The Money Mustache Community
Learning, Sharing, and Teaching => Ask a Mustachian => Topic started by: Clean Shaven on October 07, 2015, 03:50:55 PM
-
Any reason not to try to do "market timing" when investing each month?
In the past, I've had an automatic $1K/month investment transfer set up for Vanguard, for the 16th of each month. Any reason not to do this manually, and try to "time" on a day when perhaps the market is a tad bit lower that month? (either way, that $1K is going in)
Maybe this just isn't worth the effort in the long run (big picture). But I'm bored at work, so why not?
-
Here's why not: Let's say today you get your $1K all queued and and ready to go. The next day the market goes up a little bit, so you hold off waiting for it to go lower. The next day it goes up a little more, and so on for a couple weeks. Then you wind up waiting for an entry point for years.
-
Over your investment horizon, the odds and performance are probably the same no matter what single day a month you choose to invest 1000.
-
There's no good reason not to do three jumping jacks before the transaction either, but neither is going to reliably improve your returns.
-
Here's why not: Let's say today you get your $1K all queued and and ready to go. The next day the market goes up a little bit, so you hold off waiting for it to go lower. The next day it goes up a little more, and so on for a couple weeks. Then you wind up waiting for an entry point for years.
One could always have a "invest on a 2% drop or before the 20th" rule.
-
Here's why not: Let's say today you get your $1K all queued and and ready to go. The next day the market goes up a little bit, so you hold off waiting for it to go lower. The next day it goes up a little more, and so on for a couple weeks. Then you wind up waiting for an entry point for years.
One could always have a "invest on a 2% drop or before the 20th" rule.
How does that backtest?
-
There's no good reason not to do three jumping jacks before the transaction either, but neither is going to reliably improve your returns.
Agree
After all " three is a magic number" ---- cue Schoolhouse Rock
-
Here's why not: Let's say today you get your $1K all queued and and ready to go. The next day the market goes up a little bit, so you hold off waiting for it to go lower. The next day it goes up a little more, and so on for a couple weeks. Then you wind up waiting for an entry point for years.
One could always have a "invest on a 2% drop or before the 20th" rule.
How does that backtest?
No freakin' clue. It'd be interesting to see the results; if I have time, I'll run the numbers.
-
If you could reliably pick the best day to invest (you can't), you could easily make a small fortune in a year.
If you could reliably pretend you knew the best day of the month to invest, you could make a small fortune selling newsletters.
-
Thank you for posting this.
I wondered the same thing last week regarding my payroll 401k contributions every two weeks. Seems like even I (a totally uneducated investor) could "time" the market better than simply buying my normal shares every other Friday (paydays).
My proposal was to change my 401k allocation over to a 100% cash/bond type fund (something stable). Then set up parameters so that once a month, I take the dollars from that fund and distribute them into my normal allocations. Basically, I'm in charge of when I enter the market vs the forced "auto-pilot" approach based on when I get paid.
I don't really see a downside to doing this. If my rules were as follows....
- Get paid - put 401k allocation into cash/bond fund to HOLD
- Watch my index prices - when share price drops below previous payday prices, BUY
- IF share prices never drop below previous payday prices, then BUY on the next payday regardless of price
WORST CASE -- the market never goes down between paydays, and I end up investing last payday's funds on the following payday. This is essentially what's happening now anyway...I'm entering the market every two weeks on payday no matter what the markets are doing. Best case, I do a little better than that; over time perhaps "better' is enough to make a difference.
-
Better to just be consistent. You will not win trying to time the market.
-
Interesting ideas. I'm curious to see a backtest.
-
I don't see the big deal with this as long as you make the contribution. I did this with my IRA, it's really no different than buying when the market is "on sale" (as is often suggested around here). I never tried to "time" anything, just put money in on day when the market went down a little instead of up. I put more than usual in when it went down a lot. Sometimes it went down more the next day and sometimes it went up. It didn't really matter to me once the money was in there.
-
The best day would be the day you received your money. That will give you the best return.
-
In general, the market goes up more often than down, essentially restating the historical trend upwards over the available history. Taking that assumption, and extending it to each and every market day, you should have more "up" days than "down" days. So while you could manage to string together a bunch of instances where waiting a few days to invest may happen to coincide with the down days we see every month, the odds would seem to be against you. The counterpoint to this hypothesis is that it assumes that ups and downs are distributed relatively evenly and in similar degree. It would take an enormously larger amount of time to do, but one could look back in time, and see how ups and downs are distributed, and what the average degree of change is, etc. It's possible that while the market on average shows gains over time, these gains are erratically distributed to a high enough degree that your chance of "missing" the large gains goes down relative to what the average would presume. However, as mentioned above, the effects would be miniscule over time, both positive and negative, with regards to weekly/monthly distributions from paychecks and other time-dependent income. So my conclusion is that it's just not worth it to fret over these. For large distributions and windfalls, it might actually be worth looking at, but in those cases, other forum threads have often pointed to academic research that shows on average lump sum investing ends up being superior in hindsight a large majority of the time.
