Use access key #2 to skip to page content.

talotu (< 20)

Dollar Cost Avg - A Sirens' Song



June 08, 2009 – Comments (7)

 Joe has decided to invest $100,000 from a maturing CD by buying a Vanguard total market index fund. Should he

 a) Invest all in one chunk on the first day the money is available

 b) Put the money in 20% per day for 5 days, thus averaging his purchase price over 5 different days.


My short answer is 'a', in fact it seems intuitively obvious. What impresses me is how many people who would be considered intelligent (MBA from a top-ten business school, PhD from a renowned science program) not only answer 'b', but have difficulty seeing how 'a' could possibly be correct.

 In fact when a Vanguard advisor was asked this question he chose 'a', but gave a specious argument why, though it's possible that they are trained to give incorrect reasoning people would accept, rather than use a futile attempt involving logic and math which will end up confusing people. (Which is all I've been able to manage) 

7 Comments – Post Your Own

#1) On June 08, 2009 at 12:28 PM, XMFCrocoStimpy (97.53) wrote:

I'd be interested in seeing the long answer "involving logic and math"


Report this comment
#2) On June 08, 2009 at 12:46 PM, scottgeiger (< 20) wrote:

option b works only IF the stock price goes down over the investment period.  In this case you have averaged to a lower total cost.  If the stock price goes up then you have averaged higher.  This is "blind DCA" - basically just investing at set periods without regard to the price of the stock, or any other factor for that matter.

A better practice might be:

1. Invest a portion of your funds at $x

2. Wait a certain period (depends on the stock and the market)

3. If the price of the stock falls below $x AND you still think the company is solid, invest another portion (rinse and repeat).

4. If the price of the stock goes above $x, then look for other investment opportunities.


Report this comment
#3) On June 08, 2009 at 1:07 PM, dudemonkey (53.56) wrote:

"a" is effectively market timing.  You are betting that the market will never go lower than its current valuation.

Report this comment
#4) On June 08, 2009 at 1:15 PM, rofgile (99.29) wrote:

I normally would say "a", since I am investing small amounts of money with a high transaction cost.  If the transaction cost was $0, I would be fine with doing option "b" - but with a larger time scale.

If we are in a two year bear market (accepted as an assumption) and over time the market should keep going lower, then I would choose option "b" with time intervals of 2 months - 4 months per each of the five purchases.

Report this comment
#5) On June 08, 2009 at 2:27 PM, Melaschasm (71.33) wrote:

IMO, buying stocks over the course of 5 days is not what dollar cost averaging is meant to do.

Dollar cost averaging works great for 401k/IRA accounts.  This is money that naturally enters the market over time, and the time frame is very long, unless you are about to retire.

Many 401k plans require you to invest the money as you earn it, so I will instead focus on an IRA account.

If you are going to invest in an IRA each year, then dollar cost averaging is a worth consideration.  If you are using this account to invest in the Vanguard 500 Index Fund, investing $6000 per year every year is likely to be more profitable than keeping the money in cash and waiting for the right moment to invest (market timing), according to dollar cost averaging theory.  Likewise, if you have an account with very low fees, then making a $500 investment in the IRA each month is likely to result in better returns than picking one day a year to make your investment.

When it comes to investment philosophies, individual circumstances are likely to determine what is the best way to invest. 

Report this comment
#6) On June 08, 2009 at 5:07 PM, talotu (< 20) wrote:

If you assume that whatever you are investing in has a long term positive expectation (if not then why are you investing) then even with $0 transation costs contributing $500 per month is likely to result in worse returns, not better.  It will lessen your risk, but given that you are going to be investing eventually anyway, you have already decided that the risk is acceptable.

The first option is not timing the market, it is getting your money in as soon as possible in what you expect to be a worthwhile investment for the risk.

Timing the market involves *not* investing money that is available, and holding it for a later time.


Dollar cost averaging is just an attribute of having money available at different times.  It doesn't do anything.

For example, you work at an employer that lets you put up to 50% of your income in your 401K, and you make more than 33,000 (2x the IRA max).

Is it better to contribute a % such that you make equal contributions throughout the year, or is it better to choose 50% (assuming you can survive on half your net salary) until you max.


Report this comment
#7) On June 08, 2009 at 5:34 PM, Deepfryer (27.25) wrote:

""a" is effectively market timing.  You are betting that the market will never go lower than its current valuation."


Not really. If you consider "a" to be market timing, then "b" is market timing as well. With "b" you're betting that the market will go down during the period between day #2 and day #5. That's the only scenario where option "b" is better than "a".

You don't really know whether the market will go up or down in the period from day #2 through day #5. But, the market goes up more often than it goes down, so probability-wise, your "expected value" is slightly higher if you choose option "a". You also save on commissions by buying the shares all at once.

Report this comment

Blog Archive

June (1)

Featured Broker Partners