Dollar-Cost-Average over HOW MANY years?

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
MikeT
Posts: 389
Joined: Sun Jul 10, 2016 11:26 pm

Dollar-Cost-Average over HOW MANY years?

Post by MikeT » Fri Jan 13, 2017 3:01 pm

Within my IRA, my desired allocation is 80 stock / 20 bonds.
Presently, I'm at 67% stock, 20% bonds and 13% "cash".

Because the stock market is at an all-time high (nearly 20,000), I don't want to use the "cash" to buy the stock all at once..

Here's my plan: does this make sense:

Dollar cost average in over 2 years.
If the market drops to 19,000 buy stock with HALF remaining cash.
If the market drops to 18,000 bu stock with ALL remaining cash.

Thoughts?

Thanks,
Mike

Fox
Posts: 40
Joined: Tue Dec 13, 2016 2:00 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by Fox » Fri Jan 13, 2017 3:17 pm

What will you do with the cash if the Dow hits 21,000 this year? Keep waiting?

Crisium
Posts: 174
Joined: Thu Dec 22, 2016 9:52 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by Crisium » Fri Jan 13, 2017 3:23 pm

And what if stock rise fairly steadily over 24 months? Then you lost the gamble. You have bought fewer shares with the same amount of cash, and received fewer dividends.

I'm not opposed to DCA. Of all possible things, lump summing the day before a correction is the worst possible thing. But DCA is no guarentee of "better losses" (see above paragraph, what if there's a correction in month 25? DCA loses more). Consider a shorter time frame, as a common rule of thumb is never DCA more than 12 months.

edit: 2 years is 24 months, duh
Last edited by Crisium on Fri Jan 13, 2017 3:52 pm, edited 3 times in total.

MikeT
Posts: 389
Joined: Sun Jul 10, 2016 11:26 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by MikeT » Fri Jan 13, 2017 3:23 pm

I guess so.

Are you suggesting no dollar-cost averaging and just go all-in right now (1/13/17)?

_Mike

Jags4186
Posts: 2646
Joined: Wed Jun 18, 2014 7:12 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by Jags4186 » Fri Jan 13, 2017 3:26 pm

If I get a windfall I always lump sum invest. The time frame which most people talk about DCAing a lump sum of cash is always about 12 months, but no one knows what the market is going to do in 1 year. If we look at history, the market is much more likely to be higher than it is today 1 year from now rather than lower. That supports lump sum investing right now. If we look at history the market is almost always higher 10 years in the future (I don't think there is a 10 year period with losses maybe if you start at the peak before the great depression?). That also supports lump sum investing now. If I am DCAing I'm basically hoping that the market drops so I get the benefit, and if I'm hoping the market drops that means I shouldn't DCA and just hold my money in cash until he market drops. In fact it means I should cash out my current investments. Now that doesn't make any sense to me.

investordoc
Posts: 63
Joined: Wed Jan 21, 2015 1:50 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by investordoc » Fri Jan 13, 2017 3:39 pm

Agree with others. You can go all in or if it makes you more comfortable put some in this month and then each month after until at desired levels. Maybe if you see a pullback while waiting for the next month put more in then (no...i'm not suggesting market timing). You have to be comfortable with your decisions.
It is what it is until it isn't anymore

cheapskate
Posts: 748
Joined: Thu Apr 26, 2007 1:05 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by cheapskate » Fri Jan 13, 2017 4:09 pm

Why not DCA over 2 years, but put equal amounts every months (over 24 months) ?

It sounds like psychologically you might be happier DCA'ing. If that is the case go ahead and DCA over some reasonable interval. It matters much more that you stick to the plan and less whether you lump sum or DCA. DCA'ing might not be mathematically optimal but for a lot of investors, it is a perfectly fine regret minimization strategy (they are more likely to stick to their allocation once they DCA into it).

bantam222
Posts: 96
Joined: Sun Nov 23, 2014 5:57 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by bantam222 » Fri Jan 13, 2017 4:20 pm

In my opinion, DCA is simply a strategy to lower your risk(variance) in return for lower expected returns (due to money on the side lines). I would only recommend this strategy if someone cannot withstand the risk levels of being fully invested.

If you are worried about the stock market being at all time highs and feel a short term decline is certain, why are you only worried about this 13% that is currently in cash? What about the 87% of your portfolio which is currently invested and fully at risk right now? I feel like most people would agree pulling some of this money out, waiting for a decline, and then re-investing would be market timing (which most people agree is a proven inferior investment strategy).

I would argue that if DCA is the optimal strategy (holding money out due to predicted market decline), then pulling existing funds out the market and waiting for the decline would also be optimal. Most people agree the latter is not optimal, therefore the former is also not optimal.

You should not look at dollars in your portfolio differently (besides tax implications) just because it is currently invested or not currently invested. That was yesterday, moving forward, it should be in the most optimal position.

MikeT
Posts: 389
Joined: Sun Jul 10, 2016 11:26 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by MikeT » Fri Jan 13, 2017 4:24 pm

Before posting this question, I setup an automatic xfer of 1/24th of the cash each month over 24 months.

With my plan, if the market drops from 20,000 to 18,000 in 3 months, I'd dive in with nearly all the cash (21/24ths of the cash). That would be great!

