"Think you can time the stock market? Look at this chart first"

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Taylor Larimore
Advisory Board
Posts: 27300
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

"Think you can time the stock market? Look at this chart first"

Post by Taylor Larimore » Fri Dec 08, 2017 4:27 pm

Bogleheads:

Today's MarketWatch article gives convincing evidence of how foolish it is for investors to try and time the stock market:

Think You Can Time The Stock Market? Look at This Chart First

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

fantasytensai
Posts: 467
Joined: Fri Sep 30, 2016 3:30 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by fantasytensai » Fri Dec 08, 2017 4:42 pm

Thanks for sharing Taylor. This is a very helpful piece to help people think healthily when hesitating to enter the market. I never heard of this particular author before. I had mostly stayed away from MarketWatch articles lately due to the sheer number is useless articles (like to ones today on tipping guidance). But this was a great find!

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Fri Dec 08, 2017 5:04 pm

Great link and should likely be stickied. I remember reading a book awhile ago which showed missing the best x amount of days since 1970's until now produced returns WORSE then just being in cash during that whole time frame. The idea of being a buy and hold investor is so you do not miss the few good days that produce the majority of all the return.

Off the topic, does anyone know if they have looked at when those "best 20-30 days occur" and see if there is any similarities meaning in relation to valuation metrics? I do remember reading somewhere the best days are usually clumped together (or maybe that was the worst days?).

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

User avatar
Taylor Larimore
Advisory Board
Posts: 27300
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: "Think you can time the stock market? Look at this chart first"

Post by Taylor Larimore » Fri Dec 08, 2017 5:05 pm

fantasytensai wrote:
Fri Dec 08, 2017 4:42 pm
Thanks for sharing Taylor. This is a very helpful piece to help people think healthily when hesitating to enter the market. I never heard of this particular author before. I had mostly stayed away from MarketWatch articles lately due to the sheer number is useless articles (like to ones today on tipping guidance). But this was a great find!
fantasytensai:

You're right. Most articles on and off the internet are designed to attract viewers, subscribers, or to sell us something. That's why an article like this is so refreshing.

Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: "Think you can time the stock market? Look at this chart first"

Post by saltycaper » Fri Dec 08, 2017 5:11 pm

I think it would be pretty tough to miss the best days without missing some of the worst days too. Could be a wash. I'm sure someone has already analyzed the occurrence of best/worst days in relation to worst/best days, so all we need now is for someone to find it to satisfy the curious.
Quod vitae sectabor iter?

User avatar
rocket354
Posts: 190
Joined: Mon Dec 14, 2015 12:31 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by rocket354 » Fri Dec 08, 2017 5:18 pm

One thing about the article I have to question is what is the utility of pointing out that if you missed the best x days, you'd be down so much? It's of trivial interest, but as far as market timing goes I think it would be difficult to miss the best, say, 30 days without also missing some of the worst days. I know last year, for example, Brexit was an awful day, but a couple days later was a great day. A more interesting study would be: what would returns have been without the best 30 days and without the worst 30 days? Or some proportional allocation based on actual percentages of positive and negative days. Basically showing that if you miss time in the market, you're just missing out.

Edit: Salty, great minds think alike :)

User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: "Think you can time the stock market? Look at this chart first"

Post by saltycaper » Fri Dec 08, 2017 5:23 pm

rocket354 wrote:
Fri Dec 08, 2017 5:18 pm

It's of trivial interest, but as far as market timing goes I think it would be difficult to miss the best, say, 30 days without also missing some of the worst days.
Exactly. I wonder if it would be harder to miss the 30 best days without missing the 30 worst days than it would be to miss both. In which case the article would be more amusing than actually indicative of the dangers of market timing.
Quod vitae sectabor iter?

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by livesoft » Fri Dec 08, 2017 5:29 pm

Sorry, but this article is rehashing old junk. bogleheads.org has often had discussions of missing the best days and worst days. Here is one from 2010:

Missing the best and worst days

If I exchange from my bond fund to my stock fund at the end of a worst day, then by definition that money transferred missed worst day. One doesn't have to predict anything because the transfer is done after the fact. Now whether it does one any good or not is another question.

