Inverse DCA vs. Lump Sum

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Inverse DCA vs. Lump Sum

Post by VictoriaF »

A common question is:
If I have some money to invest, what is better: to invest it as a lump sum (LS) or to do Dollar Cost Averaging (DCA)?
- The response is usually "LS is better on the long run, but DCA helps with psychological quirks."

When the goal is inverse of the above, is the logic opposite? In other words,
If I want to de-allocate a non-trivial sum from equities to fixed-income, what is better: to move it as LS or to do DCA?
- Is DCA preferable in the "inverse" case for both performance and psychological reasons?
- Should Inverse DCA be performed on Really Good Days (RGD)?

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
User avatar
mhc
Posts: 5257
Joined: Mon Apr 04, 2011 10:18 pm
Location: NoCo

Re: Inverse DCA vs. Lump Sum

Post by mhc »

I think it depends on why you are changing your AA.

1. AA is much too aggressive, then LS.
2. Want to change AA due to glide slope, then DCA.

Maybe there are some other scenarios.
52% TSM, 23% TISM, 24.5% TBM, 0.5% cash
michaelsieg
Posts: 625
Joined: Mon Jan 07, 2013 10:02 am

Re: Inverse DCA vs. Lump Sum

Post by michaelsieg »

Most arguments about DCA versus LS are based on the observation that the stock market statistically over the long-term is more likely to go up than down, hence a LS statistically gives you an advantage, as all you capital "works" for you right away.
If you use this logic to your "inverse DCA question" then you would benefit longer from the stock market if you do inverse DCA a opposed to LS and probably have a higher chance to end up ahead.

But the real question is why you need to rebalance your portfolio. If you are uncomfortable with the risk of your current AA or if you have hit rebalancing bands, then I would do LS.
Obviously no one knows where markets go, but if you feel that they might be valued fairly high, that could be an other argument for LS.
User avatar
Chan_va
Posts: 869
Joined: Wed Dec 05, 2012 6:15 pm

Re: Inverse DCA vs. Lump Sum

Post by Chan_va »

I just want to point out that there are hybrid options as well.

1. Lump sum out of equities into a "short term" account, then DCA into fixed income
2. DCA out of equities into a "short term" account, then LS into fixed income

Not saying any of these is better/worse than any other option, but in true boglehead fashion, here is a "cover all bases" option :)

1. LS 25% from equities into interm grade FI
2. DCA 25% from equities into interm grade FI
3. LS 25% from equities to MM/Short term FI, then DCA into interm grade FI
4. DCA 25% from equities into MM, then LS into interm grade FI
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

mhc wrote:I think it depends on why you are changing your AA.

1. AA is much too aggressive, then LS.
2. Want to change AA due to glide slope, then DCA.

Maybe there are some other scenarios.
I am reallocating the funds for three reasons:
1. My AA is about 4.5% off.
2. The equity markets are rather high.
3. I'd like to have some cash available in case (a) equity markets drop, or (b) bond interests rise, or (c) TIPS start paying higher fixed rate.

#1 is not a big deal, it could wait.
#3 opportunities may be months (years) off.
#2 is the real reason why I started doing "inverse DCA." Now, I am interested if it is historically/theoretically prudent.

Thank you,

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
User avatar
mhc
Posts: 5257
Joined: Mon Apr 04, 2011 10:18 pm
Location: NoCo

Re: Inverse DCA vs. Lump Sum

Post by mhc »

#1 seems like a good reason to just LS
#2 and #3 seem like market timing that may or my not pay off. In that case, there is no telling if DCA or LS would be better.

If you are only talking about 4.5% of your portfolio, I don't think it will matter much what you do.
52% TSM, 23% TISM, 24.5% TBM, 0.5% cash
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

mhc wrote:#1 seems like a good reason to just LS
#2 and #3 seem like market timing that may or my not pay off. In that case, there is no telling if DCA or LS would be better.