-
The best day would be the day you received your money. That will give you the best return.
This. Investing sooner > Attempting to time it
One month you'll do great and get in at that 2% drop or whatever and feel like a genius. Then the next month the market will increase all month, and you'll keep waiting for a drop and hit the end of the month and invest on a high. Since the market tends to go up more than it goes down, the second situation will probably happen more often than the first. Sooner is better.
That being said, if it's a fun game for you and keeps you motivated investing it probably won't hurt you too much.
-
Thanks all -
It's probably a "more hassle than it's worth" concept, in the big picture.
-
You could also look at it like this. Which is how I feel about the stock market and why averaging in the best plan, especially for your mentality. Say you had a huge windfall of 500k and you had no exposure to the stock market. You decide you will invest the full 500k in stock market. Say there is a large downturn, and your 500k now becomes only 450k. That will feel like a large loss, however if you had been investing over time and down 50k, you would be either be used to the ups and downs therefore less inclined to either get out altogether or reallocate and pay a tonne of fees. Averaging in over time is a lot better for your mentality than keeping a lump some and getting in on a pullback.
-
If it were that easier or rewarding, everyone would do it.
Jack Bogle: "After nearly 50 years in this business, I do not know of anybody who has done market timing successfully and consistently. I don't even know anybody who knows anybody who has done it successfully and consistently."
-
This isn't (pure) market timing, though. Market timing requires being right twice -- at the top and at the bottom. In some cases, a market timer is in cash because "it's gonna crash at any time!". This is a hybrid approach where the money that will be invested is held until one of two conditions occur:
1) A market drop of X%;
or 2) A date in the month is reached.
Also, unlike market timing, there is no selling. There's also no prediction. The investment occurs after the price movement, rather than before.
The question, then, is whether the average drop before the 16th (or whenever) is more than the average gain before the 16h (what PathtoFIRE wrote about the grouping of gains/drops).
-
Thank you for posting this.
I wondered the same thing last week regarding my payroll 401k contributions every two weeks. Seems like even I (a totally uneducated investor) could "time" the market better than simply buying my normal shares every other Friday (paydays).
My proposal was to change my 401k allocation over to a 100% cash/bond type fund (something stable). Then set up parameters so that once a month, I take the dollars from that fund and distribute them into my normal allocations. Basically, I'm in charge of when I enter the market vs the forced "auto-pilot" approach based on when I get paid.
I don't really see a downside to doing this. If my rules were as follows....
- Get paid - put 401k allocation into cash/bond fund to HOLD
- Watch my index prices - when share price drops below previous payday prices, BUY
- IF share prices never drop below previous payday prices, then BUY on the next payday regardless of price
WORST CASE -- the market never goes down between paydays, and I end up investing last payday's funds on the following payday. This is essentially what's happening now anyway...I'm entering the market every two weeks on payday no matter what the markets are doing. Best case, I do a little better than that; over time perhaps "better' is enough to make a difference.
You left out a scenario: the market goes up x% a day for 2 weeks until you invest. Then you lost out on all those gains. Statistically/historically this will happen ~77% of the time.
-
I think this is just another set of rules around dollar cost averaging. And we know that DCA underperforms just investing the whole pot 2/3rds of the time.
I have to admit though, I think about this kind of thing all the time. The human brain seeks order in chaos, and has a hard time accepting that there is no predictable pattern there.
-
If you could reliably pick the best day to invest (you can't), you could easily make a small fortune in a year.
If you could reliably pretend you knew the best day of the month to invest, you could make a small fortune selling newsletters.
Fantastic. Cheers.
-
This isn't (pure) market timing, though. Market timing requires being right twice -- at the top and at the bottom. In some cases, a market timer is in cash because "it's gonna crash at any time!". This is a hybrid approach where the money that will be invested is held until one of two conditions occur:
1) A market drop of X%;
or 2) A date in the month is reached.
Also, unlike market timing, there is no selling. There's also no prediction. The investment occurs after the price movement, rather than before.
The question, then, is whether the average drop before the 16th (or whenever) is more than the average gain before the 16h (what PathtoFIRE wrote about the grouping of gains/drops).
Like many have said before. There is a 77% chance of the market moving up on any given day. Therefore delaying it will just incur lower returns. Yes, you will get your percentage drops that you are looking for, but the times you miss out on better prices will more than cover the few times you time it well, because of the 77% chance of an up day. Its very simple to the statistically minded. Delaying will only work in a bear market. Funds with cash reserves will always underperform the market.
-
Read this: http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/
Even if you could market time on a monthly basis with perfection, the effect on your results would be unlikely to amount to much compared to simple DCA.