If it never drops, then I've learned a valuable lesson about my terrible skills of market timing.

I thought this was a great idea (until reading the responses to my question :-)

Mike

User avatar
Aptenodytes
Posts: 3751
Joined: Tue Feb 08, 2011 8:39 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by Aptenodytes » Fri Jan 13, 2017 4:30 pm

The trick is to decide on a scheme that gets you in the market fast and calm. Your scheme may be what you need. The algorithms and incantations and magic dust that work for one person fall flat for another. Deviations from lump sum are judged in terms of how well they get you invested quickly and calmly.

User avatar
Toons
Posts: 13063
Joined: Fri Nov 21, 2008 10:20 am
Location: Hills of Tennessee

Re: Dollar-Cost-Average over HOW MANY years?

Post by Toons » Fri Jan 13, 2017 4:32 pm

Another vote for all in.
20 years from now the current all time high,
will be in the rearview mirror. :happy
"One does not accumulate but eliminate. It is not daily increase but daily decrease. The height of cultivation always runs to simplicity" –Bruce Lee

ame
Posts: 122
Joined: Sun Sep 27, 2015 12:30 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by ame » Fri Jan 13, 2017 4:36 pm

MikeT wrote:I guess so.

Are you suggesting no dollar-cost averaging and just go all-in right now (1/13/17)?

_Mike
Some bogleheads claim (I agree) it's about time invested, not timing. I fully funded/invested my 2016 and 2017 Roth IRA on 1/11/17. However, I'm using dollar-cost averaging to fully fund my traditional 401K.

MikeT
Posts: 389
Joined: Sun Jul 10, 2016 11:26 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by MikeT » Fri Jan 13, 2017 4:44 pm

That's a really compelling argument:

@bantam222 "If you are worried about the stock market being at all time highs and feel a short term decline is certain, why are you only worried about this 13% that is currently in cash? What about the 87% of your portfolio which is currently invested and fully at risk right now? "

User avatar
jimb_fromATL
Posts: 2278
Joined: Sun Nov 10, 2013 12:00 pm
Location: Atlanta area & Piedmont Triad NC and Interstate 85 in between.

Re: Dollar-Cost-Average over HOW MANY years?

Post by jimb_fromATL » Fri Jan 13, 2017 6:56 pm

MikeT wrote:Within my IRA, my desired allocation is 80 stock / 20 bonds.
Presently, I'm at 67% stock, 20% bonds and 13% "cash".

Because the stock market is at an all-time high (nearly 20,000), I don't want to use the "cash" to buy the stock all at once..

Here's my plan: does this make sense:

Dollar cost average in over 2 years.
If the market drops to 19,000 buy stock with HALF remaining cash.
If the market drops to 18,000 bu stock with ALL remaining cash
Thoughts?

Thanks,
Mike
Does the warranty on your crystal ball include paying your for any loss if its predictions are not correct? If not, then if history is any indicator at all you're more likely to come out better by just putting it all in a single lump sum.

To give you an example of how lump sum compares to DCA:
  • In Vanguard's Wellington balanced fund VWELX,
    • For 432 rolling 12 month reporting periods since 01/1980 a lump sum invested in VWELX did better than DCA over those months 82.4% of the time. In the best case, lump sum did better than DCA by about 18.5%. In the worst case, it was worse by about -9.4%.

      For 420 rolling 24 month reporting periods since 01/1980 a lump sum invested in VWELX did better than DCA over those months 90.% of the time. In the best case, lump sum did better than DCA by about 19.9%. In the worst case, it was worse by about -0.3%.
    For the more volatile S&P 500 index fund VFINX:
    • For 432 rolling 12 month reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA over those months 76.9% of the time. In the best case, lump sum did better than DCA by about 19.5%. In the worst case, it was worse by about -14.1%.

      For 420 rolling 24 month reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA over those months 81.9% of the time. In the best case, lump sum did better than DCA by about 30.3%. In the worst case, it was worse by about -7.3%.
Since all the predictors of when to get in and when to get out are also based on how things have worked in the past, and without a working crystal ball, looks to me like just a lump sum is the better bet.


jimb

bantam222
Posts: 96
Joined: Sun Nov 23, 2014 5:57 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by bantam222 » Fri Jan 13, 2017 7:28 pm

Another example:
Ignoring tax implications and transactions fees of buying/selling, if your entire portfolio was liquidated to cash today (say Fidelity no longer serviced your area and you needed to move to Vanguard or something), how would you re-invest it?

Would you jump directly into a 87/13 mix of invested/cash? No? Why are you currently in this AA then?

Just because that happens to be the current AA you have should not impact the optimal AA for tomorrow. Your current AA should not influence your target AA, and that is exactly what DCA is doing.

Would you DCA back into the market starting from $0? That makes no sense, the only thing that change is you were forced to change brokers (between two brokers that have very similar offerings). Why would this fact make you reduce your stock from 87% to 0%?

The other thing that confuses me about people who vote for DCA: once you get to 100%, would you sell off 20% and then DCA the funds back up to 100% and keep repeating? I feel like most people would agree this is silly. Then why does it make sense to do this the first time??? Which is the optimal AA? 100% invested or 80% invested? It can't be both, and it doesn't keep changing every few months (or even years) unless you are trying to market time.