If you are only talking about 4.5% of your portfolio, I don't think it will matter much what you do.
Exactly. Given that 4.5% is not going to make much difference one way or the other, I am using it as an excuse to engage in some market timing. But I am also wondering if I could come up with a quasi-theoretical concept, such as "Inverse DCA," to make my activities appear more respectable.

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Inverse DCA vs. Lump Sum

Post by Aptenodytes »

If you step back and think for a moment, the answer ought to be obvious. If you know where you need to be, you should of course get there as fast as your emotions will tolerate. Any other answer is beyond the range of logical discourse. The answer follows from the premise of knowing where you need to be.
michaelsieg
Posts: 625
Joined: Mon Jan 07, 2013 10:02 am

Re: Inverse DCA vs. Lump Sum

Post by michaelsieg »

If you have an appetite for some risk on the bond side - the emerging government bond fund just opened, it might be worth considering it with a small fraction of your bond allocation...certainly much higher risk than high quality bonds, but at the same time not directly correlated to the equity market.
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

Thank you, Michael, I will check it out,

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
User avatar
Random Musings
Posts: 6770
Joined: Thu Feb 22, 2007 3:24 pm
Location: Pennsylvania

Re: Inverse DCA vs. Lump Sum

Post by Random Musings »

If you AA is off 4.5% and this is a rebalancing exercise, rebalance.

Essentially, on this rebalance you are taking equites to target and reducing the duration of your bond portfolio. IMHO, the only "gamble" you are taking is the reduction in duration.

I guess the real question is what your written investment plan says about deploying that 4.5% out of cash.

With respect to your 2) comment, there are studies (and Larry S. has talked about them as well) about expected returns based on Shiller PE 10. As you probably already know, the range of returns as a function of PE 10 is rather large, but the avg expected return based on current Shiller PE 10 is rather muted. Same with bonds. Since you are taking it out of equity, I would lump.

RM
I figure the odds be fifty-fifty I just might have something to say. FZ
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

Random Musings wrote:If you AA is off 4.5% and this is a rebalancing exercise, rebalance.

Essentially, on this rebalance you are taking equites to target and reducing the duration of your bond portfolio. IMHO, the only "gamble" you are taking is the reduction in duration.

I guess the real question is what your written investment plan says about deploying that 4.5% out of cash.
Thank you, Random Musings,

My current IPS is silent on the bond duration and on the deployment of the 4.5% out of cash. What I wrote above in #2 (opportunistic dry powder) is as far as I have thought about it.
Random Musings wrote:With respect to your 2) comment, there are studies (and Larry S. has talked about them as well) about expected returns based on Shiller PE 10. As you probably already know, the range of returns as a function of PE 10 is rather large, but the avg expected return based on current Shiller PE 10 is rather muted. Same with bonds. Since you are taking it out of equity, I would lump.

RM
Thank you for an additional perspective.

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Inverse DCA vs. Lump Sum

Post by YDNAL »

VictoriaF wrote:A common question is:
If I have some money to invest, what is better: to invest it as a lump sum (LS) or to do Dollar Cost Averaging (DCA)?
- The response is usually "LS is better on the long run, but DCA helps with psychological quirks."

When the goal is inverse of the above, is the logic opposite? In other words,
If I want to de-allocate a non-trivial sum from equities to fixed-income, what is better: to move it as LS or to do DCA?
- Is DCA preferable in the "inverse" case for both performance and psychological reasons?
- Should Inverse DCA be performed on Really Good Days (RGD)?

Victoria
Unless we try to make too much of this scenario (not unusual in Bogleland :)), IF the Stock Market is expected* to go up over time, then DCA Equities to Fixed Income ("de-allocate") to achieve better expected* results. Otherwise, my suggestion is to match need for money with Fixed Income duration.

* expected <> guaranteed
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
User avatar
iceport
Posts: 6054
Joined: Sat Apr 07, 2007 4:29 pm

Re: Inverse DCA vs. Lump Sum

Post by iceport »

That's an interesting question. If I interpret "wbern" correctly, valuations and volatility enter into the decision to employ DCA or not when purchasing into an allocation, so I would think they should also be considered when selling out of an allocation. When buying, high valuations and high volatility both increase the likelihood that DCA will be advantageous. However, when selling I wonder if high valuations would indicate a preference for lump selling.

