Dollar Cost Averaging inferior [but Vanguard disagrees]

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Topic Author
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by coachz » Sun Nov 08, 2015 5:44 am

If DCA is inferior I sure wish Vanguard reps would stop saying it's a good idea in their webinars. They consistently mention it as a good strategy when apparently it is not.

http://www.cbsnews.com/news/dollar-cost ... r-results/

LittleD
Posts: 220
Joined: Sun Jul 31, 2011 7:23 am

Re: Dollar Cost Averaging inferior

Post by LittleD » Sun Nov 08, 2015 5:56 am

This has been hashed over many times and you are right that DCA is not a strategy
that on average will turn out better than a lump sum move it all in the middle
investing. The reason some will say to use it is that many investors are very
risk averse and can't stomach the idea of pushing all their money into the market
and the next day watching it fall 20% or more. Even if DCA is not the ideal
investing strategy, in some cases it might make the most sense if you can
do it and stay the course.

User avatar
tadamsmar
Posts: 8579
Joined: Mon May 07, 2007 12:33 pm

Re: Dollar Cost Averaging inferior

Post by tadamsmar » Sun Nov 08, 2015 6:01 am

DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.

User avatar
BeBH65
Posts: 1590
Joined: Sat Jul 04, 2015 7:28 am

Re: Dollar Cost Averaging inferior

Post by BeBH65 » Sun Nov 08, 2015 6:05 am

From our DCA versus lump sum wiki page who contains more info.
In most cases, you are moving your money from cash (or the equivalent, a low-yielding money market) to some mix of stocks and bonds. The expected return of both stocks and bonds are higher than cash. However, their volatility is higher as well. The risk is that just after making your investment, the market could crash, causing you to feel bad that you invested when you did.
Edited 2x to correct the URL.
Last edited by BeBH65 on Sun Nov 08, 2015 9:18 am, edited 3 times in total.
BeBH65. (only an investment enthusiast, not a financial adviser, perform your due diligence). | Have a look at https://www.bogleheads.org/wiki/Outline_of_Non-US_domiciles

User avatar
JoMoney
Posts: 8130
Joined: Tue Jul 23, 2013 5:31 am

Re: Dollar Cost Averaging inferior

Post by JoMoney » Sun Nov 08, 2015 6:35 am

DCA is a strategy to reduce risk, not to maximize returns.
It will get your money invested at an average price over some period of time, guaranteeing your cost won't be 'all in' at some peak price.

Outside of making regular payroll contributions, most advisers aren't fans of DCA. But their motivations and 'risk tolerance' for your money may not be aligned the same way.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

ks289
Posts: 642
Joined: Sun Mar 11, 2012 12:42 pm

Re: Dollar Cost Averaging inferior

Post by ks289 » Sun Nov 08, 2015 6:46 am

tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
For the purposes of comparing DCA and lump sum, investing a % of each paycheck really is not DCA since you are presumably not holding anything back that you would otherwise invest. While on the surface it resembles DCA, you are really lump sum investing amounts regularly as soon as the money is earned. The term some would use instead of DCA is automatic continuous investing.
Last edited by ks289 on Sun Nov 08, 2015 9:26 am, edited 1 time in total.

GettingThere
Posts: 11
Joined: Wed Oct 14, 2015 12:01 pm

Re: Dollar Cost Averaging inferior

Post by GettingThere » Sun Nov 08, 2015 6:52 am

If you have $1M in cash, by all means, invest it in a lump sum if you wish.

I personally view every other Friday as a lump sum investment. Which is to say, every payday I invest my budgeted % of income.

I believe this will give me better returns than investing at the end of a long period of saving.

I don't expect to ever have $1M in cash just sitting there. The only possibilty is inheritance, but I hope my folks spend it. Or the lottery, which is pretty unlikely since I don't participate.

User avatar
Jerry55
Posts: 706
Joined: Tue Jan 27, 2015 1:56 am
Location: That Toddlin' Town

Re: Dollar Cost Averaging inferior

Post by Jerry55 » Sun Nov 08, 2015 7:33 am

I can't agree more w/the above posters ~ Here's MY take on it now...as a former Fed Employee.

FERS (and many in the private sector) get matching contributions from their employers, let's say 5% max.

That's per pay period of their salary. So, if they put in 10K on Jan. 2, but their pay is usually 1k, the employer

will only put in, up to $50.00. Now, the next pay period,they put in 8K, the employer will only match with

another $50.00. See where I'm heading ? Employee put in Max for the year, ($18,000.00)

but only got $100.00 from their employer. However, if employee put in 18K / 26 = $692. per pay period,

employer would've added 26*50= $1,300.00, instead of $100.00.

I was CSRS, so I, on the other hand, received no matching funds, and I would've been able to put in the entire

18K in my first Jan. paycheck if I made 20K per pay period, as we did not get any employer match.

Then again, if I was making 18K bi-weekly, I'd still be working too. :sharebeer

p.s. I understand if this was a taxable account and none of the above was/is relevant :happy
Retired CSRS on 12/19/2012 @ age 57 w/39 years | Good Bye Tension, Hello Pension !!!

User avatar
ruralavalon
Posts: 17065
Joined: Sat Feb 02, 2008 10:29 am
Location: Illinois

Re: Dollar Cost Averaging inferior

Post by ruralavalon » Sun Nov 08, 2015 9:04 am

tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
That is exactly right.

The confusion arises out of the fact that the term DCA ("dollar cost averaging") is applied to two very different situations: (1) investing a large lump sum in stages over time; and (2) investing a set sum out of every pay check.
"Everything should be as simple as it is, but not simpler." - Albert Einstein | Wiki article link:Getting Started

Grt2bOutdoors
Posts: 21726
Joined: Thu Apr 05, 2007 8:20 pm
Location: New York

Re: Dollar Cost Averaging inferior

Post by Grt2bOutdoors » Sun Nov 08, 2015 9:35 am

ruralavalon wrote:
tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
That is exactly right.

The confusion arises out of the fact that the term DCA ("dollar cost averaging") is applied to two very different situations: (1) investing a large lump sum in stages over time; and (2) investing a set sum out of every pay check.
+1 and to every other poster who responded.

I suppose my DCA strategy is quite inferior, so inferior that aghast! :oops: I might be able to retire one day with out having resorting to eating Alpo for the rest of my days.
"One should invest based on their need, ability and willingness to take risk - Larry Swedroe" Asking Portfolio Questions

User avatar
iceport
Posts: 4151
Joined: Sat Apr 07, 2007 4:29 pm

Re: Dollar Cost Averaging inferior

Post by iceport » Sun Nov 08, 2015 9:45 am

JoMoney wrote:DCA is a strategy to reduce risk, not to maximize returns.
It will get your money invested at an average price over some period of time, guaranteeing your cost won't be 'all in' at some peak price.

Outside of making regular payroll contributions, most advisers aren't fans of DCA. But their motivations and 'risk tolerance' for your money may not be aligned the same way.
That's exactly right.
BeBH65 wrote:From our DCA versus lump sum wiki page who contains more info.
In most cases, you are moving your money from cash (or the equivalent, a low-yielding money market) to some mix of stocks and bonds. The expected return of both stocks and bonds are higher than cash. However, their volatility is higher as well. The risk is that just after making your investment, the market could crash, causing you to feel bad that you invested when you did.
Edited 2x to correct the URL.
Unfortunately, the wiki article is biased against DCA. With a few very minor, innocuous tweaks, its neutrality could be greatly enhanced:

First paragraph, last sentence: "The answer depends on your psychology." Well, yes. But so does the choice of one's AA, though it probably isn't attributed so entirely to psychology. What if the term "psychology" were replaced with "aversion to loss"? That would still be accurate, without implying those who DCA are weak-minded. It does seem more neutral.