DCA only makes sense for people who have a large amount of money (for them) they need to add to their portfolio. If they don't think they can handle the large swings initially, this helps to ease them in. This is the same as someone using a lower AA that has non-optimal returns, but it helps them "sleep at night". If that is what it takes to feel comfortable, sure, give up some returns. But don't try to go around saying you are maximizing your returns with DCA

Grt2bOutdoors
Posts: 19544
Joined: Thu Apr 05, 2007 8:20 pm
Location: New York

Re: Dollar-Cost-Average over HOW MANY years?

Post by Grt2bOutdoors » Fri Jan 13, 2017 7:45 pm

Toons wrote:Another vote for all in.
20 years from now the current all time high,
will be in the rearview mirror. :happy
+1 It's time in the market that counts. I went all in a few days ago, join me and if the market tanks, we can commiserate together. An IRA is a long term investment vehicle, chances are you will not be withdrawing all of your assets in one fell swoop, rather you will either draw on it over a decade or two or three leaving the remainder to your estate/heirs or you will convert the assets to a Roth leaving the assets to heirs who could choose to draw from it tax free over their lifetimes (that could be many decades).
"One should invest based on their need, ability and willingness to take risk - Larry Swedroe" Asking Portfolio Questions

User avatar
Toons
Posts: 13063
Joined: Fri Nov 21, 2008 10:20 am
Location: Hills of Tennessee

Re: Dollar-Cost-Average over HOW MANY years?

Post by Toons » Sat Jan 14, 2017 8:09 am

Grt2bOutdoors wrote:
Toons wrote:Another vote for all in.
20 years from now the current all time high,
will be in the rearview mirror. :happy
+1 It's time in the market that counts. I went all in a few days ago, join me and if the market tanks, we can commiserate together. An IRA is a long term investment vehicle, chances are you will not be withdrawing all of your assets in one fell swoop, rather you will either draw on it over a decade or two or three leaving the remainder to your estate/heirs or you will convert the assets to a Roth leaving the assets to heirs who could choose to draw from it tax free over their lifetimes (that could be many decades).

+1
Bingo :happy
"One does not accumulate but eliminate. It is not daily increase but daily decrease. The height of cultivation always runs to simplicity" –Bruce Lee

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 9:31 am

I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?

Mofritty
Posts: 27
Joined: Thu Nov 06, 2014 1:58 pm
Location: "Somewhere in the middle of America"

Re: Dollar-Cost-Average over HOW MANY years?

Post by Mofritty » Sat Jan 14, 2017 10:02 am

I am still dollar cost averaging lump sum money from two years ago. My rationale was your rationale. After all, the market was up. Looking forward, all I could see was the potential risk of buying too high. Looking back now, I see clearly something that wasn't evident at the time: the real risk of leaving my money on the sidelines.
+1 for going all in at your desired allocation.

sixty40
Posts: 223
Joined: Sat Dec 06, 2014 11:53 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by sixty40 » Sat Jan 14, 2017 10:52 am

dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
I do not necessarily believe that DCA’g is to an investors advantage, but for me personally, it is something I (have to) do. I have a problem with the movement of large amounts of money from one asset class (cash) to another (stocks or bonds, etc.), and it makes me hyperventilate. Unfortunately for me it is part of my genetic makeup.

An example was 15 yrs ago when we bought our current house for $500k, we had about $1.0M in net assets at the time and made over $250k in income. One would think it would be not be a big deal, right. Well it was all good until it sank in near the signing of the final papers. I had to go see my doctor because I was so stressed that I thought I was going to have (or was having) a heart attack. It was one of the most stressful times in my life.

If I had to move $50k from my bank to just my Vanguard MM account, I still move it in maybe 2-3 increments.

My name is Tom and I am a DCA’r.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 11:07 am

sixty40 wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
I do not necessarily believe that DCA’g is to an investors advantage, but for me personally, it is something I (have to) do. I have a problem with the movement of large amounts of money from one asset class (cash) to another (stocks or bonds, etc.), and it makes me hyperventilate. Unfortunately for me it is part of my genetic makeup.

An example was 15 yrs ago when we bought our current house for $500k, we had about $1.0M in net assets at the time and made over $250k in income. One would think it would be not be a big deal, right. Well it was all good until it sank in near the signing of the final papers. I had to go see my doctor because I was so stressed that I thought I was going to have (or was having) a heart attack. It was one of the most stressful times in my life.

If I had to move $50k from my bank to just my Vanguard MM account, I still move it in maybe 2-3 increments.

My name is Tom and I am a DCA’r.
Thanks for the reply. I agree, and most of our discussions agree, that if reducing anxiety over a transaction can be helped by doing the transaction in stages, then that is a legitimate possibility. The assumption is that the transaction is clearly warranted in the first place.

I was more interested in why people ask if they "should" DCA, as if that is some magic sauce for making more money. The background on that is that genuine DCA in the original sense actually is a magic sauce for making more money, but that concept does not apply in the "dca" vs lump sum question.

