Author Topic: Long Term Market Timing (using DCA)  (Read 6472 times)

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Long Term Market Timing (using DCA)
« on: June 18, 2014, 07:16:01 AM »
I've had this notion in the back of my mind that I'd like to explore.  I consistently buy shares every two weeks on a Friday morning.  The reason for this is simple: We get paid Wed, transfer to brokerage arrives Fri. 

The question I have is - is Friday the best day long term to buy into the market?  Now, before you yell at me for market timing consider the following.  The information is available for a fairly long historical period that shows which months and indeed, which days of the week offer the best and worst returns long term.  I'm not terribly interested in months of the year as I'm not going to pile cash waiting for the best months to invest.  The risk in this strategy would be too high for my liking.  However there is no real downside in changing the day of the week that I invest. 

So, in consideration of this study, I see some interesting information that might be actionable. 

Does anyone have any commentary about the idea that I should wait until Monday to buy in on a long term DCA strategy?  On average if I buy Monday instead of Friday, it would appear that I would buy low more often. 

I have no idea what the math would look like - perhaps someone here could run a scenario on the advantage gained on average with a fixed amount invested on different days of the week?  Bo_knows?

rmendpara

  • Pencil Stache
  • ****
  • Posts: 602
Re: Long Term Market Timing (using DCA)
« Reply #1 on: June 18, 2014, 08:23:19 AM »
The studies have probably been done already. Search day of the week stock market performance... and you'll probably find something saying Fridays are the best performing days, or whatever.

I think it's more complicated than you need to worry about, but if you're dead set on it, then why not just alternate one of your investments to a Monday (or any other day)?

http://www.investopedia.com/articles/05/seasonaltrends.asp

This suggests that Mondays are the worst performing day of the week (so maybe the best day to invest), however, this ignores the 'monthly effect' and 'seasonal effects'.


Winston

  • Stubble
  • **
  • Posts: 187
  • Age: 42
  • Location: Houston, TX
Re: Long Term Market Timing (using DCA)
« Reply #2 on: June 18, 2014, 08:29:40 AM »
I've often wondered what the "401(k) effect" is, or if it even exists. Does all of the additional buying boost stock prices on typical paydays (e.g. Fridays, 15th of the the month, 1st of the month, etc.)?

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Long Term Market Timing (using DCA)
« Reply #3 on: June 18, 2014, 08:39:38 AM »
Ottawa,

I agree that it is feasible to figure out which day has historically been the best day to invest on. What is much more difficult is figuring out the statistical significance of this finding. In other words are you looking at randomness or are you looking at causality?

What is less controversial is this: buying securities when they are cheap is likely to have higher returns than buying them when they are expensive.

This is the idea behind "value averaging"

The way to do this is to set up an investment schedule.

So let's say you have $1200 to invest over one year.

Your goal would be to have $100 invested on January 1 $200 invested in February 1 $300 invested on March 1 etc.

So let's say on January 31 the value of your portfolio has dropped in half because of a stockmarket collapse. Using value averaging you would invest $150 on February 1 instead of $100 to get your investment in the market up to $200.

But if the market had gone up at the end of January and you now had $150 invested in the market, you would only invest $50 on February 1.

In this way you are buying low preferentially over buying high.

And there is data to support such an approach.

(my favorite approach is just to dump all of my money into the market as soon as it is available.)



AlanStache

  • Handlebar Stache
  • *****
  • Posts: 1982
  • Age: 40
  • Location: South East Virginia
Re: Long Term Market Timing (using DCA)
« Reply #4 on: June 18, 2014, 08:41:16 AM »
Things will also be mixed as some people are paid on the 1/15th and some every two weeks, these are cycles that will be in and out of phase.  Also wonder how many are self directed vs 'pro' directed and if the 'pros' are more willing to sit on cash for a month then buy.  Also lots of $$$ probably goes directly into mutual funds that dont buy 'right away'.  This came up in a different thread and the general answer was something to the effect of buying sooner is better as there is some (very small) per day average compounding value. 

But if you find anything worth reading please link!

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Re: Long Term Market Timing (using DCA)
« Reply #5 on: June 18, 2014, 09:09:43 AM »
The studies have probably been done already. Search day of the week stock market performance... and you'll probably find something saying Fridays are the best performing days, or whatever.

