Canadian Money Forum banner

1 - 20 of 29 Posts

·
Registered
Joined
·
3,061 Posts
Discussion Starter #1
Most of my RRSP is in TD eFunds, but the amounts in each fund are creeping up toward the point at which it might be worth considering switching to ETFs.

But the part I've never quite understood is this: everything I've read indicates that (in the long run) dollar-cost averaging is likely to be a more successful strategy than trying to time the market. But unless you have a really big-bucks portfolio, ETFs come out ahead of index funds only if you buy shares just a few times per year (because otherwise the transaction fees could add up to more than you would have lost through your index funds' MERs).

So, if I contribute to my ETF once a month, and let's say each transaction costs me $25, I will have paid $300 per year in fees and gained some of the benefits of dollar-cost averaging. (In reality I contribute to my eFunds with each paycheck, which comes twice a month, so if I did that with ETFs it would be $600/year in fees).

Because ETFs incur transaction fees, I would probably be more likely to contribute just once or twice a year to keep my costs down, but this essentially amounts to market timing and I'd lose the probability benefits of dollar-cost averaging.

Has anyone done a study on this? Or is there some part of the picture I'm missing?
 

·
Registered
Joined
·
2,925 Posts
If you are talking about moving from an eFund to an ETF that tracks the same index then it wouldn't be timing the market. I say this because if your eFund is down so will the ETF and the same goes for when it is up they will both be up. It would only be market timing if you cashed out of your eFund and waited before buying the ETF. So I would think the idea of DCA into an eFund and then switching to an ETF once a year would be a good strategy.
 

·
Registered
Joined
·
3,061 Posts
Discussion Starter #3
No, what I'm referring to is more long term:

I mean if instead of investing in my eFunds, I switch my entire portfolio (well, the part that's in index funds) over to ETFs and invest only in ETFs going forward.

I'm saying that, due to the transaction fees associated with ETFs, I'm likely to contribute to them less often (so if I were investing $20K/year in my RRSP, I might do it in just two $10K installments each year rather than the 12 or 24 installments that I do with my eFunds), and thus lose out on the benefits of dollar cost averaging.

It just seems that this might at least partially offset the benefits of ETFs over funds. Sure, you don't pay an MER, but I guess it boils down to the theoretical difference in your end result if you do dollar-cost-averaging with frequent contributions throughout the year versus what amounts to market timing with fewer and larger contributions each year. I'm wiling to be convinced that ETFs are still better, but this worries me a little.
 

·
Registered
Joined
·
3,423 Posts
If you can handle having 2 brokerage accounts then I would suggest moving all your current investments to ETFs and then just leave that account alone.

In your index fund account you can just keep accumulating until you have enough to move the money to ETFs again.

If this is not palatable then your suggestions of just buying less often is fine.

Transaction fees are 1 drawback of regular ETF purchases. The lack of automation is another - most people are better off making their regular contributions via an automated purchase plan which doesn't exist for ETFs.
 

·
Registered
Joined
·
500 Posts
There is no definitive support for the notion that dollar-cost averaging improves returns ... in some cases it does and in other cases it doesn't ... the best argument in favour of DCA is that it is convenient and can be placed on auto-pilot ... very good for someone who wants to be hands-off.

BTW, DCA does not imply monthly purchases, it implies regular periodic purchases ... so once per quarter, or twice a year, or once a year ... all are still DCA.
 

·
Registered
Joined
·
5,464 Posts
I think the math is pretty clear that DCA does not improve returns when compared in any rational way to lump sum investing.

The perceived advantage of DCA relates to being the market sooner, as well as the behavioural aspect ("invest a little at a time" - which compares DCAing to not investing at all!).

If you google "Dollar Cost Averaging Fallacy" you get a bazillion hits.
 

·
Registered
Joined
·
3,061 Posts
Discussion Starter #7
I
The perceived advantage of DCA relates to being the market sooner, as well as the behavioural aspect ("invest a little at a time" - which compares DCAing to not investing at all!).
My understanding is that DCA's advantage was this:

If you do lump-sum investments in indexes once or twice a year, there's a chance that you will hit the market at a time when it's unusually high or unusually low, and if you're really unlucky you could hit the market at high times every time, meaning that your value basically has nowhere to go but down.

But if you invest a little each month, you're averaging the risk better by spreading it -- yes, sometimes you'll buy high, but other times you'll buy low, and with more data points you're spreading the risk around.

No?

I'm by no means an expert on this, but that was my understanding.
 

·
Registered
Joined
·
5,464 Posts
Brad: what you are doing is buffering the volatility, not increasing the expected return in the MAJORITY of all cases. DCA acts as a shock absorber (which you've described) on both the up and down sides.

Let's imagine 3 "price paths" for your investment, and two scenarios: DCA and lump sum investing.

Price path A: unit value goes up.
Price path B: unit value stays the same.
Price path C: unit value declines.

Given path A, the lump sum investor is better off.

Given path B, it's a tie.

Given path C, the DCA'er is better off.