Angelus359
Posts: 845
Joined: Tue Mar 04, 2014 12:56 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by Angelus359 » Sat Jan 14, 2017 11:55 am

I don't have money to DCA, so I just invest a set amount of money every monday from my paycheck
IT-DevOps System Administrator

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 12:39 pm

Angelus359 wrote:I don't have money to DCA, so I just invest a set amount of money every monday from my paycheck

That IS DCA.
To not be DCA you would buy a fixed number of shares every Monday, assuming you always had enough money available to do so. The difference between the two is that you pay a lower average price per share the first way compared to the second. That is the actual magic sauce that does work and is the original meaning of then term.

User avatar
iceport
Posts: 4008
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by iceport » Sat Jan 14, 2017 1:02 pm

dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
"Discipline matters more than allocation.” ─William Bernstein

sixty40
Posts: 223
Joined: Sat Dec 06, 2014 11:53 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by sixty40 » Sat Jan 14, 2017 1:14 pm

dbr wrote:
sixty40 wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
I do not necessarily believe that DCA’g is to an investors advantage, but for me personally, it is something I (have to) do. I have a problem with the movement of large amounts of money from one asset class (cash) to another (stocks or bonds, etc.), and it makes me hyperventilate. Unfortunately for me it is part of my genetic makeup.

An example was 15 yrs ago when we bought our current house for $500k, we had about $1.0M in net assets at the time and made over $250k in income. One would think it would be not be a big deal, right. Well it was all good until it sank in near the signing of the final papers. I had to go see my doctor because I was so stressed that I thought I was going to have (or was having) a heart attack. It was one of the most stressful times in my life.

If I had to move $50k from my bank to just my Vanguard MM account, I still move it in maybe 2-3 increments.

My name is Tom and I am a DCA’r.
Thanks for the reply. I agree, and most of our discussions agree, that if reducing anxiety over a transaction can be helped by doing the transaction in stages, then that is a legitimate possibility. The assumption is that the transaction is clearly warranted in the first place.

I was more interested in why people ask if they "should" DCA, as if that is some magic sauce for making more money. The background on that is that genuine DCA in the original sense actually is a magic sauce for making more money, but that concept does not apply in the "dca" vs lump sum question.
dbr, I agree with you that LS should be better than DCA, I just cannot get myself to do it.

If one goes by percentages, I would agree that Lump Sum (LS) has a higher probability of being higher at the end of the year than DCA, making LS the better bet. In any given yr the market as a whole has a higher chance of being up than down (there are more up yrs than down yrs), although the downside can be a greater drop. In the near term, the market has a higher percentage of being up, and in the long term the market is up, so LS is better than DCA.

As an investment strategy, if one has lump sum dollars to invest, I would also like to hear an argument for how DCA can be better than LS.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 1:17 pm

iceport wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
Yes, it is known that sometimes the outcome is that one would have been better off to have phased the investment. The problem is that this happens less often than the opposite and that it is not possible to know in advance which the case will be. I would not call this support for a strategy but rather the opposite. Or did I misunderstand the article?

Dandy
Posts: 5483
Joined: Sun Apr 25, 2010 7:42 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by Dandy » Sat Jan 14, 2017 1:31 pm

Unless it is a small amount I rarely recommend 100% lump or 100% DCA. When the market is high by some objective measure then I would recommend say 20% lump and automate the DCA monthly over the next 9 months to a year. When the Market drops in a month double up on the DCA that month. You should be fully invested within a year.

Gets you off the sideline, automates most of your investment over a reasonably short timeframe and if you double up when the market is down as suggested it will get you used to buying when the market is down.

If the market is down by some objective measure, then I would consider 50% or more down and automate monthly DCA the rest and again if the market drops during a month double up on the DCA amount.

This is not investment science it is just a way to deal with investing a large sum when markets are at an objective high or low and manage your fears a bit at those times. Often people just freeze and lose out when the market rebounds. Since equity markets usually rise over time most advice is just invest the lump sum immediately.

good luck

User avatar
slayed
Posts: 263
Joined: Wed Feb 01, 2012 3:07 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by slayed » Sat Jan 14, 2017 1:42 pm

MikeT wrote:Within my IRA, my desired allocation is 80 stock / 20 bonds.
Presently, I'm at 67% stock, 20% bonds and 13% "cash".

Because the stock market is at an all-time high (nearly 20,000), I don't want to use the "cash" to buy the stock all at once..

Here's my plan: does this make sense:

Dollar cost average in over 2 years.
If the market drops to 19,000 buy stock with HALF remaining cash.
If the market drops to 18,000 bu stock with ALL remaining cash.

Thoughts?

Thanks,
Mike
this is mental accounting pure and simple. if that 13% cash was currently invested in stocks, would you be taking it out of the market right now and going to cash? why not?