Second paragraph, last sentence: "The risk is that just after making your investment, the market could crash, causing you to feel bad that you invested when you did." Don't we all always "feel bad" when the market crashes and we lose money? The point is, if you had been just a bit later in your timing -- or had DCA'd -- your losses could have been substantially reduced. This alternate wording would still be accurate, without trivializing the very real, tangible loss: "The risk is that just after making your investment, the market could crash, causing your investment to lose value due to the unfortunate timing of your purchase."

Fourth paragraph, first sentence: "For a completely rational investor, lump sum investing will always produce a higher expected return..." Is that first clause really necessary? Lump sum investing always produces a higher expected return, for completely irrational and rational investors alike. The implication is that lump sum investing is completely rational (which it is), but DCA is not. If DCA can reduce risk, it can be a completely rational choice. Could the statement simply start with the fact, "Lump sum investing will always produce a higher expected return..."

Fifth paragraph, first sentence, and sixth paragraph, first sentence: rather than "regret," a more neutral word might be "loss." It is a real-life "loss," after all, that creates the "regret."
"Discipline matters more than allocation.” ─William Bernstein

Pete1981
Posts: 28
Joined: Tue May 12, 2015 11:02 am

Re: Dollar Cost Averaging inferior

Post by Pete1981 » Sun Nov 08, 2015 9:50 am

DCA certainly didn't work for me. I had about 100K and was going to invest 25K each quarter starting with Oct 1st this year. I put my first 25K in but instead of sticking to my plan I decided DCA wasn't the best strategy (probably chasing returns here) and got the rest of my money in by Oct 15th. As a result I missed out on two weeks of great returns. Luckily I still averaged about 4.5 percent gains, but I probably could have gotten closer to 10 had I just lump summed everything. Oh well, could have been worse.

User avatar
goodenyou
Posts: 2073
Joined: Sun Jan 31, 2010 11:57 pm
Location: Skating to Where the Puck is Going to Be..or on the golf course

Re: Dollar Cost Averaging inferior

Post by goodenyou » Sun Nov 08, 2015 10:00 am

DCA is a technique to overcome fear in investing. The science is settled (to steal a cliche) that lump sum investing is superior to DCA. The flaw in DCA is some investors stop investing when the market trends down because they still have control. DCA tempts market timing which is one of the greatest mistakes in investing.
"Ignorance more frequently begets confidence than does knowledge" | Do you know how to make a rain dance work? Dance until it rains.

ks289
Posts: 642
Joined: Sun Mar 11, 2012 12:42 pm

Re: Dollar Cost Averaging inferior

Post by ks289 » Sun Nov 08, 2015 10:01 am

Grt2bOutdoors wrote:
ruralavalon wrote:
tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
That is exactly right.

The confusion arises out of the fact that the term DCA ("dollar cost averaging") is applied to two very different situations: (1) investing a large lump sum in stages over time; and (2) investing a set sum out of every pay check.
+1 and to every other poster who responded.

I suppose my DCA strategy is quite inferior, so inferior that aghast! :oops: I might be able to retire one day with out having resorting to eating Alpo for the rest of my days.
It sounds like you may not actually be doing DCA at least in the sense that Larry Swedroe's piece or the wiki refers, but rather just doing #2 investing a set sum out of every paycheck as most people do. It is of course impossible to invest money that has not been earned yet so there is no alternative.
DCA being inferior with regard to expected returns is not to be confused with actual returns. I've had some good timing and some bad timing with lump sums, but if it turns out bad sometimes that's not in my control IMO.

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 10:10 am

iceport wrote:
BeBH65 wrote:From our DCA versus lump sum wiki page who contains more info.
In most cases, you are moving your money from cash (or the equivalent, a low-yielding money market) to some mix of stocks and bonds. The expected return of both stocks and bonds are higher than cash. However, their volatility is higher as well. The risk is that just after making your investment, the market could crash, causing you to feel bad that you invested when you did.
Edited 2x to correct the URL.
Unfortunately, the wiki article is biased against DCA. With a few very minor, innocuous tweaks, its neutrality could be greatly enhanced:

First paragraph, last sentence: "The answer depends on your psychology." Well, yes. But so does the choice of one's AA, though it probably isn't attributed so entirely to psychology. What if the term "psychology" were replaced with "aversion to loss"? That would still be accurate, without implying those who DCA are weak-minded. It does seem more neutral.

Second paragraph, last sentence: "The risk is that just after making your investment, the market could crash, causing you to feel bad that you invested when you did." Don't we all always "feel bad" when the market crashes and we lose money? The point is, if you had been just a bit later in your timing -- or had DCA'd -- your losses could have been substantially reduced. This alternate wording would still be accurate, without trivializing the very real, tangible loss: "The risk is that just after making your investment, the market could crash, causing your investment to lose value due to the unfortunate timing of your purchase."

Fourth paragraph, first sentence: "For a completely rational investor, lump sum investing will always produce a higher expected return..." Is that first clause really necessary? Lump sum investing always produces a higher expected return, for completely irrational and rational investors alike. The implication is that lump sum investing is completely rational (which it is), but DCA is not. If DCA can reduce risk, it can be a completely rational choice. Could the statement simply start with the fact, "Lump sum investing will always produce a higher expected return..."

Fifth paragraph, first sentence, and sixth paragraph, first sentence: rather than "regret," a more neutral word might be "loss." It is a real-life "loss," after all, that creates the "regret."
I think the difference between AA choice based on loss aversion and DCA vs Lump Sum based on loss aversion is that the former is static: "Here is how much risk I am willing to take on."
When someone DCAs instead of lump summing, they seem to be saying "my chosen AA has too much risk, so I want to slowly increase risk over time until it gets to the risk level that I chose as being right for me."
That seems irrational to me--mathematically expected returns are higher for lump sum, and the investor has already (presumably) chosen an AA with a level of risk they are comfortable with.
There is the same possibility of loss after DCA finishes... lets say you DCA in over the course of a year, and then right after all your money is in the market crashes. Same problem, right?

The answer seems to me to be: lower the risk of your AA if you aren't truly comfortable with it. If you are completely comfortable with it it doesn't make sense to me that you would have to ease slowly up to that risk level?

tibbitts
Posts: 9399
Joined: Tue Feb 27, 2007 6:50 pm

Re: Dollar Cost Averaging inferior

Post by tibbitts » Sun Nov 08, 2015 10:23 am

coachz wrote:If DCA is inferior I sure wish Vanguard reps would stop saying it's a good idea in their webinars. They consistently mention it as a good strategy when apparently it is not.

http://www.cbsnews.com/news/dollar-cost ... r-results/
I don't know what webinars you're referring to, but the definition of a "good" strategy isn't necessarily one that, on average, has produced the highest return. If that were the case, you'd be criticizing Vanguard for not recommending 100% (or higher) equities for everyone.

