DCA or one fell swoop?

[scrinch]

Gummy, nice spreadsheet. Do you have a simple generic VB routine that can be applied to other spreadsheets to collect the MC results? I've never bothered to learn enough about VB to write a MC routine.

 

[gummy]

Ed:
There's some stuff on charting, here:
http://www.gummy-stuff.org/Excel_charts.htm

The spreadsheet is here (somewhere):
http://www.gummy-stuff.org/Excel/
Probably the ones that begin with "Excel".

scrinch:
The spreadsheet has monthly S&P data from 1928 to 2001 in columns A & B.
If'n y'all got your own data, you can paste 'em there.

Come to think of it, maybe I can fiddle to allow you to select a portfolio, something like this:
http://www.gummy-stuff.org/asset-correlation.htm
where you pick any four assets (provided Yahoo has the data), download and select allocations then do the DCA vs All-at-once thing with that portfolio.

I'll play with that when I get a chance.

 

[Ed_The_Gypsy]

Ed:
There's some stuff on charting, here:
http://www.gummy-stuff.org/Excel_charts.htm

The spreadsheet is here (somewhere):
http://www.gummy-stuff.org/Excel/
Probably the ones that begin with "Excel".

Merci beaucoup, mon ami!

[I forget. Are there commas in French?]

scrinch:
The spreadsheet has monthly S&P data from 1928 to 2001 in columns A & B.
If'n y'all got your own data, you can paste 'em there.

Scrinch, how far back does the Scrinch 500 go?

 

[Ed_The_Gypsy]

gummy,

The tutorial link was live again! Thanks very much!

Ed

 

[scrinch]

Scrinch, how far back does the Scrinch 500 go? 

I was asking about a Visual Basic routine for Monte Carlo simulation in completely different Excel models, not for modeling DCA using other economic data.  As far as the Scrinch500, it goes back to 334 B.C. or so.  Why do you ask?   

 

[gummy]

Finally got around to doing that DCA or All-at-once stuff, here:

http://www.gummy-stuff.org/DCA-all-at-once.htm

You select an allocation of four assets and download a bunch of monthly returns and see which ritual would give the better result:
$12K invested All-at-once
or
$1K per month for 12 months.

It's great fun to play with

 

[Rich_by_the_Bay]

Finally got around to doing that DCA or All-at-once stuff, here:

http://www.gummy-stuff.org/DCA-all-at-once.htm

You select an allocation of four assets and download a bunch of monthly returns and see which ritual would give the better result:
$12K invested All-at-once
or
$1K per month for 12 months.

It's great fun to play with

Great way of looking at this question.

I'll have to play with different asset allocations, but it is surprising to me how all-at-once dominates the few scenarios I ran. I guess it's not surprising in retrospect: market's up more often than it's down.

Thanks for doing this.

 

[Meadbh]

Time in the markets is the key. Hence it's not surprising that the "all at once" option is more likely to be successful. DCA is a useful strategy to use when investment is dependent on cash flow that dribbles in over the year.

Meadbh

 

[gummy]

I guess it's not surprising in retrospect: market's up more often than it's down.

That's what I would have said, but (of course) it depends upon whether the increases are bigger than the decreases.

Anyway, I decided to test this commonly held belief, here:
http://www.gummy-stuff.org/stocks-up-down.htm

Interesting, eh?

 

[Nords]

Interesting, eh?

Ooh, yeah. Answers an important question, too.

Nice to see you posting here again Gummy!

 

[sgeeeee]

DCA vs lump-sum decisions or more of a risk-return decision.

Lump-sum gives you the highest probability of maximum return, but also a higher probability of maximum loss. DCA smoothes that out.

 

[mathjak107]

your right about the rewards but im not so sure about the loss being greater...seems to me with 67% per cent more up than down years the dca may cut your gains more than they would cut your losses

 

[sgeeeee]

your right about the rewards but im not so sure about the loss being greater...seems to me with 67% per cent more up than down years the dca may cut your gains more than they would cut your losses

If the security does go down, there will be a day when it reaches the highest price. If you lump sum on that day, you lose the maximum you can loose. If you DCA, you may still lose, but you lose less since some of the security is purchased at a lower cost. You can't lose as much if you DCA as it is possible to lose if you happen to invest on the worst possible day during a period.

 

[mathjak107]

statistically it would be very hard if not close to impossible to invest on the worst days only..especially when the deck is stacked against you....there have been quite a few studies done to evaluate the effectivness of dca assuming of course you have a choice of lump summing it in or dca....looking at my own portfolio track record in the last 19 years my porfolio was up 15 out of the 19 years....the odds of getting a better deal and buying lower on dips putting a little in each time say would have actually had you buying in higher and higher.... of course like monday morning quarterbacking its easy to select a few time frames where dca would pan out better but statistically its very hard to catch those moments....like selling tech before the bubble burst ,collecting interest from bonds and jumping back into the markets before the rise again...looks great on paper but near impossible to pull off successfuly

 

[mathjak107]

interesting reading:from at the dogberry patch site:

Does Dollar Cost Averaging Work?