User avatar
iceport
Posts: 4008
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by iceport » Sat Jan 14, 2017 2:09 pm

dbr wrote:
iceport wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
Yes, it is known that sometimes the outcome is that one would have been better off to have phased the investment. The problem is that this happens less often than the opposite and that it is not possible to know in advance which the case will be. I would not call this support for a strategy but rather the opposite. Or did I misunderstand the article?
It's possible. The focus of the article is limiting (short term) downside risk. It appears the data supports DCA as a reasonably effective — though clearly not perfect — means of accomplishing that objective.
"Discipline matters more than allocation.” ─William Bernstein

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 2:16 pm

iceport wrote:
dbr wrote:
iceport wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
Yes, it is known that sometimes the outcome is that one would have been better off to have phased the investment. The problem is that this happens less often than the opposite and that it is not possible to know in advance which the case will be. I would not call this support for a strategy but rather the opposite. Or did I misunderstand the article?
It's possible. The focus of the article is limiting (short term) downside risk. It appears the data supports DCA as a reasonably effective — though clearly not perfect — means of accomplishing that objective.
And the downside risk as soon as the sum is fully invested and for decades thereafter? If one wants to limit downside risk one should not make the investment at all. This leads to an observation that the real issue may lie with people contemplating more risk than they really want to take.

User avatar
iceport
Posts: 4008
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by iceport » Sat Jan 14, 2017 2:41 pm

dbr wrote:
iceport wrote:
dbr wrote:
iceport wrote:
dbr wrote:I am trying to figure out the persistence of the idea that DCA is somehow a strategy to an investor's advantage. I wonder if some of the posters that ask about this could give some background on how the idea has occurred to them -- was it something read about or something someone told them, or was it an idea that seemed an obvious possibility, or what?

?
The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
Yes, it is known that sometimes the outcome is that one would have been better off to have phased the investment. The problem is that this happens less often than the opposite and that it is not possible to know in advance which the case will be. I would not call this support for a strategy but rather the opposite. Or did I misunderstand the article?
It's possible. The focus of the article is limiting (short term) downside risk. It appears the data supports DCA as a reasonably effective — though clearly not perfect — means of accomplishing that objective.
And the downside risk as soon as the sum is fully invested and for decades thereafter? If one wants to limit downside risk one should not make the investment at all. This leads to an observation that the real issue may lie with people contemplating more risk than they really want to take.
Notice I included the "short term" descriptor in the reply above.

I am led to an observation that those so furiously opposed to DCA typically argue simultaneously 1) that short term volatility matters little decades down the road; but also 2) that not being fully invested for 11 months (out of decades to come) is a huge error in logic or judgement.
"Discipline matters more than allocation.” ─William Bernstein

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 2:55 pm

iceport wrote:
dbr wrote:
iceport wrote:
dbr wrote:
iceport wrote: The idea persists because sometimes it actually is to an investor's advantage to DCA — believe it or not! That's just a simple fact, as you can clearly see, should you elect to review this interesting analysis of history: Do Not Dollar-Cost-Average for More than Twelve Months.

(The OP may be interested in reviewing that analysis as well...)
Yes, it is known that sometimes the outcome is that one would have been better off to have phased the investment. The problem is that this happens less often than the opposite and that it is not possible to know in advance which the case will be. I would not call this support for a strategy but rather the opposite. Or did I misunderstand the article?
It's possible. The focus of the article is limiting (short term) downside risk. It appears the data supports DCA as a reasonably effective — though clearly not perfect — means of accomplishing that objective.
And the downside risk as soon as the sum is fully invested and for decades thereafter? If one wants to limit downside risk one should not make the investment at all. This leads to an observation that the real issue may lie with people contemplating more risk than they really want to take.
Notice I included the "short term" descriptor in the reply above.

I am led to an observation that those so furiously opposed to DCA typically argue simultaneously 1) that short term volatility matters little decades down the road; but also 2) that not being fully invested for 11 months (out of decades to come) is a huge error in logic or judgement.
No, the objection is not that there is something wrong with short term DCA. The objection is that there is nothing helpful about it other than psychological comfort. There is no reason to answer the question "should I DCA?" with a yes. It is patently obvious that if someone wants to avoid volatility in the short run, then one should not invest in the short run. So why would someone ask? When one has decided to invest then one is exposed to both short run and long run volatility. What do people expect, that DCA is a magic potion to immunize a portfolio from risk?

User avatar
iceport
Posts: 4008
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by iceport » Sat Jan 14, 2017 3:07 pm

dbr wrote:No, the objection is not that there is something wrong with short term DCA. The objection is that there is nothing helpful about it other than psychological comfort.
Except in the non-trivial percentage of the time when it does serve a purpose beyond psychological comfort, which you are conveniently disregarding.
"Discipline matters more than allocation.” ─William Bernstein

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sat Jan 14, 2017 3:09 pm

iceport wrote:
dbr wrote:No, the objection is not that there is something wrong with short term DCA. The objection is that there is nothing helpful about it other than psychological comfort.
Except in the non-trivial percentage of the time when it does serve a purpose beyond psychological comfort, which you are conveniently disregarding.
Well, if a person wants to postpone taking risk for a short time, and that is considered helpful, then I have no objection to that. I don't consider it helpful.

User avatar
Peculiar_Investor
Posts: 1164
Joined: Thu Oct 20, 2011 12:23 am
Location: Calgary, AB
Contact:

Re: Dollar-Cost-Average over HOW MANY years?

Post by Peculiar_Investor » Sat Jan 14, 2017 10:22 pm

As would be expect, Vanguard has a viewpoint. I would suggest the OP read Dollar-cost averaging just means taking risk later
Vanguard wrote:In this paper, we compare the historical performance of dollar-cost averaging (DCA) with lump-sum investing (LSI) across three markets: the United States, the United Kingdom, and Australia. On average, we find that an LSI approach has outperformed a DCA approach approximately two-thirds of the time, even when results are adjusted for the higher volatility of a stock/bond portfolio versus cash investments. This finding is consistent with the fact that the returns of stocks and bonds exceeded that of cash over our study period in each of these markets.