User avatar
goingup
Posts: 3714
Joined: Tue Jan 26, 2010 1:02 pm

Re: Dollar Cost Averaging inferior

Post by goingup » Sun Nov 08, 2015 10:25 am

I like Rick Ferri's thinking on this decision. He suggests that if the amount is 20% or less of your current portfolio, lump it in. If it's more than 20% then DCA over time.
http://www.rickferri.com/blog/investmen ... -lump-sum/

dbr
Posts: 31156
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar Cost Averaging inferior

Post by dbr » Sun Nov 08, 2015 10:36 am

goingup wrote:I like Rick Ferri's thinking on this decision. He suggests that if the amount is 20% or less of your current portfolio, lump it in. If it's more than 20% then DCA over time.
http://www.rickferri.com/blog/investmen ... -lump-sum/
I would be sure to read and understand that article rather than just follow the advice. I personally neither endorse nor dispute the advice. I think the article is a good one.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Dollar Cost Averaging inferior

Post by Maynard F. Speer » Sun Nov 08, 2015 10:43 am

I've got to say I've always felt that study is an example of Vanguard using overly reductive logic - to be point it gives a misleadingly simple answer

We could take the same logic further with leverage ... Lump sum with 2x leverage at most points in history and you'll do even better - why miss out on those returns? .. but what if you'd invested your $1million (x2) in 2000? Or 2008? Or in Japan 1990?

We can find these 30-60 year periods of flat stock returns, but the mistake would've been lump-summing ... By DCA, you're reducing your exposure to worst-case-timing events so dramatically, you can't possibly reduce the whole argument to an average return

An average lump sum return - over different periods - is effectively DCA (just with more compounding) .. When you lump sum, you'll never get the average .. It's misleading

However I'd also agree it's really an AA issue .. My own suggestion (which I think is about optimal in terms of risk/reward) would be to lump sum, but with no more than a 35% equity allocation ... Which is the current market allocation to stocks ... If I wanted an equity-heavy portfolio, I'd use some kind of DCA, or AA adjustment, and I'd want to spread asset purchases over about 5-10 years ... Otherwise it's too much like gambling - overweighting stocks against the market, and purchasing in any period under 5 years
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

User avatar
LadyGeek
Site Admin
Posts: 58615
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by LadyGeek » Sun Nov 08, 2015 10:56 am

I retitled the thread. The OP can change the thread title further by editing the Subject: line in Post #1.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 11:11 am

Maynard F. Speer wrote:I've got to say I've always felt that study is an example of Vanguard using overly reductive logic - to be point it gives a misleadingly simple answer

We could take the same logic further with leverage ... Lump sum with 2x leverage at most points in history and you'll do even better - why miss out on those returns? .. but what if you'd invested your $1million (x2) in 2000? Or 2008? Or in Japan 1990?

We can find these 30-60 year periods of flat stock returns, but the mistake would've been lump-summing ... By DCA, you're reducing your exposure to worst-case-timing events so dramatically, you can't possibly reduce the whole argument to an average return

An average lump sum return - over different periods - is effectively DCA (just with more compounding) .. When you lump sum, you'll never get the average .. It's misleading

However I'd also agree it's really an AA issue .. My own suggestion (which I think is about optimal in terms of risk/reward) would be to lump sum, but with no more than a 35% equity allocation ... Which is the current market allocation to stocks ... If I wanted an equity-heavy portfolio, I'd use some kind of DCA, or AA adjustment, and I'd want to spread asset purchases over about 5-10 years ... Otherwise it's too much like gambling - overweighting stocks against the market, and purchasing in any period under 5 years
I understand following market weights within an asset class--a particular asset class has certain risk/reward properties, and trying to pick individual assets within that class will generally be inferior in expected returns to relying on the market weighting of individual assets within that class.

I'm not following when you extrapolate this to market weights across asset classes--different investors have different needs. An individual investor likely has very different goals than an intistution. E.g., many large companies may wish to keep a rather large cash buffer for various reasons--these will increase market weightings of cash/short term bonds.
I don't see how that would lead me to think that a larger weighting of cash/short term bonds is right for me?

It seems to me that I should choose an allocation of different asset classes based on my needs, and then let the market's knowledge (market weightings) handle the asset classes themselves. E.g., if I am going to invest in stock, I invest at market weightings.


As to the DCA aspect, again, how is this any different once you have fully invested?
So you spread your purchase over 5-10 years (holding cash for the uninvested part? ), and then after you are done your risk is in the same place it would have been at the start if you lump summed. If it seems like gambling to put your investable money into the market in a lump, how is it not gambling to leave all your investable money in the market after your DCA has finished?
Would you sell it all again as soon as DCA has finished and start the process over again?

If not, why is the situation different than when you had some cash to invest in the first place? Money is fungible--it shouldn't make a difference in your strategy whether or not the money is currently invested in the market or held in cash; you can easily change its state from one to the other.
If having all the money in the market is too risky/gambling/etc, that should hold true after your DCA has finished too...

User avatar
ruralavalon
Posts: 17065
Joined: Sat Feb 02, 2008 10:29 am
Location: Illinois

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by ruralavalon » Sun Nov 08, 2015 11:14 am

"Dollar cost averaging" (DCA) by investing a large lump sum in stages over time is usually not the better idea. "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." Vanguard paper, "Dollar-cost averaging just means taking risk later". This Vanguard paper reaches the same conclusion about investing a large lump sum in stages over time as the Swedroe article linked in the original post. So Vanguard does not disagree about that.

"Dollar cost averaging" (DCA) by investing a set sum out of every pay check is usually a good idea.

Is this so difficult to grasp?

EDITED to add link to Vanguard paper and quote.
Last edited by ruralavalon on Sun Nov 08, 2015 11:25 am, edited 2 times in total.
"Everything should be as simple as it is, but not simpler." - Albert Einstein | Wiki article link:Getting Started

selftalk
Posts: 1096
Joined: Thu Mar 08, 2012 10:08 am

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by selftalk » Sun Nov 08, 2015 11:22 am

If I acquire money from my job bi- weekly and put it in the bank to lump sum invest the money let`s say every 365 days and during that time the market is a bull market and rises mostly throughout the year and then I put in the money after a large rise. How can that be better than putting in the money throughout the year dollar cost averaging ? When you have the money in your possession then invest it and let the compounding work for you.
Last edited by selftalk on Sun Nov 08, 2015 11:37 am, edited 1 time in total.

dbr
Posts: 31156
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by dbr » Sun Nov 08, 2015 11:23 am

ruralavalon wrote:"Dollar cost averaging" (DCA) by investing a large lump sum in stages over time is usually not the better idea.

"Dollar cost averaging" (DCA) by investing a set sum out of every pay check is usually the better idea.

Is this so difficult to grasp?
It is for people who . . .

1) suddenly have a large amount of cash and are gobsmacked about the prospect of losing some of it in a risky investment and think there is a magic process by which that can be avoided. The actual answer for such people is to not invest all of that money in a risky investment ever.

2) think somehow that some kind of market timing can improve returns or reduce risk.

3) are confused when they have read about this: http://www.investopedia.com/terms/d/dol ... raging.asp and think they should apply it to investing a lump sum.

ks289
Posts: 642
Joined: Sun Mar 11, 2012 12:42 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by ks289 » Sun Nov 08, 2015 11:26 am

ruralavalon wrote:"Dollar cost averaging" (DCA) by investing a large lump sum in stages over time is usually not the better idea.

"Dollar cost averaging" (DCA) by investing a set sum out of every pay check is usually the better idea.

Is this so difficult to grasp?
Some people are truly in disagreement about #1, but I do not think that anyone is arguing that #2 is a bad idea. What is confusing is why this continues to enter into the DCA discussion at all.
Is regular paycheck investing good? Sure. It is better compared to what?

quantAndHold
Posts: 3522
Joined: Thu Sep 17, 2015 10:39 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by quantAndHold » Sun Nov 08, 2015 11:38 am

It's a risk/reward thing. Given that "the stock market always rises over time," investing a lump sum as a lump has the potential for higher rewards. But it also riskier than DCAing the money in. The "right" answer depends on the investor's tolerance for risk. If I'd gotten a lump sum when I was 25, I would have just plopped it into the market. If the market dropped 25% the next day, I would have been annoyed, but mostly unperturbed. Now I'm two years from retirement and more risk averse. When I got a lump sum earlier this year, I decided to dollar cost average it into my regular AA over the course of a year. The lump sum was enough that it matters to my retirement, and a 25% drop right after I invested the whole lump would have changed my retirement plans.