I think it's more complicated than you need to worry about, but if you're dead set on it, then why not just alternate one of your investments to a Monday (or any other day)?

http://www.investopedia.com/articles/05/seasonaltrends.asp


Yes, this link is a more visually pleasing one, compared to the one I linked to in the OP, which some of you may have missed.  http://pages.stern.nyu.edu/~adamodar/New_Home_Page/invfables/pricepatterns.htm.  Thanks for that!

Yes, this:

Quote
This suggests that Mondays are the worst performing day of the week (so maybe the best day to invest), however, this ignores the 'monthly effect' and 'seasonal effects'.

is what I'm referring to:


arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28267
  • Age: -999
  • Location: Seattle, WA
Re: Long Term Market Timing (using DCA)
« Reply #6 on: June 18, 2014, 09:14:12 AM »
We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

mveill1

  • 5 O'Clock Shadow
  • *
  • Posts: 87
  • Location: United Kingdom
Re: Long Term Market Timing (using DCA)
« Reply #7 on: June 18, 2014, 09:24:00 AM »
From your graph, it looks like it's safer to put a market order after trading hours on Monday, so you get the Tuesday morning price? Daily return uses closing. Unless I'm missing something.

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Re: Long Term Market Timing (using DCA)
« Reply #8 on: June 18, 2014, 09:29:06 AM »
Short previous discussion on this topic:
http://forum.mrmoneymustache.com/investor-alley/is-there-a-better-day-of-the-week-to-invest/

Thanks for the link.

Except the previous discussion was largely anecdotal.  I'm interested in the paper I linked to.  Monday is on average (over the 116 years covered) the 'downest' day as of 2001.  Does this trend still hold?  Does it amount to a hill of beans?  I mean, we are only talking about a small effect.  Actually, the graph I linked to is misleading.  See the original article I linked to for the information on the right side axis.

waltworks

  • Magnum Stache
  • ******
  • Posts: 4052
Re: Long Term Market Timing (using DCA)
« Reply #9 on: June 18, 2014, 10:58:42 AM »
I'd guess (again assuming the graph is accurate) that there are a few famous Mondays/outliers/black swans that are pulling down the average, so what you're really hoping is that there will be another black Monday and you'll happen to buy some shares. Is there standard deviation/median information available? Averages are notoriously misleading.

Even if all Mondays are universally bad, unless I am misreading your graph, you are talking about a ~.09% discount if you buy on a Monday. So 9/100 of 1%. Unless you are putting in huge money, that's not enough to worry about IMO (ie if you are buying $1k of shares, you'd get a $0.90 extra gain).

-W

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28267
  • Age: -999
  • Location: Seattle, WA
Re: Long Term Market Timing (using DCA)
« Reply #10 on: June 18, 2014, 11:29:53 AM »
I'd guess (again assuming the graph is accurate) that there are a few famous Mondays/outliers/black swans that are pulling down the average, so what you're really hoping is that there will be another black Monday and you'll happen to buy some shares. Is there standard deviation/median information available? Averages are notoriously misleading.

Even if all Mondays are universally bad, unless I am misreading your graph, you are talking about a ~.09% discount if you buy on a Monday. So 9/100 of 1%. Unless you are putting in huge money, that's not enough to worry about IMO (ie if you are buying $1k of shares, you'd get a $0.90 extra gain).

-W

And if you have to WAIT until the next Monday to invest, how much are you losing out by not being invested an extra few days?  Probably enough to balance it out.
We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

waltworks

  • Magnum Stache
  • ******
  • Posts: 4052
Re: Long Term Market Timing (using DCA)
« Reply #11 on: June 18, 2014, 11:36:35 AM »
Yes, I'd agree with that, given that the existing market timing research says you lose a little money "buying the dips" because of the gains/dividends missed in the interim while waiting for your chosen size of dip. This is just a micro-version of the same idea, so I would expect that at best, it's a very slightly losing proposition.

If it makes the OP happier, sure, set up your buying to happen every other Tuesday morning or whatever. It'll still end up being dollar cost averaging if you stick with it for the long term.