We conclude that if an investor expects such trends to continue, is satisfied with his or her target asset allocation, and is comfortable with the risk/return characteristics of each strategy, the prudent action is investing the lump sum immediately to gain exposure to the markets as soon as possible. But if the investor is primarily concerned with minimizing downside risk and potential feelings of regret (resulting from lump-sum investing immediately before a market downturn), then DCA may be of use. Of course, any emotionally based concerns should be weighed carefully against both (1) the lower expected long-run returns of cash compared with stocks and bonds, and (2) the fact that delaying investment is itself a form of market-timing, something few investors succeed at.
So the answer seems to be, it depends on your comfort level.
Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams

User avatar
Dale_G
Posts: 3125
Joined: Tue Feb 20, 2007 5:43 pm
Location: Central Florida - on the grown up side of 81

Re: Dollar-Cost-Average over HOW MANY years?

Post by Dale_G » Sat Jan 14, 2017 10:35 pm

If the OP is interested in the maximum return, the probability is that DCA should be done over 1/365th of a year. Sometimes this has to be delayed to the next day the market is open.

Dale
Volatility is my friend

ralph124cf
Posts: 2059
Joined: Tue Apr 01, 2014 11:41 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by ralph124cf » Sun Jan 15, 2017 12:15 am

You have said that your desired asset allocation is 80/20, but the question that you asked seems to indicate that you are not really comfortable with that allocation. Perhaps revisit your IPS.

Ralph

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by *3!4!/5! » Sun Jan 15, 2017 12:36 am

MikeT wrote:Within my IRA, my desired allocation is 80 stock / 20 bonds.
Presently, I'm at 67% stock, 20% bonds and 13% "cash".


Because the stock market is at an all-time high (nearly 20,000), I don't want to use the "cash" to buy the stock all at once..
Here's my plan: does this make sense:
Dollar cost average in over 2 years.
If the market drops to 19,000 buy stock with HALF remaining cash.
If the market drops to 18,000 bu stock with ALL remaining cash.
This discussion of what to do in the future is pointless if we don't know what happened in the past.

How did you get into this situation to start with?

User avatar
jimb_fromATL
Posts: 2278
Joined: Sun Nov 10, 2013 12:00 pm
Location: Atlanta area & Piedmont Triad NC and Interstate 85 in between.

Re: Dollar-Cost-Average over HOW MANY years?

Post by jimb_fromATL » Sun Jan 15, 2017 8:42 am

Dale_G wrote:If the OP is interested in the maximum return, the probability is that DCA should be done over 1/365th of a year. Sometimes this has to be delayed to the next day the market is open.

Dale
A sampling of the S&P 500 index suggests that it wouldn't make enough difference compared to doing DCA monthly, and still not as good a bet as just investing the whole lump ASAP. Certainly not it there were any transaction fees, and probably not worth the time to key it in every day unless it could be set up to be done automatically.

Allowing for weekends and holidays, there are about 262 +/- days when stocks are traded in a year.
  • Daily: For 9075 rolling 262 daily reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA over those days 77.3% of the time.

    Monthly: For 432 rolling 12 monthly reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA over those months 76.9% of the time.
jimb

Dandy
Posts: 5483
Joined: Sun Apr 25, 2010 7:42 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by Dandy » Sun Jan 15, 2017 9:28 am

No doubt that since stocks rise over time the lump sum will probably turn out better than DCA. Seems like it is about 70/30 in favor of lump sum. What are the stats if one only did DCA when, say the S&P was at or within 2% of its historic high (or some better measure of high valuation). You would think the results would be closer than 70/30. I assume when the S&P is 20% below its historic high that Lump sum investing would beat DCA at even a higher rate than 70/30.

I would think that there is some more useful guideline than just do 100% lump sum all the time. Something that takes some objective measure of market value (even if that is somewhat flawed) and some combination of percentage of lump sum and DCA and some DCA timeframe. I wish I had the math/research skills to do it.

It would be great if such a study would suggest a range of outcomes something like:
1. When the market is at X (historic high)less y% (e.g. -20%) do 100% lump sum. This beats a 12 month DCA - 90% of the time

2. When it is at X less z% (say 10%) do 50% lump sum and DCA for 9 to 12 months. 100% lump sum beats this approach 60% of the time.

3. When it is within 2% or at its historic high do 100% DCA for 18 months. 100% lump sum and this approach is too close to call.