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by Morik » Sun Nov 08, 2015 11:41 am

quantAndHold wrote:It's a risk/reward thing. Given that "the stock market always rises over time," investing a lump sum as a lump has the potential for higher rewards. But it also riskier than DCAing the money in. The "right" answer depends on the investor's tolerance for risk. If I'd gotten a lump sum when I was 25, I would have just plopped it into the market. If the market dropped 25% the next day, I would have been annoyed, but mostly unperturbed. Now I'm two years from retirement and more risk averse. When I got a lump sum earlier this year, I decided to dollar cost average it into my regular AA over the course of a year. The lump sum was enough that it matters to my retirement, and a 25% drop right after I invested the whole lump would have changed my retirement plans.
I'm guessing your AA isn't 100% equities though--at 2 years from retirement you presumably have a reasonably high bond allocation.
So a 25% drop in equities would have dropped your portfolio by how much? (50/50 stock/bond, 12.5% drop?)

I still am trying to get an answer to this (as a general question to those who DCA): if you felt it was too risky to just drop the money in, why is it not still to risky once all the money is in?

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Dollar Cost Averaging inferior

Post by Maynard F. Speer » Sun Nov 08, 2015 11:46 am

Morik wrote:
Maynard F. Speer wrote:I've got to say I've always felt that study is an example of Vanguard using overly reductive logic - to be point it gives a misleadingly simple answer

We could take the same logic further with leverage ... Lump sum with 2x leverage at most points in history and you'll do even better - why miss out on those returns? .. but what if you'd invested your $1million (x2) in 2000? Or 2008? Or in Japan 1990?

We can find these 30-60 year periods of flat stock returns, but the mistake would've been lump-summing ... By DCA, you're reducing your exposure to worst-case-timing events so dramatically, you can't possibly reduce the whole argument to an average return

An average lump sum return - over different periods - is effectively DCA (just with more compounding) .. When you lump sum, you'll never get the average .. It's misleading

However I'd also agree it's really an AA issue .. My own suggestion (which I think is about optimal in terms of risk/reward) would be to lump sum, but with no more than a 35% equity allocation ... Which is the current market allocation to stocks ... If I wanted an equity-heavy portfolio, I'd use some kind of DCA, or AA adjustment, and I'd want to spread asset purchases over about 5-10 years ... Otherwise it's too much like gambling - overweighting stocks against the market, and purchasing in any period under 5 years
I understand following market weights within an asset class--a particular asset class has certain risk/reward properties, and trying to pick individual assets within that class will generally be inferior in expected returns to relying on the market weighting of individual assets within that class.

I'm not following when you extrapolate this to market weights across asset classes--different investors have different needs. An individual investor likely has very different goals than an intistution. E.g., many large companies may wish to keep a rather large cash buffer for various reasons--these will increase market weightings of cash/short term bonds.
I don't see how that would lead me to think that a larger weighting of cash/short term bonds is right for me?

It seems to me that I should choose an allocation of different asset classes based on my needs, and then let the market's knowledge (market weightings) handle the asset classes themselves. E.g., if I am going to invest in stock, I invest at market weightings.


As to the DCA aspect, again, how is this any different once you have fully invested?
So you spread your purchase over 5-10 years (holding cash for the uninvested part? ), and then after you are done your risk is in the same place it would have been at the start if you lump summed. If it seems like gambling to put your investable money into the market in a lump, how is it not gambling to leave all your investable money in the market after your DCA has finished?
Would you sell it all again as soon as DCA has finished and start the process over again?

If not, why is the situation different than when you had some cash to invest in the first place? Money is fungible--it shouldn't make a difference in your strategy whether or not the money is currently invested in the market or held in cash; you can easily change its state from one to the other.
If having all the money in the market is too risky/gambling/etc, that should hold true after your DCA has finished too...
The market-weighted allocation across asset classes is actually the logical conclusion of Efficient Markets Theory - as recommended by William Sharpe

Studies (I think from Ibbotson and Arnott?) show asset allocation accounts for over 100% of long-term returns - being that other variables, such as market timing and fees, have an aggregate negative return

You can have a cap-weighted stocks portfolio, but that doesn't tell you whether the market believes there's excessive risk (vs return) in stocks, and isn't holding many ... And this is really what market efficiency comes down to - weighing up risks across securities, sectors, regions and asset classes

It can be argued a 60:40 portfolio is no less an active bet against the market than one which overweights Biotech and micro-caps, or Apple and Tullow Oil

But of course, your investing aims and horizon don't necessarily match the market's - although the market's may be most efficient ... The most efficient solution should therefore be to leverage the global market portfolio - if you can leverage cheaply .. And that is difficult to argue against

On the other hand, as the market portfolio is probably best at dealing with risk, you could add risk by holding more e.g. small-cap value, or private equity, but still maintaining the AA ... It's worth pondering

Re: DCA

One way you can think of it is that every time period is its own asset class - investing in stocks in the 1980s vs the 2000 resulted in very different returns and drawdowns ... Lined up, they'd look like two different asset classes

You're not just buying stocks, you're buying the economic conditions of the era, relative to the one you're moving into

Diversification by time has just as significant an impact on risk/reward as asset allocation ... If you'd lump summed in 2000 (top of the market) you'd have done quite poorly by 2010 .. If you'd DCAed, you'd have invested the same amount by 2010, but would have bought at many better prices getting there
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 12:00 pm

Maynard F. Speer wrote: The market-weighted allocation across asset classes is actually the logical conclusion of Efficient Markets Theory - as recommended by William Sharpe

Studies (I think from Ibbotson and Arnott?) show asset allocation accounts for over 100% of long-term returns - being that other variables, such as market timing and fees, have an aggregate negative return

You can have a cap-weighted stocks portfolio, but that doesn't tell you whether the market believes there's excessive risk (vs return) in stocks, and isn't holding many ... And this is really what market efficiency comes down to - weighing up risks across securities, sectors, regions and asset classes

It can be argued a 60:40 portfolio is no less an active bet against the market than one which overweights Biotech and micro-caps, or Apple and Tullow Oil
I'm not convinced--Biotech, micro-caps, and individual stocks are all equities, and I expect that over time they should perform roughly similarly. (I.e., I have seen no evidence that the expected return & volatility of each differs much.)

Bonds, on the other hand, are a different animal than these--the legal regulations around bankruptcy & the nature of how loans work vs how stocks work (from a financial/mechanical standpoint) means that I very much expect a bond to have a different expected return & expected volatility compared to stocks.

Shouldn't I choose asset class weightings based on my goals, given that broad asset classes have different expected returns & volatility?