So is it a tie? Well, actually, no - and this is where probabilities come in. In the long run, price path A is more likely to occur than paths B or C. Therefore, since A has the highest odds of occurring, and lump-sum investing is the best choice for that scenario, then lump-sum investing gives you the best odds.

Think about it this way: if you DCA, and price path A is the norm (over time, diversified portfolio) - then you will suffer a penalty, not gain a reward, from DCA. The penalty comes from NOT being in the market.

Complicated math-y version of the above argument: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=148754

Much more basic explanation, from which this post was quickly cribbed: Chap 2 of Money Logic.

Ultimately, I would argue there are better ways to buffer portfolio volatility than by DCAing. Of course, the behavioural arguments for investing a little at a time are strong. They are just not related to the actual probable investing outcome measured in dollar terms.
 

·
Registered
Joined
·
500 Posts
Brad ... there is nothing in the concept of DCA that requires purchases to be monthly ... therefore, you are comparing DCA to DCA, and concluding that DCA is better.

The best time to invest is when you have the money ... if that means monthly purchases, so be it ... weekly? so be it ... quarterly? so be it.
 

·
Registered
Joined
·
3,061 Posts
Discussion Starter #10
Cool, thanks for that great explanation! This is starting to make sense to my tiny mind.
 

·
Banned
Joined
·
419 Posts
My thoughts are the best time to invest in the market is when you have the money. The whole point of DCA is to invest. Wether it improves returns or not does not matter.
Set up DCA to your paycheck, and "pay yourself first". If you are worried about a yearly lump sum purchase and then a stock market drop then you are trying to time the market.
Some people are not suited to saving for a year and then a lump sum. A big wad of cash may be tempting to purchase that 60" plasma your friend just bought.
In my case, my TFSA looks tempting to splurge , but when i see the dividend income rolling in it keeps me focused on why I saved the money in the first place.
Cash is cash, and people like to spend cash. Once it is DCA'd into your RSP, or whatever there seems to be a mental barrier (in my case anyway) that prevents spending.
5K cash in a bank account can easily be taken out for that week end to Vegas, or flatscreen that you "deserve".

Besides, DCA allows you to put the money into the lowest priced slice of the portfolio pie, so you get a built in buy low, sell high.

For the OP, I believe the Claymore ETF's allow you to make monthly investments into their ETF's.
 

·
Premium Member
Joined
·
2,686 Posts
Brad: what you are doing is buffering the volatility, not increasing the expected return in the MAJORITY of all cases. DCA acts as a shock absorber (which you've described) on both the up and down sides.

Let's imagine 3 "price paths" for your investment, and two scenarios: DCA and lump sum investing.

Price path A: unit value goes up.
Price path B: unit value stays the same.
Price path C: unit value declines.

Given path A, the lump sum investor is better off.

Given path B, it's a tie.

Given path C, the DCA'er is better off.

So is it a tie? Well, actually, no - and this is where probabilities come in. In the long run, price path A is more likely to occur than paths B or C. Therefore, since A has the highest odds of occurring, and lump-sum investing is the best choice for that scenario, then lump-sum investing gives you the best odds.

Think about it this way: if you DCA, and price path A is the norm (over time, diversified portfolio) - then you will suffer a penalty, not gain a reward, from DCA. The penalty comes from NOT being in the market.

Complicated math-y version of the above argument: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=148754

Much more basic explanation, from which this post was quickly cribbed: Chap 2 of Money Logic.

Ultimately, I would argue there are better ways to buffer portfolio volatility than by DCAing. Of course, the behavioural arguments for investing a little at a time are strong. They are just not related to the actual probable investing outcome measured in dollar terms.
MoneyGal's explanation is great and I grant that DCA appears to be safe but it is not. The problem is you could hit a 50% market correction as soon as you complete your DCA purchases. The likelihood is exactly the same if you make a lump-sum investment.

In reality, DCA is great because most people (at least those with a savings habit) are saving money periodically. It is pretty rare to come up on a lump sum to invest.
 

·
Registered
Joined
·
5,464 Posts
I completely agree that compared to not investing at all, then DCA is a great strategy.

However, the financial arguments for DCAing when compared to lump-sum-ing don't hold any water...they're like urban myths. :D
 

·
Registered
Joined
·
88 Posts
I'm with moneygal on this one too. This link is for that gummystuff website I mentioned in my last post. At the bottom of the link is a downloadable spreadsheet that proves lump sum investing fairly convincingly. He takes 1000 random 12 month periods of the SP500 and graphs it out. Lump sum usually wins more than 2/3 of the time.
http://www.gummy-stuff.org/DCA_for_masochists.htm
I know a lot of people can't / don't have the discipline to hoard up money under their mattress and then lump it into the market all at once. So as usual, each to his own.
 

·
Registered
Joined
·
2,925 Posts
DCA will beat lump sum, like real estate beats investing in the stock market.

You look on in shock, but you must realize we are all human and that changes everything. Long term can kill us before it works or our emotions can ruin us before success.