Say you decided to put $4,000 into a Roth IRA for yourself and your spouse. Would it be smarter to dump the $8,000 in as a lump sum or would you be better off dividing it up in some fashion and for example put in $1000 a month over 8 months? The thinking behind splitting up the investment is that if the market is going up and down over the course of the year you have a good chance of buying some stocks at the drops as well as minimizing the chance that if you dropped the $8,000, today's price would turn out to be the highest for the whole year.

There is an interesting article at the moneychimp regarding dollar cost averaging a large sum of money into the stock market vs just dumping the whole amount in. He has a calculator in the article that you can pick the month and year you would have started and see if the money would have done better as a lump sum investment or spread out over a year using historical returns.

The moneychimp article says that dollar cost averaging will loose 2 out of 3 times according to the calculator.

Of course, dollar cost averaging will win if your start date falls right before a dramatic crash (like October 1987) or at the start of an overall 12 month slump (like most of 2000). But unless you can predict these downturns ahead of time, you have no scientific reason to believe that dollar cost averaging will give you an advantage.

I really liked his closing paragraph where he wonders why it is then that so many people persist in believing that dollar cost averaging is better. His answer:

Maybe because it has a psychological appeal: if the market dips, people will be happy because DCA will be saving them money; and if the market goes up, people will be happy regardless.

 

[gummy]

I really liked his closing paragraph where he wonders why it is then that so many people persist in believing that dollar cost averaging is better.
His answer:if the market dips, people will be happy because DCA will be saving them money; and if the market goes up, people will be happy regardless.

I suspect it's because people who invest with every paycheck like to say (with pride):
"I do DCA!"

PS#1 Because some believe that daily, weekly and monthly up / down days are quite different, I've modified the spreadsheet to do these.

PS#2 There's a dark side to DCA.

 

[bpp]

I suspect it's because people who invest with every paycheck like to say (with pride):
"I do DCA!"

I lumpsum with every paycheck.

 

[gummy]

For thirty years I DCA'd monthly ... with every paycheck.

It sounds more sanitary than "lumpsum".

P.S.
Because some claim that the fraction of returns that are positive decreases to 50% as the time period decreases, I've added that to the tutorial.

 

[youbet]

1. DCA wins in a flat market as well as a declining market if the amount of investment is equal from period to period.

2. Some of the discussion above seems to be intertwining DCA with timing. Definitely not the same.

3. I think DCA, timing and lump sum investing all have their pros and cons and each has scenarios where they make sense. If you think one technique is best for all purposes all the time, you haven't thought it through.

 

[bosco]

I like to use synchrovest when I have periodic money to invest. it's a little more trouble than DCA but beats it over time. Here's an explanation

http://www.geocities.com/lostcowboy5/Spreadsheets/synchrovest.txt

It was developed by the same guy who invented AIM. I know Gummy is familiar with AIM because he has a discussion of it on his website.

 

[FinanceDude]

Well, the simpe answer of when to invest is: When you have the money to invest............

However, on the basis of this discussion, your core values come into play. Very conservative investors tend to DCA, more aggressive investors tend to lum sum.

Murphy's Law will come into play...........i.e..............there is an OUTSTANDING chance the market will go down the DAY after you invest lump sum...........

Long term.............TIME IN the market outperforms TIMING the market...........

 

[sgeeeee]

statistically it would be very hard if not close to impossible to invest on the worst days only..especially when the deck is stacked against you....there have been quite a few studies done to evaluate the effectivness of dca assuming of course you have a choice of lump summing it in or dca....looking at my own portfolio track record in the last 19 years my porfolio was up 15 out of the 19 years....the odds of getting a better deal and buying lower on dips putting a little in each time say would have actually had you buying in higher and higher.... of course like monday morning quarterbacking its easy to select a few time frames where dca would pan out better but statistically its very hard to catch those moments....like selling tech before the bubble burst ,collecting interest from bonds and jumping back into the markets before the rise again...looks great on paper but near impossible to pull off successfuly

DCA does reduce both the maximum gains and maximum losses achievable as compared to lump-sum. So lump-sum investing increases potential variations. Those variations are the definition of risk.

I am not advocating DCA over lump-sum and in fact have always chosen to invest lump-sum when the opportunity presented itself. But it is a risk-reward decision. If fear of a market drop the day after you invest drives you, then DCA. If you are more interested in the highest probability of maximum return and are willing to take the risk, then lump-sum.