[Sampson]

Is anyone is aware of some primary research that describes dollar cost averaging (DCA) and if there is an optimal size of a contribution (either dollar or percentage of total funds to invest).

Obviously this is highly dependent on how the market is acting, but I'd like to find out if 'guidelines' exist.

In the past, my individual transactions have been 0.5-2% of my entire portfolio, and the minimum was based on keeping the expense ratio below 0.5% (I've now softened this rule to 1%).

So lets say I've got $10-$20k sitting, and all needs to go to one holding to fit my asset allocation, would it be recommended to buy in $1k chunks (1% expense ratio) or larger?

Pointers to primary research appreciated.

 

[Mockingbird]

Dollar Cost Averaging concept has been debunked many times since the first article appeared in the Journal of Financial & Quantitative Analysis in 1979. Most study will yield that the dollar-cost averaging is unlikely to produce superior results to lump-sum investing. However, I believe it is a great tool for implementing a "forced" savings habit.

Article on DCA
http://moneycentral.msn.com/content/P104966.asp


Original article by George Constantinides in 1979
A Note on the Suboptimality of Dollar Cost Average As an Investment Policy

You can do your own math.
Does Dollar Cost Averaging Work?

 

[Sampson]

Thanks for the links.

Been reading some primary literature -

Knight & Mandell (1993). Nobody Gains From Dollar-Cost Averaging. Analytical, Numerical and Empirical Results. Financial Services Review. 2(1):51-61.

As well as the original Constantinides article.

It appears the main reason DCA does not work in these models is because of the opportunity cost. These studies assume that the money to be invested is averaged in over the entire holding period (when compared to initial lump sum investment followed by buy and hold).

However, over a shorter time frame, where the opportunity cost is low, then presumably DCA (say over a month or two) would not consistently be outperformed by a simple lump investment strategy.

Still reading, but it appears there may be an optimal investment amount - which produces better returns not because of the DCA, but because of reduced fees.

 

[mfd]

you should look at value averaging which has shown to produce better results then dollar cost averaging and random investing.

http://www.studyfinance.com/jfsd/pdffiles/v13n1/marshall.pdf


Basically you determine that the value of your investment needs grow 8% every year to meet some final goal. Every couple of months you check the performance...if you performed less than 8% then you buy more to bring the value up to what it should be. If you perform better then 8% then you sell some of the investment.

Its basically rebalancing based on performance rather then asset allocation

 

[Dividend Growth Investor]

Hmm if you have money just sitting around, chances are you are going to underperform your benchmark for the year. But If you are looking to invest it all over one year I would put 20K/12 months or 1667 every month automatically and thus smooth out your risk..

 

[CanadianCapitalist]

Is anyone is aware of some primary research that describes dollar cost averaging (DCA) and if there is an optimal size of a contribution (either dollar or percentage of total funds to invest).

Obviously this is highly dependent on how the market is acting, but I'd like to find out if 'guidelines' exist.

Doesn't this depend on the size of the portfolio relative to the contributions? When we were starting out, a big lump sum contribution (say from a surprise bonus) might work out to 10 or even 20 percent of the portfolio. These days, we'd be lucky if it is 2 or 3 percent.

Personally, I keep investing whenever I have money. It is very hard to time the markets right. Take the current market situation -- the markets are up 30 percent or so from the bottom. Some say a correction is inevitable. Others say bull markets climb a wall of worry. As it is always the case, there are conflicting views and it may be best to keep investing.