(Whereas when trying to choose amongst things that have similar expected return & volatility, I choose to own all of them, in the weightings the market has. Because the market should theoretically best choose the mix to optimize the expected return & volatility of that class. At least, it can theoretically optimize it better than anything I could do, or anyone else I could rely on.)
Maynard F. Speer wrote: Re: DCA

One way you can think of it is that every time period is its own asset class - investing in stocks in the 1980s vs the 2000 resulted in very different returns and drawdowns ... Lined up, they'd look like two different asset classes

You're not just buying stocks, you're buying the economic conditions of the era, relative to the one you're moving into

Diversification by time has just as significant an impact on risk/reward as asset allocation ... If you'd lump summed in 2000 (top of the market) you'd have done quite poorly by 2010 .. If you'd DCAed, you'd have invested the same amount by 2010, but would have bought at many better prices getting there
And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).

dbr
Posts: 31156
Joined: Sun Mar 04, 2007 9:50 am

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by dbr » Sun Nov 08, 2015 12:04 pm

ks289 wrote: Is regular paycheck investing good? Sure. It is better compared to what?
The origin of the term DCA, and the only actually meaningful definition, is the distinction between a practice of buying a fixed number of shares every period and a practice of investing a fixed dollar amount every period. If you invest a fixed dollar amount you buy more shares when stock prices are down and fewer when stock prices are up. This results in buying stock at a lower average price than if one buys the same fixed number of shares each period. The result is obviously desirable and probably more convenient when one receives money in constant amounts at fixed intervals. The issue only even arises in an older time when buying shares in round lots of fixed size was standard. Today a person investing in mutual funds would hardly even consider investing by buying 100 shares exactly every month; your just put in $429.67 (or whatever) every paycheck.

It is this lingering misunderstanding that is behind some of the mythology going on here.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Dollar Cost Averaging inferior

Post by Maynard F. Speer » Sun Nov 08, 2015 12:52 pm

Morik wrote:I'm not convinced--Biotech, micro-caps, and individual stocks are all equities, and I expect that over time they should perform roughly similarly. (I.e., I have seen no evidence that the expected return & volatility of each differs much.)

Bonds, on the other hand, are a different animal than these--the legal regulations around bankruptcy & the nature of how loans work vs how stocks work (from a financial/mechanical standpoint) means that I very much expect a bond to have a different expected return & expected volatility compared to stocks.

Shouldn't I choose asset class weightings based on my goals, given that broad asset classes have different expected returns & volatility?

(Whereas when trying to choose amongst things that have similar expected return & volatility, I choose to own all of them, in the weightings the market has. Because the market should theoretically best choose the mix to optimize the expected return & volatility of that class. At least, it can theoretically optimize it better than anything I could do, or anyone else I could rely on.)
Well there's no guarantee stocks will beat bonds over any period - they trailed bonds for about 73 years in the 19th century, and stocks are trailing long-term bonds now since about 1997 ... But if you follow the idea that the market's essentially compensating you to take risk, then riskier assets should yield higher long-term returns:

Since the 70s and 80s for example (CAGR):
- US Stock Market: 10.4
- Small-Cap Value: 13.36
- Active Micro-cap: 14.38
- Private Equity: 15.3


https://en.wikipedia.org/wiki/Fama%E2%8 ... ctor_model

And yes, the prevailing wisdom here is that you adjust your risk exposure through your stock and bond allocations ... But that approach is open to criticism - overweighting stocks means you do have to be a little lucky:

The Biggest Urban Legend in Finance - Rob Arnott
https://www.researchaffiliates.com/Our% ... egend.aspx

And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).
This is what Vanguard's study shows: on average you'd have done better lump-summing .. The market's spent more time going up than going down, so you're more likely to have benefited from bull markets and compounding than you'd have saved by avoiding lump-summing in bad years

But the problem with lump-summing is: you don't get the average ... The average completely ameliorates risk ... The reason you DCA or hold a defensive AA is to avoid bad timing, and minimise the damage of worst-case or black swan events

If someone came to me looking to invest $1million, I'd suggest a defensive AA, in line with the market's .. You'll always meet people who invested heavily right before a market crash, and as every crash is different, it's not as simple as saying "Stay the course", and guaranteeing a swift recovery - we often use Japan's 1990 market crash as a recent worst-case scenario - many people will have invested their retirement savings in 1990

Image
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

engineer1969
Posts: 102
Joined: Tue Jul 05, 2011 6:56 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by engineer1969 » Sun Nov 08, 2015 1:15 pm

Isn't DCA just the "effect" of periodically investing a nearly constant amount of funds?
I don't really view it as a strategy, but the result of employing a periodic investment strategy.

Whether or not to invest in a lump sum or periodically comes down to asset allocation and what risks you are willing to take. It seems what "feels" right is never what makes the most sense looking at the raw numbers.

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 1:24 pm

Maynard F. Speer wrote:
And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).
This is what Vanguard's study shows: on average you'd have done better lump-summing .. The market's spent more time going up than going down, so you're more likely to have benefited from bull markets and compounding than you'd have saved by avoiding lump-summing in bad years

But the problem with lump-summing is: you don't get the average ... The average completely ameliorates risk ... The reason you DCA or hold a defensive AA is to avoid bad timing, and minimise the damage of worst-case or black swan events

If someone came to me looking to invest $1million, I'd suggest a defensive AA, in line with the market's .. You'll always meet people who invested heavily right before a market crash, and as every crash is different, it's not as simple as saying "Stay the course", and guaranteeing a swift recovery - we often use Japan's 1990 market crash as a recent worst-case scenario - many people will have invested their retirement savings in 1990

Image
And my question was--why would you choose to DCA a large amount of money over time, but would not sell once you are fully invested and start DCAing again?

The same rationales apply--once you are fully invested you aren't getting the average, right?

Also, can you explain more about how getting the average price completely ameliorates risk?
(Can you clarify your meaning? I assume you don't mean it renders the investment risk-free... but complete amelioration implies that.)

engineer1969
Posts: 102
Joined: Tue Jul 05, 2011 6:56 pm

Re: Dollar Cost Averaging inferior

Post by engineer1969 » Sun Nov 08, 2015 1:32 pm

Morik wrote:
And my question was--why would you choose to DCA a large amount of money over time, but would not sell once you are fully invested and start DCAing again?

The same rationales apply--once you are fully invested you aren't getting the average, right?

Also, can you explain more about how getting the average price completely ameliorates risk?
(Can you clarify your meaning? I assume you don't mean it renders the investment risk-free... but complete amelioration implies that.)
This is a good point. I think investing the lump sum and rebalancing at a predetermined interval or allocation shift makes the most sense for me.

User avatar
Aptenodytes
Posts: 3762
Joined: Tue Feb 08, 2011 8:39 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by Aptenodytes » Sun Nov 08, 2015 1:38 pm

There are demonstrations of complexity theory that show how by combining a few very simple mechanisms in a linked system you can generate seemingly bizarre behaviors. I think of our collective instability in this debate as reflecting a similar kind of dynamic.

These are the simple elements in question.

1) A portfolio of stocks and bonds outperforms cash on average, with the likelihood of outperformance increasing with time invested.

2) investors are prone to predictable psychological foibles centered around feelings of regret, and these foibles are highly influenced by the way losses are framed. Losses that are materially identical can trigger very different feelings of regret if they are framed differently. Moving a lot of money into stocks just before a crash and leaving money in stocks just before a crash are examples of materially identical losses that can trigger very different feelings of regret because of how they are framed.

3) an investor's success depends on having a sound plan and being able to execute the plan. A good plan poorly executed can be as damaging as a poor plan well executed.

4) poor execution is common, and often triggered by anxiety over losses. Such anxiety will be shaped for most people by both the material and the psychological.

I think we would all accept each of these four fairly simple statements as completely true. Yet their interaction generates complexity that has rendered it impossible for anyone to write a simple guide to the lump sum question that we all agree on.

The complexity affects how you think about AA choice, withdrawal strategies and other things in the same way. There is no simple equilibrium in the things because of the interaction of the financial and psychological elements.

