Log in

No account? Create an account

The truth about dollar cost averaging

« previous entry | next entry »
Dec. 4th, 2005 | 12:17 pm
mood: analytic

Having read another point of view on DCA in the book "All About Exchange-Traded Funds" by Archie Richards Jr. (McGraw-Hill), I would like to talk about it some more.

He makes an important point that you cannot look at DCA effectively without considering the situation that the investor is currently in.

First, an assumption: Market timing is not possible. By this I mean that it is impossible for the average investor to predict if the market will go up or down. Even well-informed investors make mistakes all the time, and it is absurd for someone whose profession is not finane to think that they have the knowledge to predict the stock-market.

Second assumption: The market goes up more often than it goes down. Historically this is true, of the market as a whole. Individual stocks do not follow this rule as their value can be wiped out completely or make huge losses, leaving no chance to recover.

Case 1: Investor has no savings. In this case, the only options the investor has is to save up and then invest, or to invest as money comes in. If the money is saved first, then it is not invested at all for that time. The money which is not invested is guaranteed to receive low returns (or zero returns if kept in a standard savings account). Whereas the invested money is more likely to receive higher returns than lower (and possible negative) returns. In this case it is clearly less risky and more valuable in terms of returns to invest the money as it comes in. Note that as market timing cannot be used, there is no advantage to saving and investing later.

Case 2: Investor has savings in a fixed interest account, such as a savings account (0.025% p.a.) or Netbank Saver (5.4% p.a.). The investor is choosing between investing the entire sum now, or investing in dollar cost averaged blocks over the course of a year. Clearly the time the money is invested will be maximized by investing it all now. But risk will also be increased.

Let's assume that the stock-market rises 12% p.a. annualized (*). When not considering risk, it is better to invest all the money now, because 12% is bigger than 5.4%. Remember that it is more likely for the market to go up and down, and DCA only helps when the market is going down.

When considering risk, DCA will give less risk but will also reduce the expected return.

So, if you plan to hold for the long term, then it is better to invest all the money now, as the fluctuations will even out over time. DCA will give you less return, as you are investing your money for less time (Remember we are still in the case where the investor has savings already, and is choosing between investing all now, or investing in chunks).

Case 3: Investor has money in a high-risk investment alread, and is considering switching to another high-risk investment. As in case 2, the options are to use DCA over 1 year (could be monthly, quarterly) or to switch in one block. But here, keeping the money in the first investment is just as risky as having it in the second investment. But still, DCA can reduce risk.

If using DCA, it is less likely that you will transfer all the money from investment A to investment B when A is low and B is high. By transferring in chunks, you are averaging out the fluctuations in prices of both investments. All the time your money is getting you "high-risk" return in both investments, so you are not losing out on investment time by transferring in chunks.


In summary:

- If you are starting to invest with no capital, then DCA is your only real choice. There is no point delaying your investment and thereby increasing your risk.
- If you have savings already, your choice is DCA with lower risk and lower return, or a single lump sum with higher risk and higher return.

If you want to do market timing (trying to guess and invest when the market is low) then none of these rules apply of course. But I am assuming that the investor does not have the knowledge or experience to do market timing.

(*) Annualized return is calculated over a period, such as 5 years (or even 1 month if you want a hugely inaccurate result). It is the annual return per year that would give the total return over that period (with any returns re-invested). For example, 50% return over 10 years will be a little less than 5% per year annualized (less because of compounding).

Link | Leave a comment |

Comments {2}

hey Brian

from: anonymous
date: Dec. 6th, 2005 07:29 am (UTC)

Hey it's Jerry. Trying to get in contact with you - this place I got a while ago from Paula. Send me an e-mail address so we can chat in a non-public web space. Congrats on the baby! Talk soon, love Jerry.

Reply | Thread

(no subject)

from: anonymous
date: Dec. 6th, 2005 11:49 am (UTC)

I've graduated from uni, but my undergraduate e-mail is still active - I have a gmail account but lost the password, so the uni one is the only one I have right now: j.smyrk@ugrad.unimelb.edu.au


Reply | Thread