bat, you spent \$100 a share for stocks now worth \$50, but if you used dollar cost averaging on a monthly basis, you only paid an average of \$58.66 per share.

Dollar cost averaging in the real world
To get some real numbers to work with, I downloaded historical data for the S&P 500 since 1950. I used the S&P 500 because that is the index matched by my index fund investment of choice, the Vanguard 500.

I decided to focus on monthly investments in three periods: the last year, January 2006 – January 2007; the entire 2000 decade so far, January 2000 – January 2007; and the dot-com bust, January 2000 – January 2003.

Dollar cost averaging in 2006
In 2006, the S&P 500 went steadily upwards. Using dollar cost averaging and investing \$1000 at the start of each month in 2006, you would wind up on January 2, 2007 with an investment worth \$13,110.37, a gain of 9.26%.

However, if you had invested the entire \$12,000 in on January 3, 2006, your investment would be worth \$13,482.66 on January 2, 2007, a gain of 12.36%.

Thus, in a market that’s steadily rising, dollar cost averaging causes you to lose some of your gains.

Dollar cost averaging in 2000-2002
From January 2000 to January 2003, the S&P 500 took a major hit, dropping from a month-ending peak of 1454.60 to a month-ending valley of 815.28. Using dollar cost averaging and investing \$1,000 at the start of each month in 2000-2002, you would wind up on January 2, 2003 with an investment worth \$26,463.88, a loss of 26.48%. Ouch.

But it’s far, far worse if you invested that \$36,000 up front. If you put \$36,000 into the S&P 500 on January 3, 2000, it has a value of \$22,091.13 on January 2, 2003, a loss of 38.63%.

Thus, in a market that’s steadily falling, dollar cost averaging protects you against some losses.

Dollar cost averaging in the 2000s so far
What about over a longer span that incorporates both the down market of 2000-2003 and the subsequent rise since then? Using dollar cost averaging and again investing \$1,000 each month, on January 2, 2007 your investment of \$84,000 would have a value of \$105,086.15, a gain of 25.1%.

On the other hand, if you invested the entire \$84,000 on January 3, 2000, it would have a value of \$86,637.23 on January 2, 2007, a gain of 3.14%. This is an astounding win for dollar cost averaging.

In a market that falls and then rises, dollar cost averaging is incredibly strong.

What can we conclude?
Dollar cost averaging (what you do by default in your 401(k) investments) is a great strategy to use if you’re uncertain about the future of the market. However, if you are unquestionably confident of a bull market, the best strategy is to buy in completely right now rather than averaging it out (which is effectively what you do when you move into a new investment in your portfolio).

In short, for the average investor who is not a prognosticator, your best strategy to minimize your risk is to set up a regular, automatic investment plan and then forget about it. That plan is effectively going to handle the dollar cost averaging for you; just trust that the strategy will protect you against terrible losses in the long run.

1. Rick Dahl says:

Yes the numbers you chose showed a much better return using the DCA method as opposed to the up front method. A lot of that increase has to do with the speed at which the market fell and the slower recovery effort. The faster the market hits the bottom, the sooner you can by appreciating assets and the sooner you will hit the break even point, the point at which your portfolio is equal to the amount you have invested. You also chose a period that started with a huge loss. The ’00 – ’07 period works really well, but I am sure you can find an 8 year period that lags the market.

That said, dollar cost averaging is the easiest way to invest. With this method, you don’t have to be “right” in your pick of a stock or time the market. This method really only works if you have a long horizon.

2. Ted Valentine says:

Investing regular amounts of your income over time is not really dollar cost averaging. Its just regularly investing your income, and people agree that’s a good thing. Mutual fund companies have used and confused the terminology to increase revenues in my opinion.

True dollar cost averaging is when you have a lump sum of cash to invest and you want to invest it over certain period of time in regular amounts and intervals, presumably to reduce the risk of market timing.

True DCA (and NOT regular investing) has been studied in depth and has been shown to be a losing strategy over the long term. The studies showed that you are almost always better off investing money you have right away and not waiting. (Use Google to find studies.)