Suppose those made up stats above were true - wouldn't that be more helpful than lump sum wins 70% of the time so always do lump sum?

henrikk
Posts: 36
Joined: Sat Oct 06, 2007 9:23 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by henrikk » Sun Jan 15, 2017 9:49 am

sixty40 wrote: As an investment strategy, if one has lump sum dollars to invest, I would also like to hear an argument for how DCA can be better than LS.
To me this discussion has derailed. On average LS will yield a higher return since the market goes up more often than down. The problem is that with LS your probability of higher *or* lower returns are greater than if you DCA. The possible spread of outcomes is wider with LS than if you DCA. To me this is not a matter of reducing short term volatility but the probability you will have less (or more) whenever you need your money. Is the possibility of an increased downside with LS worth the possibility of an increased upside to you? The caveat is that the possibility of increased upside with LS *on average* is guaranteed (assuming historical behavior of the stock market continues). No one can answer that except you. I personally opt for DCA all the way.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sun Jan 15, 2017 9:58 am

henrikk wrote:
sixty40 wrote: As an investment strategy, if one has lump sum dollars to invest, I would also like to hear an argument for how DCA can be better than LS.
To me this discussion has derailed. On average LS will yield a higher return since the market goes up more often then down. The problem is that with LS your probability of higher *or* lower returns are greater than if you DCA. The possible spread of outcomes is wider with LS than if you DCA. To me this is not a matter of reducing short term volatility but the probability you will have less (or more) whenever you need your money. Is the possibility of an increased downside with LS worth the possibility of an increased upside to you? The caveat is that the possibility of increased upside with LS *on average* is guaranteed (assuming historical behavior of the stock market continues). No one can answer than except you. I personally opt for DCA all the way.
The derailment is with the thought process for why someone is investing at all if he thinks he might come out ahead by not investing. Of course it is possible to end up with more money after 12 or 18 months if you don't invest than if you did. If that is what one wants to try, then one should again not invest for another 12 or 18 months, and then again. In short, if the investor wants no exposure to losses, he shouldn't invest ever.

You are absolutely correct that it is about a wider range of outcomes if you invest. If the investor wants to narrow the range of outcomes they should invest less in risky investments and keep more in "safe" investments. If they want the ultimately narrow range of outcomes, they should not invest at all. And if the investor is afraid of negative outcomes and thinks DCA is a way to avoid them, then when everything is invested he should take it all out and to the same thing again. It might be more direct and less work to simply invest all the money in a less risky portfolio.

If a person really is afraid of putting in the lump sum today because he might need the money tomorrow, he absolutely should not invest in the first place.

This whole DCA thing is trying to have one's cake and eat it too.

henrikk
Posts: 36
Joined: Sat Oct 06, 2007 9:23 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by henrikk » Sun Jan 15, 2017 10:18 am

dbr wrote: If a person really is afraid of putting in the lump sum today because he might need the money tomorrow, he absolutely should not invest in the first place.

This whole DCA thing is trying to have one's cake and eat it too.
I disagree. If I ask someone to choose an investment with a 95% probability the return will be 7% +- 3% versus a similar alternative investment with a 95% probability the return 7.5 +- 5% then there are some who choose the first and others the second. Neither is right or wrong. The idea if you DCA instead of LS you should not be invested is simply wrong. This is not about trying to beat the market or "eating your cake" --- it is a choice we make on the investment returns and spread we want. Note: The numbers are made up and are not suggested to be representative of DCA vs LS.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sun Jan 15, 2017 10:27 am

henrikk wrote:
dbr wrote: If a person really is afraid of putting in the lump sum today because he might need the money tomorrow, he absolutely should not invest in the first place.

This whole DCA thing is trying to have one's cake and eat it too.
I disagree. If I ask someone to choose an investment with a 95% probability the return will be 7% +- 3% versus a similar alternative investment with a 95% probability the return 7.5 +- 5% then there are some who choose the first and others the second. Neither is right or wrong. The idea if you DCA instead of LS you should not be invested is simply wrong. This is not about trying to beat the market or "eating your cake" --- it is a choice we make on the investment returns and spread we want. Note: The numbers are made up and are not suggested to be representative of DCA vs LS.
That is not the debate. The issue between DCA and lump sum is only how quickly one wants to assume whatever choice one is making. If one wants to invest at 7% +/- 3% then one would surely invest there. The person who wants the other investment would surely invest there. Why would someone invest in the first investment for a few months and end up finally investing in the second one? That would make no sense at all.

This is one reason I believe that choosing to DCA is a cover up for not being sure what risk one really wants to take.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by dbr » Sun Jan 15, 2017 10:28 am

One thing for sure, I am starting to understand now why asking about DCA is such a persistent question.

pkcrafter
Posts: 13135
Joined: Sun Mar 04, 2007 12:19 pm
Location: CA
Contact:

Re: Dollar-Cost-Average over HOW MANY years?

Post by pkcrafter » Sun Jan 15, 2017 11:01 am

MikeT,

Maybe look at it this way and then decide. You want to be at 80% stock, so if it takes you two years you will then be exposed to the same potential market loss you are worried about now. :happy It can happen any time. If you are going to be worried about losing some amount at 80% stock, then maybe you should not go to 80%.

Another point is, we use 50% as a max drop, which of course it is not, but if you figure on that number, then at 80% equity your portfolio would lose 40% overall (50% of 80%). You are now at 67% stock, so max drop at that % is 33.5%. That is hardly enough difference to worry about, not to mention taking two years to make the change.

Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.

User avatar
Earl Lemongrab
Posts: 5873
Joined: Tue Jun 10, 2014 1:14 am

Re: Dollar-Cost-Average over HOW MANY years?

