23

I am investing in mutual funds and hence reading and trying to learn about it.

I was reading this answer which says lump sum investment is far more efficient than DCA. Well, that does not surprise me. I am aware that lump-sum will outperform DCA.

There is a link to PDF in that answer which I downloaded. Title of PDF is "Dollar-cost averaging just means taking risk later". It is the risk part that I do not understand.

My understanding is that, lump sum will be far more efficient but it will incur far more risk as well. On other hand, DCA will be less efficient but it will minimize the risk considerably.

The PDF mentioned above contradicts my understanding. It says that DCA also incur same risk; it is just delayed.

I am sure I am misunderstanding something here. Many finance-guru suggest to invest in mutual funds with equal monthly contributions. Many answers on this site also recommend it.

Can someone please explain what I am misunderstanding?

  • 2
    You may wish to bear in mind that there are potentially different audiences and goals for these sources. Source #1 (lump-sum answers) may be talking to people who are 100% going to invest, the question is how to optimize that investment. Source #2 (gurus) may be talking to people who are less likely to invest, so the monthly contributions automatic setup and such is desirable, since it induces people to actually invest. NAA, since I'm just taking the sources at face value. – Jeutnarg Aug 27 '18 at 18:42
  • With DCA you are basically hoping that when you eventually sell your sell price will be more than your average buy price. You should never buy into a falling market. – Victor Aug 27 '18 at 22:52
11

The answer you link, and the Vanguard paper it references, implicitly define dollar-cost averaging (DCA) as an execution strategy where you acquire a large amount of cash all at once, and then decide to not invest all of it right away. They contrast this with lump-sum investing, where all the cash is invested immediately when you receive it.

This is not a typical situation. A lot of people are working and get paid periodically. Some of that money goes to living expenses. The rest gets invested. Investing some of each paycheck is not DCA by this definition.

Likewise, dollar-cost averaging a sale of an asset is starting to sell off the asset in advance of when the cash is required.

The more typical situation is selling some assets periodically to cover living expenses in retirement. Again, this is not DCA.

In other words, if the strategy does not involve holding a large amount of cash for a while, it's not DCA.

The typical advice "to invest in mutual funds with equal monthly contributions" isn't about DCA. It advocates a steady, sustained, methodical approach to retirement saving. Specifically, do not:

  • Blow all your income on luxuries and invest none of it
  • Accumulate cash in a mattress until you decide to go all-in on some silly money-making scheme
  • Try to time the market

This approach (investing some of each paycheck) has the advantage of DCA: the risk of purchasing at a bad time on any particular day is minimized because your portfolio is the sum of hundreds or thousands of purchases, and no individual purchase has much impact overall. But also it doesn't have the disadvantage of DCA: a large accumulation of cash which is not earning returns. This is the reasoning for the common advice of periodic investments.

  • 1
    This seems to be answering a different question than the one asked. – Barmar Aug 27 '18 at 21:03
  • Although the question did end with the finance gurus advocating regular monthly investing. I guess this explains why they do that -- regular, automated investing avoids trying to "time" the market. – Barmar Aug 27 '18 at 21:17
  • 1
    @Barmar That's true -- I'm attempting to address what might be a misconception behind the question. – Phil Frost Aug 27 '18 at 23:21
  • 2
    Maybe it's just a personal difference but I disagree with the initial premise that "investing some of each paycheck is not DCA". Sure it is. By definition if you invest a consistent amount of money on a regular schedule, you are averaging your purchase price. I have never heard of it having anything at all to do with how much total cash the investor has on hand. Or even if it's cash at all. – Jonathan van Clute Aug 28 '18 at 18:49
  • 3
    @JonathanvanClute Context is important here. Both the answer linked in the question and the Vanguard paper implicated begin with the premise of a large lump sum, and contrast the possibilities of investing it all now, versus over time in equal dollar amounts. – Phil Frost Aug 28 '18 at 18:57
32

The paper you are referring to is likely the Vanguard report Dollar-cost averaging just means taking risk later.

The thrust of the paper is that lump sum investing outperforms dollar cost averaging the majority of the time (which makes sense; if the market, on average, goes up, then on average, it will go up after you invest, so investing earlier captures more gain).

However, you are right that dollar cost averaging distributes risk. This is even mentioned in the paper, on page 2:

But if the investor is primarily concerned with minimizing downside risk and potential feelings of regret (resulting from lump-sum investing immediately before a market downturn), then DCA may be of use.

And later on page 5:

Even though LSI’s average outperformance and risk-adjusted returns have been greater than those of DCA, risk-averse investors may be less concerned about averages than they are about worst-case scenarios, as well as the potential feelings of regret that would occur if a lump-sum investment were made immediately prior to a market decline. These concerns are not unreasonable. We found that DCA performed better during market downturns, so DCA may be a logical alternative for investors who prefer some short-term downside protection.

The "taking risk later" is really in reference to an investor who does not fully invest in the market now, for fear of a market downturn, not fully investing in the market in the future due to the same fear.