TANSTAAFL, folks. The low hanging fruit in stocks is long gone. If you want bigger returns you have to take more risk or work hard to gain an information edge of some kind.

-W

I'd guess (again assuming the graph is accurate) that there are a few famous Mondays/outliers/black swans that are pulling down the average, so what you're really hoping is that there will be another black Monday and you'll happen to buy some shares. Is there standard deviation/median information available? Averages are notoriously misleading.

Even if all Mondays are universally bad, unless I am misreading your graph, you are talking about a ~.09% discount if you buy on a Monday. So 9/100 of 1%. Unless you are putting in huge money, that's not enough to worry about IMO (ie if you are buying $1k of shares, you'd get a $0.90 extra gain).

-W

And if you have to WAIT until the next Monday to invest, how much are you losing out by not being invested an extra few days?  Probably enough to balance it out.

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Re: Long Term Market Timing (using DCA)
« Reply #12 on: June 18, 2014, 11:43:39 AM »
Yes, I'd agree with that, given that the existing market timing research says you lose a little money "buying the dips" because of the gains/dividends missed in the interim while waiting for your chosen size of dip. This is just a micro-version of the same idea, so I would expect that at best, it's a very slightly losing proposition.

If it makes the OP happier, sure, set up your buying to happen every other Tuesday morning or whatever. It'll still end up being dollar cost averaging if you stick with it for the long term.

TANSTAAFL, folks. The low hanging fruit in stocks is long gone. If you want bigger returns you have to take more risk or work hard to gain an information edge of some kind.

-W

I'd guess (again assuming the graph is accurate) that there are a few famous Mondays/outliers/black swans that are pulling down the average, so what you're really hoping is that there will be another black Monday and you'll happen to buy some shares. Is there standard deviation/median information available? Averages are notoriously misleading.

Even if all Mondays are universally bad, unless I am misreading your graph, you are talking about a ~.09% discount if you buy on a Monday. So 9/100 of 1%. Unless you are putting in huge money, that's not enough to worry about IMO (ie if you are buying $1k of shares, you'd get a $0.90 extra gain).

-W

And if you have to WAIT until the next Monday to invest, how much are you losing out by not being invested an extra few days?  Probably enough to balance it out.

Yeah, I prety much agree with you guys.  It's not worth much, if anything.  It was more a technical note.  Technically, I think you'd come out ahead...over decades but, like walt says - it would be cents...

In reality it is probably not worth the stress of sitting on cash for a weekend :-)

waltworks

  • Magnum Stache
  • ******
  • Posts: 4052
Re: Long Term Market Timing (using DCA)
« Reply #13 on: June 18, 2014, 11:49:18 AM »
I don't think you'd come out ahead, unless you think that Mondays have a high probability of a big crash for whatever reason. I'm pretty sure the mean/average numbers are being strongly influenced by a few really bad Mondays. Throw those out and the whole effect goes away, probably. Like I said, you'd really want median numbers, not mean, to make a decision.

-Walt

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Re: Long Term Market Timing (using DCA)
« Reply #14 on: June 18, 2014, 12:14:08 PM »
I don't think you'd come out ahead, unless you think that Mondays have a high probability of a big crash for whatever reason. I'm pretty sure the mean/average numbers are being strongly influenced by a few really bad Mondays. Throw those out and the whole effect goes away, probably. Like I said, you'd really want median numbers, not mean, to make a decision.

-Walt

No, I don't think the data is due to a few bad Mondays.  Going to the original paper...

Quote
A surprising feature of stock returns is the existence of what is called the weekend effect, another return phenomenon that has persisted over extraordinary long periods and over a number of international markets. It refers to the differences in returns between Mondays and other days of the week.

The returns on Mondays are significantly negative, whereas the returns on every day of the week are not.  In addition, returns on Mondays are negative more often than returns on any other trading day. There are a number of other findings on the Monday effect that researchers have fleshed out.