This variable is not accounted for when the examination is done. Because we don't live a long life the math does not account for this and religious stock investors never get this fact.
 

·
Registered
Joined
·
5,464 Posts
This thread is persistently mixing up two things: periodic investing and dollar-cost-averaging, and calling them both DCAing.

For most people, periodic investing is clearly the "right" way to invest, for pragmatic and behavioural reasons. It's how I invest. It's how most people should invest, I'd wager (not that I'm a betting kind of gal).

DCAing is a side effect of periodic investing. It isn't a strategy all on its own, when considered only as an effect of periodic (weekly, monthly, semi-annually, etc.) investing.

The problem I have is when people start to argue for the positive impacts of DCAing as an investment strategy. When you are arguing for the merits of an investment strategy, you must compare your strategy against another investing strategy - not against not investing at all. Of course periodic investing, with its epiphenomenon of DCAing, is preferable to not investing at all.

But if you want to argue that DCAing is a preferable strategy compared to another investing strategy, the math and logic (and reality) is clearly not in your favour.

In addition, most people describe the benefits of DCAing in precisely the way Brad did: as a buffer on the downside. However, DCAing is a buffer on the upside as well - it reduces investment volatility. It does not increase expected return: it *decreases* expected return. (And, I would suggest there are other, probably better ways to reduce investment volatility than DCAing.) It is inappropriate to argue that DCAing will increase your expected return. It will not.

My problem with statements like "You can't lose with DCAing!" is that they belie the lack of thinking and critical reasoning upon which the statement is built. Mushly mutual funds salespeople, who stand to earn a commission from every measly monthly payment Joe Investor deposits to his RRSP, love this kind of crappy, inaccurate, feel-good platitude. (There, my bias is pretty strong.)

OK, I am sure I sound like a fanatic now. But can I just be viewed as a fanatic for clear thinking about this point?
 

·
Registered
Joined
·
3,061 Posts
Discussion Starter #17
OK, I am sure I sound like a fanatic now. But can I just be viewed as a fanatic for clear thinking about this point?
You betcha! Your reasoning is crystal-clear to me, and beautifully explained.

I can't remember where I learned about DCA, it was probably from an early edition of one of Andrew Tobias's books, because (as he said) it was the only investment guide I ever needed, so I don't think I've read any others since. :eek:
 

·
Registered
Joined
·
2,626 Posts
But if you want to argue that DCAing is a preferable strategy compared to another investing strategy, the math and logic (and reality) is clearly not in your favour.
But you yourself explain there is a scenario (1) where DCA outperforms lump sum investing.

I'm not pro-either strategy, and have almost as much conviction as you to stand up against the uber-DCA proponents, but being an uber-lump sum proponent is as 'illogical'.
 

·
Registered
Joined
·
5,464 Posts
But I'm not a lump-sum proponent.

I'm simply saying that the advantages of dollar-cost-averaging are not to increase expected returns. They are to reduce volatility.

This isn't about periodic investing versus lump-sum investing. I'm arguing for clear thinking about the benefits, such as they are, of dollar-cost-averaging.

However: fundamentally: if you are investing in stock markets, presumably you are optimistic about the future performance of your investments. And if we are comparing dollar-cost-averaging against lump sum investing, then - if you are optimistic, why would you *not* make a lump sum investment?

Keep in mind that I said you need to compare one form of investing against another. If you are making periodic investments because that jibes with your cash flow, so be it. However, that isn't a DCA "strategy" - that's a periodic investment strategy.
 

·
Registered
Joined
·
3,061 Posts
Discussion Starter #20 (Edited)
if you are optimistic, why would you *not* make a lump sum investment?
I guess the caveat I would make is that if you're going to make lump sum investments, it seems like you need to pay more attention to the market.

I am indeed optimistic about the market long-term, so my goal is to buy as many shares as I can with each purchase. I am betting that their value will increase over the currently 15-20-year time horizon I'm using for my investments.

If I want to buy as many shares as possible, that means I want them to be as cheap as possible at the moment when I buy them. If I do lump-sum investing and don't pay attention to where the market is when I buy, compared with how it had been doing previously during the year, I could risk always buying when the market is high, and therefore I would be buying fewer shares and setting myself up for less growth.

I'm a highly disciplned saver and am perfectly willing to do lump-sum investing, but I do think that requires me to be much more market-aware than I have been. On the other hand, the realistic scenario would be that I would save up during the year for my lump-sum investment and probably make it in the last quarter of the year (e.g., October), which is when markets frequently hit their low point for the year. So maybe in this case market timing would work out. ;) (I know there's no way to know whether the markets will continue to drop if they've been falling, so it's silly to wait until you think it's hit bottom because you have no real way of knowing, it's little more than a hunch.)

I suppose the best bet is to spread my risk by making two lump-sum investments, one in February and one in October (this is also a realistic scenario, because I get a "retirement bonus" in lieu of a regular retirement plan from my company, and it arrives in February, just in time to contribute to my RRSP before the March 1 deadline).
 
1 - 20 of 29 Posts
Top