  • Thanks for your efforts; +1. I will take another look into the paper. So, do you mean that monthly equal contribution in mutual fund (DCA) is preferred option over LSI if I am MORE concerned about short term downs than (hopefully) far better returns? – Aastik Aug 27 '18 at 18:05
  • Yes. If you are more concerned with minimizing risk than maximizing gain, dollar cost averaging is what you would want to do. – Magua Aug 27 '18 at 19:12
  • 11
    As a quick followup, the LSI and DCA discussed in the paper are in regards to people who currently have the money and are looking to invest it. If you decide to invest part of your paycheck every month, that is not DCA -- you are not holding back money earmarked for investment, you are investing it as you receive it. – Magua Aug 27 '18 at 19:15
  • @Magua What if I don't want to invest the entire portion earmarked for investment every month, but rather want to spread it out across a longer period of time? (haha!) – Michael Aug 27 '18 at 19:21
  • 6
    Then that would become DCA. But it would also be a situation where you have a constantly growing pile of uninvested money (since you would be investing less than you are earmarking to invest every month). – Magua Aug 27 '18 at 19:30
9

I agree with the other answers but I think the confusion comes from the phrase "lump sum" and can be explained much more simply:

In a market that is on average increasing, you want to get your money into the market as soon as possible.

Based on this, if you plan to put $1200 into the market in a particular year, if the market is on average increasing, then some common options are:

  1. $1200 lump sum at the beginning of the year: on average this is the best choice (of these 3).
  2. Dollar-cost-averaging with $100 per month for the entire year: on average this is the medium choice (of these 3).
  3. $1200 lump sum at the end of the year: on average this is the worst choice (of these 3).

Notice the "lump sum" is both the best and worst choice depending on the timing of it.

Regarding risk, it's not that DCA by itself is "delaying" risk; notice the lump sum in option 3 would technically be delaying risk too. Instead, with option 2 (DCA) you are simply taking less risk each month rather than all at once which is very similar to, but subtly different from "delaying" risk. IMHO the title of that article isn't the best choice of words because it's not just a delay, it's also lessening the risk.

  • I think the implicit assumption is that only options 1 and 2 are being compared, i.e. you've decided to start investing now, how should you do it? – Barmar Aug 27 '18 at 21:05
  • @Barmar - Agreed. Seems like the confusion OP is having is due to two things: the term "lump sum", and also because the risk isn't the "same". Spreading it out "lessens the risk". – TTT Aug 27 '18 at 21:14
  • @TTT – 'Spreading it out "lessens the risk"' – only in short term fluctuations in your balance, but generally gives you a worse return in the medium term. If you're sitting on a lump sum in your savings account. – Nigel Touch Aug 28 '18 at 11:28
  • @NigelTouch right. That's why spreading it out is #2 in my list instead of #1. – TTT Aug 28 '18 at 12:06
2

Typically your investments can be separated in to two buckets:

  1. Assets that are allocated already (cash is an asset allocation)

  2. The allocation of assets I receive on a going forward basis

The issue with dollar cost averaging and the many other names for the same concept is that it addresses part 1 not part 2. Dollar cost averaging is about reallocation of part 1; for example reallocating some cash to a mutual fund.

Dollar cost averaging is about delaying the allocation of money you posses now. You have $1,000 right now, you've decided that all of it will be put in XYZ mutual fund, do you buy $1,000 of that mutual fund today or do you buy $100 increments for 10 months? Many people describe taking $100 from each paycheck to buy $100 of XYZ mutual fund every month as dollar cost averaging. Allocating part of your pay check as soon as you receive it is not dollar cost averaging, really, that's investing as soon as the funds become available to you.

Dollar cost averaging is about delaying risk, because dollar cost averaging is about timed relocation of assets you possess currently. As an example, you're reallocating $1,000 of your cash position to a mutual fund position $100 at a time for 10 months. The Vanguard paper researches investment results of either investing an amount right now, or parsing it out in to a predetermined investment strategy over a variety of increments and time periods.

According to Vaguard's research, if you have $1,000 right now, and you've decided that you want to allocate $1,000 to the stock market, you should just do it now, not spread it out over time. Dollar cost averaging can win in a down market, the price is falling you buy for an ever reducing price. The issue is that over time the market trends upward so dollar cost averaging is more likely to result in an increased average share price.

  • If I am understanding this correctly, equal monthly contributions to mutual fund is not different than LSI. Because, the amount I am investing each month (immediately) is available to me only through monthly paycheck. Am I correct now? – Aastik Aug 27 '18 at 18:24
  • 1
    @Aastik, yes. As seemingly counter intuitive as it sounds, if you receive $100 every month and you've decided that this $100 will be allocated to the market then you would be lump sum investing $100 every month. The issue is the really technical definition of dollar cost averaging. Ending up with an average unit price over time is not dollar cost averaging, dollar cost averaging is an allocation strategy. – quid Aug 27 '18 at 18:28
  • 1
    @RonJohn - what do you mean by a trivial example? Wouldn't the ROI be the same either way? The answer would be the same using $1 instead of $1000. Or are you just pointing out how DCA differs with a crystal ball? :D – TTT Aug 27 '18 at 19:53
  • 1
    @RonJohn - ok, i see what you mean, though even then on average you'd still be better off with lump sum. You pin pointed one of the few times you wouldn't be, but even if you did it at that unlucky moment, you wouldn't be much worse off because of it. – TTT Aug 27 '18 at 20:05
  • 2
    @RonJohn - well if that was the case, I think investing your emergency fund in stocks is the much bigger mistake than the fact that you did a lump sum. :D – TTT Aug 27 '18 at 20:18
2

Consider the following four options:

(A) Invest $2000 now, hold for 12 months.