-
The question, then, is whether the average drop before the 16th (or whenever) is more than the average gain before the 16h (what PathtoFIRE wrote about the grouping of gains/drops).
It pretty much can't be, because if the average 'downs' were more significant that the average "ups" the market would be going down over time, not up, right?
Unless the 'ups' were consistently relegated to a certain part of the month, and you could identify that.
-
Like many have said before. There is a 77% chance of the market moving up on any given day.
Citation, please. I read a 58% chance on any given day but it's from some no-name website.
Therefore delaying it will just incur lower returns. Yes, you will get your percentage drops that you are looking for, but the times you miss out on better prices will more than cover the few times you time it well, because of the 77% chance of an up day. Its very simple to the statistically minded. Delaying will only work in a bear market. Funds with cash reserves will always underperform the market.
Let me quote myself,
The question, then, is whether the average drop before the 16th (or whenever) is more than the average gain before the 16h (what PathtoFIRE wrote about the grouping of gains/drops).
That should be "very simple to the statistical minded." If most of the returns happen at the end of the month (..because people invest excess money at the end of the month), then this idea might have some merit. If they're spread out evenly, it's less likely.
This is, incidentally, why someone asked about backtesting. We can argue about it all day but the proof is in the data. I suspect you're right, and no one is suggesting to act on this idea, but that hardly means we need to toss out the idea because it violates a sacred notion. After all, that's why we know that lump sum investing is better than DCA (someone actually backtested and wrote a paper about it).
-
The question, then, is whether the average drop before the 16th (or whenever) is more than the average gain before the 16h (what PathtoFIRE wrote about the grouping of gains/drops).
It pretty much can't be, because if the average 'downs' were more significant that the average "ups" the market would be going down over time, not up, right?
Unless the 'ups' were consistently relegated to a certain part of the month, and you could identify that.
Yeah, if the average gain at day 16 was .15%, with one blip of -1% on day 8, and then the rest of the month gained .40%, it could work.
So, really, you need to not only identify when most of the gains occurred in a month but also when the largest drop occurred.
-
This is EXACTLY what my DH insists on doing.
I'm sending him this thread...dunno if it will make a difference since once he's stuck on an idea, it's pretty set in stone. Good for laughs though :)
(and honey, when you read this, I love you - stubbornness and all!)
-
It's all very negligible as long as you invest it within the month. I wouldn't stress over it.
-
This is EXACTLY what my DH insists on doing.
I'm sending him this thread...dunno if it will make a difference since once he's stuck on an idea, it's pretty set in stone. Good for laughs though :)
(and honey, when you read this, I love you - stubbornness and all!)
Well - I was right. He read this thread and says he still thinks he's right. And he gets frustrated when his mother won't let go of her opinions despite being proven wrong. Like mother, like son LOL!
Now I just have to convince him to quit stashing investments in about a million different accounts. Yesterday, he "found" $8000 in a Questrade account he had opened up half a year ago and "forgotten" about. I counted - between non-reg and registered investments, we have about 15+ accounts (each account holding an average of 3-4 funds) between the 2 of us. *sigh*
-
What I do is automatic investments 4x per month. If the market goes down rather significantly, I throw a few extra bucks at it (I do not change my auto investments). It's not market timing, but it does help me feel like I got a good deal at least on that trade.
I think it brings a little more fun to the investing thing also.
-
Citation, please. I read a 58% chance on any given day but it's from some no-name website.
What date range are you looking for? http://measuringworth.com/DJA/result.php provides daily djia closure numbers and it's -SUPER- easy to do something like this in R-
dja<-read.csv("c:/dja.csv",skip=4)
dja$difference<-c(0,diff(dja$DJIA))
table(dja$difference>0)
That shows 52% (18,692 vs 17119) positive day over day closures from 5/2/1885 through yesterday.
-
Here's a post about dollar cost averaging, the reason that this is usually a loser is also why what you're suggesting would usually be a loser.
http://jlcollinsnh.com/2014/11/12/stocks-part-xxvii-why-i-dont-like-dollar-cost-averaging/
-
If this was possible, why would a hedge fund or mutual fund not already be doing it?
-
Over your investment horizon, the odds and performance are probably the same no matter what single day a month you choose to invest 1000.
This is not true, the odds are that the performance will be better if you do it the first possible day.
-
If this was possible, why would a hedge fund or mutual fund not already be doing it?
+1
If this worked. I would just sell on payday and buy back at the supposed pullback criteria or on the final date allowed to get in. Easy money! Why not trade on Margin as well to inflate the gains. Sounds idiotic because it is..
-
Market timing doesn't work, not now, not ever. It's always a good time to invest according to your asset allocation and risk tolerance.
The worst case scenario of investing at a market high is that your money is invested.