[edited to correct typos]

cherijoh
Posts: 6379
Joined: Tue Feb 20, 2007 4:49 pm
Location: Charlotte NC

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by cherijoh » Sun Nov 08, 2015 1:57 pm

ks289 wrote:
ruralavalon wrote:"Dollar cost averaging" (DCA) by investing a large lump sum in stages over time is usually not the better idea.

"Dollar cost averaging" (DCA) by investing a set sum out of every pay check is usually the better idea.

Is this so difficult to grasp?
Some people are truly in disagreement about #1, but I do not think that anyone is arguing that #2 is a bad idea. What is confusing is why this continues to enter into the DCA discussion at all.
Because depending on the source, "dollar cost averaging" can refer to either #1 or #2. :happy If everyone agreed to consistently use something like "periodic payroll investing" for # 2 then it wouldn't be a big issue.

But I have seen recommendations that employees with high salaries (or a big bonus in 1Q) front load their 401k contributions to get the benefit of compounding. So IMO, something of a hybrid between lump summing a windfall and periodically investing a biweekly paycheck.

FWIW, this year I decided to temporarily increase my 401K contribution to absorb a large chunk of my bonus and then reset it back to a rate that would evenly space out my remaining contributions. (Each bi-weekly pay period still has a large enough contribution to get the full match). Based on the mutual fund prices and dividend payments to date in my 401K, I think I am currently behind where I would have been had I evenly spaced my contributions. Just an example of "expected" and "actual" performance diverging.

tibbitts
Posts: 9399
Joined: Tue Feb 27, 2007 6:50 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by tibbitts » Sun Nov 08, 2015 2:05 pm

Morik wrote:
quantAndHold wrote:It's a risk/reward thing. Given that "the stock market always rises over time," investing a lump sum as a lump has the potential for higher rewards. But it also riskier than DCAing the money in. The "right" answer depends on the investor's tolerance for risk. If I'd gotten a lump sum when I was 25, I would have just plopped it into the market. If the market dropped 25% the next day, I would have been annoyed, but mostly unperturbed. Now I'm two years from retirement and more risk averse. When I got a lump sum earlier this year, I decided to dollar cost average it into my regular AA over the course of a year. The lump sum was enough that it matters to my retirement, and a 25% drop right after I invested the whole lump would have changed my retirement plans.
I'm guessing your AA isn't 100% equities though--at 2 years from retirement you presumably have a reasonably high bond allocation.
So a 25% drop in equities would have dropped your portfolio by how much? (50/50 stock/bond, 12.5% drop?)

I still am trying to get an answer to this (as a general question to those who DCA): if you felt it was too risky to just drop the money in, why is it not still to risky once all the money is in?
With DCA you end up with less equity risk for a shorter time. As you gradually invest more into equities using DCA - and the simplest example is over an entire investing lifetime, which is how most people invest - you're probably also decreasing equities as a percentage in your allocation.

Looking back on my career, if I could have divided up my income equally over the years, and my corresponding equity (retirement) investments as well, I would have done that. There was definitely negative time value in terms of equity investments (but there was positive value at times with fixed income investments.) I wound up generally investing by far the most at the worst possible times for equities, which were the best times for my business.

User avatar
tadamsmar
Posts: 8579
Joined: Mon May 07, 2007 12:33 pm

Re: Dollar Cost Averaging inferior

Post by tadamsmar » Sun Nov 08, 2015 2:14 pm

ks289 wrote:
tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
For the purposes of comparing DCA and lump sum, investing a % of each paycheck really is not DCA since you are presumably not holding anything back that you would otherwise invest. While on the surface it resembles DCA, you are really lump sum investing amounts regularly as soon as the money is earned. The term some would use instead of DCA is automatic continuous investing.
I'm describing how the term is used. The op was not comparing anything. He was making unsorced claims. I asked for a source.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Dollar Cost Averaging inferior

Post by Maynard F. Speer » Sun Nov 08, 2015 2:29 pm

Morik wrote:
Maynard F. Speer wrote:
And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).
This is what Vanguard's study shows: on average you'd have done better lump-summing .. The market's spent more time going up than going down, so you're more likely to have benefited from bull markets and compounding than you'd have saved by avoiding lump-summing in bad years

But the problem with lump-summing is: you don't get the average ... The average completely ameliorates risk ... The reason you DCA or hold a defensive AA is to avoid bad timing, and minimise the damage of worst-case or black swan events

If someone came to me looking to invest $1million, I'd suggest a defensive AA, in line with the market's .. You'll always meet people who invested heavily right before a market crash, and as every crash is different, it's not as simple as saying "Stay the course", and guaranteeing a swift recovery - we often use Japan's 1990 market crash as a recent worst-case scenario - many people will have invested their retirement savings in 1990

Image
And my question was--why would you choose to DCA a large amount of money over time, but would not sell once you are fully invested and start DCAing again?

The same rationales apply--once you are fully invested you aren't getting the average, right?

Also, can you explain more about how getting the average price completely ameliorates risk?
(Can you clarify your meaning? I assume you don't mean it renders the investment risk-free... but complete amelioration implies that.)
Sequence of returns ... If you lump-sum then hit all your bad years, it may take decades to recover your losses .. a 50% loss takes a 100% return to break even .. an 80% loss requires a 400% return to break even

If your 50% loss came during the DCA years, you'd at least be buying an once-in-a-generation prices ... If it came after 10 years of DCA, you may still be up on your initial investment

Re: averaging ... If investing in the stock market incurred a 10% chance the broker would just run off with your money, it wouldn't affect average returns very much ... It may affect your decision to lump sum, however ... Averaging blurs and distorts risk
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by Morik » Sun Nov 08, 2015 2:39 pm

tibbitts wrote:
Morik wrote:
quantAndHold wrote:It's a risk/reward thing. Given that "the stock market always rises over time," investing a lump sum as a lump has the potential for higher rewards. But it also riskier than DCAing the money in. The "right" answer depends on the investor's tolerance for risk. If I'd gotten a lump sum when I was 25, I would have just plopped it into the market. If the market dropped 25% the next day, I would have been annoyed, but mostly unperturbed. Now I'm two years from retirement and more risk averse. When I got a lump sum earlier this year, I decided to dollar cost average it into my regular AA over the course of a year. The lump sum was enough that it matters to my retirement, and a 25% drop right after I invested the whole lump would have changed my retirement plans.
I'm guessing your AA isn't 100% equities though--at 2 years from retirement you presumably have a reasonably high bond allocation.
So a 25% drop in equities would have dropped your portfolio by how much? (50/50 stock/bond, 12.5% drop?)

I still am trying to get an answer to this (as a general question to those who DCA): if you felt it was too risky to just drop the money in, why is it not still to risky once all the money is in?
With DCA you end up with less equity risk for a shorter time. As you gradually invest more into equities using DCA - and the simplest example is over an entire investing lifetime, which is how most people invest - you're probably also decreasing equities as a percentage in your allocation.

Looking back on my career, if I could have divided up my income equally over the years, and my corresponding equity (retirement) investments as well, I would have done that. There was definitely negative time value in terms of equity investments (but there was positive value at times with fixed income investments.) I wound up generally investing by far the most at the worst possible times for equities, which were the best times for my business.
There is a difference between investing over your entire career as you obtain money to invest, and having $X to invest and holding it in cash while slowly investing it.

Yes, the effect of buying regularly as you get money does have lower equity exposure. But if I had the option, I'd take all the earnings I'd make over my career right now, up front, and invest them in a lump sum.

I am talking just about the situation where you have $X (whether in cash or in investments), and are choosing how to allocate it.
From a psychological perspective, I understand loss aversion studies and how investing in a lump sum, followed by a loss, is more painful than a loss 10 years after you invested (because it feels like its less your fault/less in your control).