(B) Invest $1000 now, another $1000 one month from now, and then hold 11 more months.

(C) Invest $2000 a month from now, hold for 11 more months.

(D) Don't invest anything.

Option (B) is a simple DCA strategy. Note that it's basically a mix of (A) and (C). A pure (C) strategy would mean that you're avoiding the risk of the stock going down this month. I would guess that that's what is meant by "taking risk later": you're not taking risk now, but you are in a month. Although, really, what's happening is that you're not ever taking this month's risk, and you're taking the rest of the risk at the same time as you would with Option (A). Option (B), as a mix of (A) and (C), is avoiding some of the risk of this month, so in the same sense that (C) is taking risk later, (B) is taking some risk later. Of course, if you don't want to take this month's risk, you should take Option (B), and if you want to completely minimize your risk, you should take Option (D). Now, one argument for DCA is that it's "diversifying" by taking a mix of (A) and (C). However, that misunderstands the nature of diversification. When you diversify across stocks, you decrease your risk. If one stock has a lower return than the other, then buying both will result in lower return than buying just the one with higher return, but there is always some mix of the two stocks that results in the risk decreasing more than the return decreases. But with DCA, you are decreasing your risk and return in tandem. Either you think the return on the stock is worth the risk, in which case you should go with Option (A), or you don't, in which case you should go with Option (D). If you think that the return isn't currently worth the risk, but it will be in a month, you should go with Option (C). There's no consistent set of preferences for which Option (B) is the best choice.

2

For people talking about investing $100 (or whatever) out of each check being LSI in relation to that one specific payment, how do you rationalize people electing not to do something like maintaining a large emergency fund and then going 100% of pay into 401k in month 1 and then 0% of pay into 401k in months 2 - 12 and then you refill month 1's drain on the EF in the other 11 months?

For the sake of argument, this is a whole lot more like LSI than what people are normally doing by putting money in every month.

It also leaves intact the meaning "put money in every month rather than all in one shot" that most people associate with DCA.

I have actually heard of people tossing around the above idea, strangely enough.

I also want to point out that the Vanguard paper has the beginning and ending portfolio values at a difference of about 2.3%, favoring the LSI side. I think that it's worth noting that this is the average gain after 10y for doing the riskiest possible play.

It may not be prudent to do a much riskier strategy if the average payoff is only 2.3% ahead of a plan that's got a lot more downside protection and a lot more upside potential as well.

Not that I am advocating being a market timer or anything, but if I had 120k and I was doing DCA 10k/m in order to reduce my downside risk and in month 3 or something the market went down by 33%, I could easily pivot back to LSI at that point and have a far superior return as compared with LSI when we are 10y into the future. It's no big deal if it then dips temporarily farther after you go in, as long as you are still in for the long term.

It's absolutely dumb to sell in down markets, but it's in no respect dumb to buy more in down markets, as long as you want to leave that money in for 10y+ no matter what afterwards. There's basically no scenario where buying as much stock as possible in significantly down markets and then holding for 10y+ fails to work out well.

I think that the people at Vanguard should probably consider a DCA pivoting to LSI if the market goes down far enough as part of their paper. It absolutely is a valid strategy one can only pursue when one begins with DCA and it would significantly increase the average result for any DCA scenario where the market goes down in the DCA period.

In a general sense, I think we see a lot of this "dumbing down the losing strategy" theme in a lot of places and it bothers me. The winning side of this (LSI) gets to run with all its' horsepower, but the losing side gets hamstringed and can't use all the opportunities available to it. No wonder it consistently loses.

  • Front-loading tax advantaged accounts in the beginning of the year is not uncommon among high earners; typically administrivia and/or means make this decision more or less moot. Also many matching programs match through the year, in many cases front-loading would lose out on a match. You're correct though, if an investor doesn't have a long enough investment horizon to weather a down market the money shouldn't be allocated to the market at all regardless of LSI or DCA strategy. – quid Aug 28 '18 at 1:38
1

The risk part I believe refers to delaying the investment in DCA as opposed to lump sum, i.e. in both cases, risk is taken, lump sum is risk now, DCA is chopping risk & taking it in installments (And hence later), so it's a risk delay rather than risk mitigation or minimization.

DCA can (in some times) be better than lump sum if combined with statistics & probability. Mutual funds (And markets in general) usually alternate in terms of returns and several consecutive months (10+) of negative/positive returns are rare occurrences and are usually signs of major turning points.

If DCA is done in conjunction with very basic historical returns statistics, it can greatly enhance general performance.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.