- The Monday effect is really a weekend effect since the bulk of the negative returns are manifested in the Friday close to Monday open returns. In other words, the negative returns on Monday are generated by the fact that stocks tend to open lower on Mondays than from what happens during the day. The returns from intraday returns on Monday (the price changes from open to close on Monday) are not the culprits in creating the negative returns.
- The Monday effect is worse for small stocks than for larger stocks.  This mirrors our findings on the January effect.
- The Monday effect is no worse following three-day weekends than two-day weekends. 
- Monday returns are more likely to be negative if the returns on the previous Friday were negative. In fact, Monday returns are, on average, positive following positive Friday returns, and are negative 80% of the time following negative Friday returns.[

From the graph in the "Weekend Effect" section of the paper we can discern the following (numbers eyeballed):

- Monday has a positive return on 43% of days.  It has an average daily return of -0.12%
- Tuesday has a positive return on 51% of days.  It has an average daily return of 0.04%
- Wednesday has a positive return on 54% of days.  It has an average daily return of 0.11%
- Thursday has a positive return on 51% of days.  It has an average daily return of 0.05%
- Friday has a positive return on 53% of days.  It has an average daily return of 0.1%

Here is an updated version:



So, I do think it is significant...but, you'd need access to historical data and "bo_knows" level information access/skills to quantify the difference between investing thousands of dollars every other Monday instead of every other Friday over say 5 years.

waltworks

  • Magnum Stache
  • ******
  • Posts: 4052
Re: Long Term Market Timing (using DCA)
« Reply #15 on: June 18, 2014, 12:37:19 PM »
Speaking as a former statistician, Bo knows only clowns use mean without looking at std. deviation and median numbers when doing descriptive statistics. Bo also knows someone out there has probably done an OLS regression (or something more complicated if they wanted to get crazy) of the same basic stuff and found zilch.

-W

Ottawa

  • Handlebar Stache
  • *****
  • Posts: 1026
Re: Long Term Market Timing (using DCA)
« Reply #16 on: June 18, 2014, 01:03:50 PM »
Speaking as a former statistician, Bo knows only clowns use mean without looking at std. deviation and median numbers when doing descriptive statistics. Bo also knows someone out there has probably done an OLS regression (or something more complicated if they wanted to get crazy) of the same basic stuff and found zilch.

-W

Thanks!  You are right. 


The following extract summarizes the graph below:

Quote
The following chart summarizes raw average S&P 500 Index daily returns for all trading days and for each day of the week over the entire sample period, with one standard deviation variability ranges. Results suggest that Tuesdays and Wednesdays are the good days, while Mondays, Thursdays and Fridays are the bad days. However, standard deviations of day-of-the-week returns range from 1.05% (for Fridays) to 1.35% (for Mondays), so variability tends to swamp averages.

So, here is a graph showing SD from 1981 to 2010...



Thanks for letting me muddle through this!! 

:-)

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28267
  • Age: -999
  • Location: Seattle, WA
Re: Long Term Market Timing (using DCA)
« Reply #17 on: June 18, 2014, 01:47:09 PM »
Thanks for letting me muddle through this!! 

Thanks for sharing your results as you did.  :)
We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

AlanStache

  • Handlebar Stache
  • *****
  • Posts: 1982
  • Age: 40
  • Location: South East Virginia
Re: Long Term Market Timing (using DCA)
« Reply #18 on: June 18, 2014, 02:57:00 PM »
So some quick excel math shows that if you bought the open and sold the close on every tuesday with no transaction costs you could return 3.6% over one year!  But expanding this strategy if you bought the open and sold the close five days per week (without transaction costs) you could make 9.8% in the sample period.  This might be worth looking into :-p

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28267
  • Age: -999
  • Location: Seattle, WA
Re: Long Term Market Timing (using DCA)
« Reply #19 on: June 18, 2014, 06:27:39 PM »
So some quick excel math shows that if you bought the open and sold the close on every tuesday with no transaction costs you could return 3.6% over one year!  But expanding this strategy if you bought the open and sold the close five days per week (without transaction costs) you could make 9.8% in the sample period.  This might be worth looking into :-p

We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

kyleaaa

  • Bristles
  • ***
  • Posts: 327
    • Kyle Bumpus
Re: Long Term Market Timing (using DCA)
« Reply #20 on: June 20, 2014, 02:24:47 PM »
The evidence is not strong enough to have any confidence in it, but it certainly wouldn't HURT anything to wait until Monday.  I get paid twice a month (not every 2 weeks), so the day my 401k contributions hit varies month to month. I don't think it's worth really thinking too much about.