Also, I have wondered this: If the decision is made to diversify sales prices through a periodic sale of assets over time, would a "constant share" approach be more advantageous than a "constant dollar" approach? After all, wouldn't you be better served accepting less proceeds when prices are depressed with a constant share approach, rather than selling more shares when prices are depressed with a constant dollar approach?

--Pete
"Discipline matters more than allocation.” |—| "In finance, if you’re certain of anything, you’re out of your mind." ─William Bernstein
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Inverse DCA vs. Lump Sum

Post by Aptenodytes »

petrico wrote:That's an interesting question. If I interpret "wbern" correctly, valuations and volatility enter into the decision to employ DCA or not when purchasing into an allocation, so I would think they should also be considered when selling out of an allocation. When buying, high valuations and high volatility both increase the likelihood that DCA will be advantageous. However, when selling I wonder if high valuations would indicate a preference for lump selling.

Also, I have wondered this: If the decision is made to diversify sales prices through a periodic sale of assets over time, would a "constant share" approach be more advantageous than a "constant dollar" approach? After all, wouldn't you be better served accepting less proceeds when prices are depressed with a constant share approach, rather than selling more shares when prices are depressed with a constant dollar approach?

--Pete
Hey, maybe DCA comes out on top in both directions and we can all increase our returns by continually moving money in and out of the market.

Seriously, if you have determined the best AA then that's the best position to be; this is true by definition. Sometimes emotions must be accommodated, so you adjust to let yourself sleep at night. That's an adjustment based on intangibles. But please don't think you are going to somehow improve your situation in a tangible manner by taking longer to get where you want to be. When you cross a busy street because you want to get to the other side, do you ever deliberately walk slowly because you think that if you take longer to cross you'll improve your outcomes?
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

Aptenodytes wrote:When you cross a busy street because you want to get to the other side, do you ever deliberately walk slowly because you think that if you take longer to cross you'll improve your outcomes?
Crossing a busy street slowly never improves outcomes. A better analogy is choice between taking an expressway vs. a scenic route. A scenic route may be preferable for esthetic reasons. In case of a major problem on the expressway, the scenic route may also be faster. Inverse DCA out of equities is a scenic route.

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Inverse DCA vs. Lump Sum

Post by Aptenodytes »

VictoriaF wrote:
Aptenodytes wrote:When you cross a busy street because you want to get to the other side, do you ever deliberately walk slowly because you think that if you take longer to cross you'll improve your outcomes?
Crossing a busy street slowly never improves outcomes. A better analogy is choice between taking an expressway vs. a scenic route. A scenic route may be preferable for esthetic reasons. In case of a major problem on the expressway, the scenic route may also be faster. Inverse DCA out of equities is a scenic route.

Victoria
I'll buy the esthetic analogy. That corresponds to the emotional element I mentioned. It is an intangible benefit of going slow, not a tangible one. There's never any reason to expect a tangible benefit unless you relax the assumption that you know where you want to end up. If I am about to head out on a long drive and I have zero information about traffic conditions and the most important thing is getting to my destination on time, I take the freeway every time. I'd never say, based on zero information, "hey maybe I'll take the slower road because it just might turn out better."
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

Aptenodytes wrote:
VictoriaF wrote:
Aptenodytes wrote:When you cross a busy street because you want to get to the other side, do you ever deliberately walk slowly because you think that if you take longer to cross you'll improve your outcomes?
Crossing a busy street slowly never improves outcomes. A better analogy is choice between taking an expressway vs. a scenic route. A scenic route may be preferable for esthetic reasons. In case of a major problem on the expressway, the scenic route may also be faster. Inverse DCA out of equities is a scenic route.