There is another method called Dollar Value Averaging. This strategy has been shown to be much more effective than DCA. I’m not going to explain it, but if someone’s interested, again, use the Google.

3. Thanks for the simple explanation. It’s good to see examples from the “real world” since I’ve mostly just seen people say “it’s bad” and “it’s good”.

4. MFJ says:

Would be interesting to see the approach used from 1950-2007 to have a longer timeframe – rather than letting short-term market fluctuations skew the data one way or the other. I have no idea but my guess is that DCA would actually do worse than say investing a lump sum at the beginning of each year. Seeing as how the market goes up a lot more than it goes down chances are DCA would result in a higher cost basis. I have no real reason to believe this other than it just makes sense, would be interesting to see that actual results.

I think DCA is a great way to invest, but think you might be slightly hurting your returns, especially if you have the money now and decide to wait and spread it out over the next x months so that you can DCA. Eventually though if you are a regular investment we all get to a point where you should DCA (IE get your money into the market ASAP).

If you wouldn’t mind doing lump sum once a year vs spreading it out over 12 months from 1950-2007 it would be very interesting to see what the results are.

5. Good findings Trent. For the average investor this is likely the best way to invest. If you don’t have the lump sum to begin with, automatic monthly contributions forces you to save money and utilize DCA. An index fund and monthly contributions seem to be the best bet.

6. Vicky says:

I would love to see some numbers that take brokerage fees into consideration, especially when investing smaller amounts per month.

7. Scrappykid says:

How do the above figures pan out if you have to pay a transaction fee each time you buy in?

Setting zecco aside for a minute, the lowest per-transaction fee i’ve found is \$4 per trade. At 12 transactions/year, over seven years… that’s \$336, which comes out over the whole amount for compound gains.

8. Trent says:

Scrappykid: I invest directly with Vanguard – no fees.

9. Anthony says:

Something to point out, as Trent says, but doesn’t mention is about Mutual Funds managed by a Brokerage in general. Whoever you hold your accounts with, be it Fidelity, Vanguard or any others, when you invest in their respective mutual funds it is usually a NTF (No Transaction Fee) Fund and it should be explicitly stated. Note: there are still maintenance fees associated with them, but no commission on the purchase/sale of said Funds. So if you choose your mutual funds making sure they are NTF funds, you should set up automatic investments and worry not about losses associated with commissions. However, you will have to pay taxes when you choose to sell these investments, unless it is in a tax-free or tax-advantaged account.

This doesn’t hold true with stocks, because your brokerage is acting on your behalf to purchase a said shares of a company. However, with a few companies (I can’t remember them off the top of my head) you can purchase stocks directly from them and avoid a commission.

10. lorax says:

The trick is to know the market direction in the future. Let me know if you find something precise.

DCA has been backtested to show that it is worse than lump sum investing 2/3 of the time. Think about it. Markets generally move up! Google for a reference. So again, unless you can call the market moves DCA loses you money most of the time.

11. Shaine says:

It occurred to me at some point that DCA is a good argument to keep peoples’ money locked in with a brokerage. Although you do average the cost of your savings plan, it also averages your gains.

It would seem more prudent to invest to win, not invest not to lose.

12. The inherant problem I have with DCA (and the reason why it is, on average, a bad play is due to the opportunity cost of not investing all at once.

Say the market returns an average of say 8 percent over a 10 year perion. If you have investment funds available, you could invest monthly over a year. But, on average, you will lose 1/2 year return doing it this way. You could put it in a high yield savigns acocunt, but you will be sacrificing average yield for security.

Being 27, I ignore DCA. I get a bonus every Feb. I immediately make my previous year IRA contributions. Sometimes, I can and will get burned on this but hitting a peak in the market. But, if I do this every year for 30 years, It will average out to getting a better return than DCAing.

If you are young and are going to do this a lot, it is better to take the risk. Near retirement, not so much.

13. drew says:

“Being 27, I ignore DCA. I get a bonus every Feb. I immediately make my previous year IRA contributions.”

You are missing out by not paying current year at that time as well.

To avoid the “January Effect”, buy prior year right before Christmas.