But from a mechanical perspective, taking $X and putting it into your allocation slowly is just starting at no risk (cash), and slowly moving up to whatever risk your chosen AA has.

Afterwards, you are constantly at the risk your AA has.

So if the argument is that DCA can make sense as a risk management strategy (above and beyond the psychological aspect), why is it ok to take the chosen AA risk after the DCA is complete, but not ok to take that amount of risk before the DCA is complete? Why do you only need to manage risk for the first 1, 5, 10 years (however long you DCA for)?

Mechanically, DCAing in over 10 years and now having $X+Y dollars invested (Y being gains), is no different from, 10 years from now, suddenly gaining $X+Y dollars. But in the former case no one (that I have seen) is saying a reasonable risk management strategy is to sell your entire investment (which you have just finished DCAing) and start DCAing again from scratch. But in the second case I do see this argument being made.

Mechanically they are no different--it is just a difference in the psychology attached to where/how the money came to belong to you.
(Exception: Capital gains taxes... so lets pretend there are no taxes. If there were no cap gains taxes, and you DCA into the market, would you also sell after you finish DCA and start all over again?)

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 2:44 pm

Maynard F. Speer wrote:
Morik wrote:
Maynard F. Speer wrote:
And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).
This is what Vanguard's study shows: on average you'd have done better lump-summing .. The market's spent more time going up than going down, so you're more likely to have benefited from bull markets and compounding than you'd have saved by avoiding lump-summing in bad years

But the problem with lump-summing is: you don't get the average ... The average completely ameliorates risk ... The reason you DCA or hold a defensive AA is to avoid bad timing, and minimise the damage of worst-case or black swan events

If someone came to me looking to invest $1million, I'd suggest a defensive AA, in line with the market's .. You'll always meet people who invested heavily right before a market crash, and as every crash is different, it's not as simple as saying "Stay the course", and guaranteeing a swift recovery - we often use Japan's 1990 market crash as a recent worst-case scenario - many people will have invested their retirement savings in 1990

Image
And my question was--why would you choose to DCA a large amount of money over time, but would not sell once you are fully invested and start DCAing again?

The same rationales apply--once you are fully invested you aren't getting the average, right?

Also, can you explain more about how getting the average price completely ameliorates risk?
(Can you clarify your meaning? I assume you don't mean it renders the investment risk-free... but complete amelioration implies that.)
Sequence of returns ... If you lump-sum then hit all your bad years, it may take decades to recover your losses .. a 50% loss takes a 100% return to break even .. an 80% loss requires a 400% return to break even

If your 50% loss came during the DCA years, you'd at least be buying an once-in-a-generation prices ... If it came after 10 years of DCA, you may still be up on your initial investment

Re: averaging ... If investing in the stock market incurred a 10% chance the broker would just run off with your money, it wouldn't affect average returns very much ... It may affect your decision to lump sum, however ... Averaging blurs and distorts risk
But in the scenario where you invest $X over the next 10 years, ending up with $X+$Y, vs receiving a check for $X+$Y 10 years from now...
That is mechanically identical; why DCA in one case, but not the other, aside from psychological considerations?

And if you would DCA in both cases, doesn't that mean you should sell your investment after the 10 years of DCAing and start DCAing again?

ks289
Posts: 642
Joined: Sun Mar 11, 2012 12:42 pm

Re: Dollar Cost Averaging inferior

Post by ks289 » Sun Nov 08, 2015 2:52 pm

tadamsmar wrote:
ks289 wrote:
tadamsmar wrote:DCA has two meanings. One meaning is just investing a % of each paycheck over time. That is a good policy that has nothing to do with investing a lump sum.

So, I would have to see the webinars you are are talking about.
For the purposes of comparing DCA and lump sum, investing a % of each paycheck really is not DCA since you are presumably not holding anything back that you would otherwise invest. While on the surface it resembles DCA, you are really lump sum investing amounts regularly as soon as the money is earned. The term some would use instead of DCA is automatic continuous investing.
I'm describing how the term is used. The op was not comparing anything. He was making unsorced claims. I asked for a source.
The link to Larry Swedroe's piece clarifies which meaning of the term is being used. I agree with you that vanguard's purported position on DCA is not established.
Last edited by ks289 on Sun Nov 08, 2015 2:54 pm, edited 1 time in total.

tibbitts
Posts: 9399
Joined: Tue Feb 27, 2007 6:50 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by tibbitts » Sun Nov 08, 2015 2:53 pm

Morik wrote:
tibbitts wrote:
Morik wrote:
quantAndHold wrote:It's a risk/reward thing. Given that "the stock market always rises over time," investing a lump sum as a lump has the potential for higher rewards. But it also riskier than DCAing the money in. The "right" answer depends on the investor's tolerance for risk. If I'd gotten a lump sum when I was 25, I would have just plopped it into the market. If the market dropped 25% the next day, I would have been annoyed, but mostly unperturbed. Now I'm two years from retirement and more risk averse. When I got a lump sum earlier this year, I decided to dollar cost average it into my regular AA over the course of a year. The lump sum was enough that it matters to my retirement, and a 25% drop right after I invested the whole lump would have changed my retirement plans.
I'm guessing your AA isn't 100% equities though--at 2 years from retirement you presumably have a reasonably high bond allocation.
So a 25% drop in equities would have dropped your portfolio by how much? (50/50 stock/bond, 12.5% drop?)

I still am trying to get an answer to this (as a general question to those who DCA): if you felt it was too risky to just drop the money in, why is it not still to risky once all the money is in?
With DCA you end up with less equity risk for a shorter time. As you gradually invest more into equities using DCA - and the simplest example is over an entire investing lifetime, which is how most people invest - you're probably also decreasing equities as a percentage in your allocation.

Looking back on my career, if I could have divided up my income equally over the years, and my corresponding equity (retirement) investments as well, I would have done that. There was definitely negative time value in terms of equity investments (but there was positive value at times with fixed income investments.) I wound up generally investing by far the most at the worst possible times for equities, which were the best times for my business.
There is a difference between investing over your entire career as you obtain money to invest, and having $X to invest and holding it in cash while slowly investing it.

Yes, the effect of buying regularly as you get money does have lower equity exposure. But if I had the option, I'd take all the earnings I'd make over my career right now, up front, and invest them in a lump sum.
What I'm saying is that my career income was, as you say you'd prefer, highly front-loaded and concentrated into certain years, and that didn't turn out to be beneficial for me. If you invest in only fixed income, then yes, provided you don't go long, you're always better off getting the money sooner rather than later. Fixed income compounds (and used to in the real sense, not just barely nominal as now); equities just do whatever they do.

Admittedly, "cash" was more valuable back when I was investing heavily in equities, and I'm comparing how I'd have been better off collecting some more of that real interest while gradually DCAing to what turned out to be cheaper markets over many years. Today it's hard to say; you may lose about the same no matter what you do.

Yes, when you're done DCAing that 40 years of salary you got paid on your first day of employment, the idea would be that if no more money were coming in, you'd effectively begin DCAing out to pay expenses or to (continue to) rebalance to a lower equity allocation, depending on circumstances, of course. So you'd never have that high percentage of equity exposure, for better or for (historically, no doubt) worse.

User avatar
tadamsmar
Posts: 8579
Joined: Mon May 07, 2007 12:33 pm

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by tadamsmar » Sun Nov 08, 2015 2:56 pm

DCA is superior if and only if taking money out of your current investments and DCAing back in is superior.