Victoria
I'll buy the esthetic analogy. That corresponds to the emotional element I mentioned. It is an intangible benefit of going slow, not a tangible one. There's never any reason to expect a tangible benefit unless you relax the assumption that you know where you want to end up. If I am about to head out on a long drive and I have zero information about traffic conditions and the most important thing is getting to my destination on time, I take the freeway every time. I'd never say, based on zero information, "hey maybe I'll take the slower road because it just might turn out better."
I agree. Scenic route it is. Thanks for working out an appropriate analogy,

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
Scooter57
Posts: 2019
Joined: Thu Jan 24, 2013 8:20 am

Re: Inverse DCA vs. Lump Sum

Post by Scooter57 »

Another approach is what William Bernstein suggests for retirement withdrawals in his Four Pillars of Investing book. He suggests using reverse Value Averaging. That way you take out more when your fund is high and less as it drops (and adjust your spending accordingly). That's one way to make sure you don't run out of savings. You'd figure out how much you would want to have left in your fund at the end of some time period, and then create a value path that takes you there. Each month you'd take out the amount that left your fund holding whatever the value path said it should. At the end of the time period you'd have drawn down to your stated amount or stopped taking money out if the fund had really tanked.
User avatar
iceport
Posts: 6054
Joined: Sat Apr 07, 2007 4:29 pm

Re: Inverse DCA vs. Lump Sum

Post by iceport »

Aptenodytes wrote:
petrico wrote:That's an interesting question. If I interpret "wbern" correctly, valuations and volatility enter into the decision to employ DCA or not when purchasing into an allocation, so I would think they should also be considered when selling out of an allocation. When buying, high valuations and high volatility both increase the likelihood that DCA will be advantageous. However, when selling I wonder if high valuations would indicate a preference for lump selling.

Also, I have wondered this: If the decision is made to diversify sales prices through a periodic sale of assets over time, would a "constant share" approach be more advantageous than a "constant dollar" approach? After all, wouldn't you be better served accepting less proceeds when prices are depressed with a constant share approach, rather than selling more shares when prices are depressed with a constant dollar approach?

--Pete
Hey, maybe DCA comes out on top in both directions and we can all increase our returns by continually moving money in and out of the market.

Seriously, if you have determined the best AA then that's the best position to be; this is true by definition. Sometimes emotions must be accommodated, so you adjust to let yourself sleep at night. That's an adjustment based on intangibles. But please don't think you are going to somehow improve your situation in a tangible manner by taking longer to get where you want to be. When you cross a busy street because you want to get to the other side, do you ever deliberately walk slowly because you think that if you take longer to cross you'll improve your outcomes?
If you feel comforted by ignoring the fact that the timing (i.e. price point) of your entry into or out of the market can have very tangible consequences, far be it for me to judge.

--Pete
"Discipline matters more than allocation.” |—| "In finance, if you’re certain of anything, you’re out of your mind." ─William Bernstein
sashwin
Posts: 60
Joined: Wed Feb 23, 2011 11:07 pm

Re: Inverse DCA vs. Lump Sum

Post by sashwin »

petrico wrote:If you feel comforted by ignoring the fact that the timing (i.e. price point) of your entry into or out of the market can have very tangible consequences, far be it for me to judge.

--Pete
Pete I remember our debates from a while back on DCA vs LS, particularly your roulette wheel analogy that eliminated time from the equation. As is often the case I don't think either of us budged from our positions :D. If I may summarize, your position can be stated as "DCA has lower expected rewards, but also lower risk", ie SD is lower. My position is "Sure, any AA with higher cash will have these characteristics, if that's the AA you want, just move to it". I understand your perspective on this also, even though I may not agree.

Would you also change your strategy based on the purpose of the divestment ? Ie, if the goal of selling is to change AA into higher cash (but leave the cash in your portfolio, thus it is still rebalanced) would you do something different, vs if you are pulling the money out permanently ? Just curious to hear your reasoning.