Post by Earl Lemongrab » Sun Jan 15, 2017 11:16 am

MikeT wrote:Are you suggesting no dollar-cost averaging and just go all-in right now (1/13/17)?
How much do you already hold in stocks? Are you going to sell those and DCA that money? Do you understand that there's very little difference?
This week's fortune cookie: "Your financial life will be secure and beneficial." So I got that going for me, which is nice.

User avatar
jimb_fromATL
Posts: 2278
Joined: Sun Nov 10, 2013 12:00 pm
Location: Atlanta area & Piedmont Triad NC and Interstate 85 in between.

Re: Dollar-Cost-Average over HOW MANY years?

Post by jimb_fromATL » Sun Jan 15, 2017 11:23 am

henrikk wrote:
dbr wrote: If a person really is afraid of putting in the lump sum today because he might need the money tomorrow, he absolutely should not invest in the first place.

This whole DCA thing is trying to have one's cake and eat it too.
I disagree. If I ask someone to choose an investment with a 95% probability the return will be 7% +- 3% versus a similar alternative investment with a 95% probability the return 7.5 +- 5% then there are some who choose the first and others the second. Neither is right or wrong. The idea if you DCA instead of LS you should not be invested is simply wrong. This is not about trying to beat the market or "eating your cake" --- it is a choice we make on the investment returns and spread we want. Note: The numbers are made up and are not suggested to be representative of DCA vs LS.
Let's use some real numbers to put the risk versus reward in perhaps a better perspective:
  • For 432 rolling 12 month reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA over those months 76.9% of the time.

    The worst 12 month period for a lump sum of $10,000 with a -43.33% APY would have been worth $5,667 while DCA had an APY of -34.02% and would have been worth $6,598. The best lump sum would have been worth $15,655 compared to the best DCA of $13,098.
It's actually even more complicated because you need to consider the potential gain or loss in the fund(s) where you're holding the lump sum while you disburse it via DCA to some other fund. That's because you might be able to contribute more or less to the DCA amount depending on how the remaining lump is doing in the original fund(s).

In the next example, I use the PMT() math formula for the increased amount that you can withdraw periodically from a lump sum that is invested in a fund that has a constant return.
  • For 432 rolling 12 month reporting periods since 01/1980 a lump sum invested in VFINX did better than DCA out of a fund earning a constant 3.% over those months 70.8% of the time.

    The worst 12 month period for a lump sum of $10,000 with a -43.33% APY would have been worth $5,667 while DCA out of the 3% fund had an an APY of -32.94% and would have been worth $6,706. The best lump sum would have been worth $15,655 compared to the best DCA of $13,312.

Also remember that the surplus or deficit at the end of the DCA period will compound for the rest of the time you have it. So the potential loss or gain in the short run can have a much bigger effect over a few more decades.

jimb

User avatar
iceport
Posts: 4008
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by iceport » Sun Jan 15, 2017 11:30 am

dbr wrote:One thing for sure, I am starting to understand now why asking about DCA is such a persistent question.
Right. It would appear that many smart, rational people are incapable of seeing this issue in more than one way.

For the record I can clearly and easily see the perspective you've been arguing from, though it is obvious you are unable or unwilling to see the problem in a different way. You are locked into focusing on an unintended consequence — the timing element — rather than the true, underlying intent: to diversify purchase prices. This feels a little like trying to get someone to see an autostereogram, with the people who simply can't see the three-dimensional image insisting it can't possibly be there. Well, technically, you're right. But some of us can see the ink pattern in a meaningfully different way.

Here's my best attempt to describe how DCA might look if you allow yourself to take the timing element out of the decision:
Pretend there is a roulette wheel that you can use to purchase into your desired allocation. There are 24 pockets, each with a different weighted portfolio price. You don't know today's portfolio value (if you were fully invested), and you don't know how each of the 24 pockets compares to today's value, but you are told that the average of the prices is slightly higher than today's value. However, there is significant variability among the prices, and 10 of the pockets are priced higher than today's prices, some significantly so. You'd love to be able to pick a pocket that has a price lower than or equal to today's portfolio value, otherwise you lose money. The trouble is, you are at the mercy of the roulette wheel.

You only have two options available to you today for investing in your portfolio:

1) You can place one big bet on one pocket, with your entire lump sum, hoping you get one of the lower prices; or
2) You can spread out your bets equally among all 24 pockets.

Which would you choose?


And with all that said, I feel obliged to divulge that I value and respect your contributions to this forum greatly, dbr. You are one of the unsung heroes here, providing consistently superb commentary and advice. This discussion is an interesting diversion, but I think we can both agree that the choice to DCA or not is *not* a critical decision.
"Discipline matters more than allocation.” ─William Bernstein

*3!4!/5!
Posts: 1256
Joined: Fri Sep 02, 2016 1:47 pm

Re: Dollar-Cost-Average over HOW MANY years?

Post by *3!4!/5! » Sun Jan 15, 2017 11:50 am

I think some people are falling for the "sunk-cost-fallacy"-fallacy.
*3!4!/5! wrote:
MikeT wrote:Within my IRA, my desired allocation is 80 stock / 20 bonds.
Presently, I'm at 67% stock, 20% bonds and 13% "cash".
This discussion of what to do in the future is pointless if we don't know what happened in the past.

How did you get into this situation to start with?

Post Reply