This was proved by George Constantinides back in 1979, using pure logic.

Is their a citation where Vanguard says DCA is superior? I don't know of one.

PS: If one gets a windfall, there might be some financial planning to do, immediately investing in your current AA is not always best.

User avatar
Maynard F. Speer
Posts: 2139
Joined: Wed Mar 18, 2015 10:31 am

Re: Dollar Cost Averaging inferior

Post by Maynard F. Speer » Sun Nov 08, 2015 3:00 pm

Morik wrote:
Maynard F. Speer wrote:
Morik wrote:
Maynard F. Speer wrote:
And if you DCAed over 10 years from 1989 to 1999, and then you are all in by the time of the 2000 crash, again, its poor results compared to if you just sell everything again as soon as you get all-in and start DCAing all over again.
But I don't see people arguing for doing so...

Why is doing this:
1) I have $X in cash, and I will hold it in cash and slowly move it into a portfolio with a certain AA

different from:
2) I have $X in a portfolio with a certain AA. I will sell it for cash. Go to 1).
This is what Vanguard's study shows: on average you'd have done better lump-summing .. The market's spent more time going up than going down, so you're more likely to have benefited from bull markets and compounding than you'd have saved by avoiding lump-summing in bad years

But the problem with lump-summing is: you don't get the average ... The average completely ameliorates risk ... The reason you DCA or hold a defensive AA is to avoid bad timing, and minimise the damage of worst-case or black swan events

If someone came to me looking to invest $1million, I'd suggest a defensive AA, in line with the market's .. You'll always meet people who invested heavily right before a market crash, and as every crash is different, it's not as simple as saying "Stay the course", and guaranteeing a swift recovery - we often use Japan's 1990 market crash as a recent worst-case scenario - many people will have invested their retirement savings in 1990

Image
And my question was--why would you choose to DCA a large amount of money over time, but would not sell once you are fully invested and start DCAing again?

The same rationales apply--once you are fully invested you aren't getting the average, right?

Also, can you explain more about how getting the average price completely ameliorates risk?
(Can you clarify your meaning? I assume you don't mean it renders the investment risk-free... but complete amelioration implies that.)
Sequence of returns ... If you lump-sum then hit all your bad years, it may take decades to recover your losses .. a 50% loss takes a 100% return to break even .. an 80% loss requires a 400% return to break even

If your 50% loss came during the DCA years, you'd at least be buying an once-in-a-generation prices ... If it came after 10 years of DCA, you may still be up on your initial investment

Re: averaging ... If investing in the stock market incurred a 10% chance the broker would just run off with your money, it wouldn't affect average returns very much ... It may affect your decision to lump sum, however ... Averaging blurs and distorts risk
But in the scenario where you invest $X over the next 10 years, ending up with $X+$Y, vs receiving a check for $X+$Y 10 years from now...
That is mechanically identical; why DCA in one case, but not the other, aside from psychological considerations?

And if you would DCA in both cases, doesn't that mean you should sell your investment after the 10 years of DCAing and start DCAing again?
Well I think you've engineered a situation in which someone sends you a check for 10 years of cost-averaged stock returns

If you consider your lump sum a big inheritance check, lump-summing gives you a wider distribution of absolute losses
"Economics is a method rather than a doctrine, an apparatus of the mind, a technique of thinking, which helps its possessor to draw correct conclusions." - John Maynard Keynes

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 3:06 pm

Maynard F. Speer wrote: Well I think you've engineered a situation in which someone sends you a check for 10 years of cost-averaged stock returns

If you consider your lump sum a big inheritance check, lump-summing gives you a wider distribution of absolute losses
But so does selling all your investments (after you are done DCAing) and starting again.

Should anyone with current investments just sell their positions into cash and DCA right now? If not, why not?

User avatar
JoMoney
Posts: 8130
Joined: Tue Jul 23, 2013 5:31 am

Re: Dollar Cost Averaging inferior

Post by JoMoney » Sun Nov 08, 2015 3:09 pm

Morik wrote:
Maynard F. Speer wrote: Well I think you've engineered a situation in which someone sends you a check for 10 years of cost-averaged stock returns

If you consider your lump sum a big inheritance check, lump-summing gives you a wider distribution of absolute losses
But so does selling all your investments (after you are done DCAing) and starting again.

Should anyone with current investments just sell their positions into cash and DCA right now? If not, why not?
Some people are effectively doing that by rebalancing between stocks and bonds.
The argument against DCA is silly. It's like arguing against rebalancing as being inferior to buy and hold.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

Morik
Posts: 731
Joined: Tue Nov 25, 2014 12:26 pm

Re: Dollar Cost Averaging inferior

Post by Morik » Sun Nov 08, 2015 3:15 pm

JoMoney wrote:
Morik wrote:
Maynard F. Speer wrote: Well I think you've engineered a situation in which someone sends you a check for 10 years of cost-averaged stock returns

If you consider your lump sum a big inheritance check, lump-summing gives you a wider distribution of absolute losses
But so does selling all your investments (after you are done DCAing) and starting again.

Should anyone with current investments just sell their positions into cash and DCA right now? If not, why not?
Some people are effectively doing that by rebalancing between stocks and bonds.
How so? Can you give me an example?

If you DCA into a 70/30 allocation over 10 years, then what... move to 60/40? 50/50?
That is very different from selling it all and have 90% cash, 7% stock, 3% bonds...

User avatar
ruralavalon
Posts: 17065
Joined: Sat Feb 02, 2008 10:29 am
Location: Illinois

Re: Dollar Cost Averaging inferior [but Vanguard disagrees]

Post by ruralavalon » Sun Nov 08, 2015 3:21 pm

dbr wrote:
ks289 wrote: Is regular paycheck investing good? Sure. It is better compared to what?
The origin of the term DCA, and the only actually meaningful definition, is the distinction between a practice of buying a fixed number of shares every period and a practice of investing a fixed dollar amount every period. If you invest a fixed dollar amount you buy more shares when stock prices are down and fewer when stock prices are up. This results in buying stock at a lower average price than if one buys the same fixed number of shares each period. The result is obviously desirable and probably more convenient when one receives money in constant amounts at fixed intervals. The issue only even arises in an older time when buying shares in round lots of fixed size was standard. Today a person investing in mutual funds would hardly even consider investing by buying 100 shares exactly every month; your just put in $429.67 (or whatever) every paycheck.

It is this lingering misunderstanding that is behind some of the mythology going on here.
Yes, I believe that buying a fixed dollar amount every pay period is the original meaning of the term "dollar cost averaging (DCA)".

It created a lot of confusion when someone started applying the term to investing a large lump sum in stages, an entirely different problem.
"Everything should be as simple as it is, but not simpler." - Albert Einstein | Wiki article link:Getting Started

User avatar
JoMoney
Posts: 8130
Joined: Tue Jul 23, 2013 5:31 am

Re: Dollar Cost Averaging inferior

Post by JoMoney » Sun Nov 08, 2015 3:23 pm

Morik wrote:...
If you DCA into a 70/30 allocation over 10 years, then what... move to 60/40? 50/50?
That is very different from selling it all and have 90% cash, 7% stock, 3% bonds...
That's a ridiculous scenario that nobody actually suggests except as a straw-man argument to show some superiority of 'lump sum' investing, while ignoring the risk of lumping in which doesn't guarantee you'll get a better price and exposes you to the chance you could be at a peak or a high that's far from the 'average'.

The example doesn't follow for someone who 'averages' anyway. Why would this straw-man suddenly turn to lumping out instead slowly averaging out and back in like rebalancers do?
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

Post Reply