-- Ashwin
User avatar
iceport
Posts: 6054
Joined: Sat Apr 07, 2007 4:29 pm

Re: Inverse DCA vs. Lump Sum

Post by iceport »

sashwin wrote:
petrico wrote:If you feel comforted by ignoring the fact that the timing (i.e. price point) of your entry into or out of the market can have very tangible consequences, far be it for me to judge.

--Pete
Pete I remember our debates from a while back on DCA vs LS, particularly your roulette wheel analogy that eliminated time from the equation. As is often the case I don't think either of us budged from our positions :D. If I may summarize, your position can be stated as "DCA has lower expected rewards, but also lower risk", ie SD is lower. My position is "Sure, any AA with higher cash will have these characteristics, if that's the AA you want, just move to it". I understand your perspective on this also, even though I may not agree.

Would you also change your strategy based on the purpose of the divestment ? Ie, if the goal of selling is to change AA into higher cash (but leave the cash in your portfolio, thus it is still rebalanced) would you do something different, vs if you are pulling the money out permanently ? Just curious to hear your reasoning.

-- Ashwin
Hi Ashwin,

I really don't have a strategy for the withdrawal phase yet. In direct answer to your question, though, the relative magnitude of the sales would matter, not the purpose. In a move of a few percent, the risks of unlucky timing of a lump sale are limited. Need to liquidate 30% of the whole? Whoa! Let's think this through!

Actually, when it comes right down to it, my best guess is that in the withdrawal stage there will be an element of market timing! Here's the broad outline: Maintain a 70/30 equity/fixed income allocation permanently, but by retirement have about 20% of the total portfolio separated from that risky allocation and placed into riskless assets. (That works out to an overall AA of 56/44 equity/fixed income.) So how to replenish the "cash" bucket? First would be any dividends. But with roughly 5 years' worth of withdrawals in cash-like assets, I'd certainly not do any selling in a down market -- thus the market timing element. (The premise is that essential living expenses will be just about covered with a pension.) Beyond that, I haven't thought it out. But I'll reiterate, if I do ever choose to use periodic selling, I'd probably use a constant share approach rather than a constant dollar approach.

Just to clarify, I'm not an advocate of DCA, nor am I an opponent of lump-summing. I'm simply a defender of DCA as a perfectly rational way to address a certain form of risk. And I really do understand the valid arguments in favor of lump summing. That being said, the notion that there is never a tangible difference between the two methods (noted in a post above) isn't one of them, because it just isn't true.

--Pete
"Discipline matters more than allocation.” |—| "In finance, if you’re certain of anything, you’re out of your mind." ─William Bernstein
umfundi
Posts: 3361
Joined: Tue Jun 07, 2011 5:26 pm

Re: Inverse DCA vs. Lump Sum

Post by umfundi »

Victoria,

You should resist the urge to tinker and tamper with your portfolio and IPS.

Let me suggest Sudoku. The mathematics is interesting, the puzzles are fun, and the complexity is up to you. Or, genealogy. A potentially infinite jigsaw puzzle. I indulge in both.

The idea that you should DCA out of an investment is a beautiful counterexample to the argument that you should DCA in.

Keith
Déjà Vu is not a prediction
User avatar
Topic Author
VictoriaF
Posts: 20122
Joined: Tue Feb 27, 2007 6:27 am
Location: Black Swan Lake

Re: Inverse DCA vs. Lump Sum

Post by VictoriaF »

umfundi wrote:Victoria,

You should resist the urge to tinker and tamper with your portfolio and IPS.

Let me suggest Sudoku. The mathematics is interesting, the puzzles are fun, and the complexity is up to you. Or, genealogy. A potentially infinite jigsaw puzzle. I indulge in both.

The idea that you should DCA out of an investment is a beautiful counterexample to the argument that you should DCA in.

Keith
Keith,

Sudoku is lame. When you have a chance try the SET game.

Victoria
Inventor of the Bogleheads Secret Handshake | Winner of the 2015 Boglehead Contest. | Every joke has a bit of a joke. ... The rest is the truth. (Marat F)
Post Reply