I want to be fully invested to my asset allocation so that my money does not miss the best days, but that doesn't help normal people who fear losses and have regrets if they buy investments that go down.
Wiki This signature message sponsored by sscritic: Learn to fish.

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Fri Dec 08, 2017 6:24 pm

saltycaper wrote:
Fri Dec 08, 2017 5:11 pm
I think it would be pretty tough to miss the best days without missing some of the worst days too. Could be a wash. I'm sure someone has already analyzed the occurrence of best/worst days in relation to worst/best days, so all we need now is for someone to find it to satisfy the curious.
I don't understand this comment? Are you saying if one market times it is not as bad as the results as they "may" miss the best 20 (for example) days because they would also miss some of he worst days as well thus cancelling them out?

If so, what data do you have to make that assumption? My default in logic without any evidence without some data would be folks would market time randomly and miss the best days AND also miss avoiding the worst days. I would need some data to support any other view. But assuming the opposite without data to support it is an unusual POV.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

wrongfunds
Posts: 1728
Joined: Tue Dec 21, 2010 3:55 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by wrongfunds » Fri Dec 08, 2017 6:25 pm

what exactly is the mechanics involved in missing the best days or for that matter missing the worst days? are you talking about liquidating and then re-entering the market after staying off the market for 24 hour period? isn't this pretty much impossible to do or at least total ridiculous sequence of events? and if so, why are we even starting with "if pigs could fly" premise?

User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: "Think you can time the stock market? Look at this chart first"

Post by saltycaper » Fri Dec 08, 2017 6:39 pm

staythecourse wrote:
Fri Dec 08, 2017 6:24 pm
saltycaper wrote:
Fri Dec 08, 2017 5:11 pm
I think it would be pretty tough to miss the best days without missing some of the worst days too. Could be a wash. I'm sure someone has already analyzed the occurrence of best/worst days in relation to worst/best days, so all we need now is for someone to find it to satisfy the curious.
I don't understand this comment? Are you saying if one market times it is not as bad as the results as they "may" miss the best 20 (for example) days because they would also miss some of he worst days as well thus cancelling them out?

If so, what data do you have to make that assumption? My default in logic without any evidence without some data would be folks would market time randomly and miss the best days AND also miss avoiding the worst days. I would need some data to support any other view. But assuming the opposite without data to support it is an unusual POV.

Good luck.
You understand the comment, except it's not my point of view--it is a guess, speculation, a thing I am wondering about. I have no data, only anecdotal observations. I was wondering if anyone did have the data. More unusual than either POV is to have a strong POV in the absence of any data.
Quod vitae sectabor iter?

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Fri Dec 08, 2017 6:43 pm

saltycaper wrote:
Fri Dec 08, 2017 6:39 pm
staythecourse wrote:
Fri Dec 08, 2017 6:24 pm
saltycaper wrote:
Fri Dec 08, 2017 5:11 pm
I think it would be pretty tough to miss the best days without missing some of the worst days too. Could be a wash. I'm sure someone has already analyzed the occurrence of best/worst days in relation to worst/best days, so all we need now is for someone to find it to satisfy the curious.
I don't understand this comment? Are you saying if one market times it is not as bad as the results as they "may" miss the best 20 (for example) days because they would also miss some of he worst days as well thus cancelling them out?

If so, what data do you have to make that assumption? My default in logic without any evidence without some data would be folks would market time randomly and miss the best days AND also miss avoiding the worst days. I would need some data to support any other view. But assuming the opposite without data to support it is an unusual POV.

Good luck.
You understand the comment, except it's not my point of view--it is a guess, a thing I am wondering about. I have no data, only anecdotal observations. I was wondering if anyone did have the data. More unusual than either POV is to have a strong POV in the absence of any data.
I would be interested as well. Personally, I ALWAYS think of a default on any opinion and then test against the default to see if it right or wrong. Not saying everyone should, but that is how I think.

Either way I was not sure why you would think that the two have anything to do with each other. Thought you had some logic to that opinion and not just a query. My bad.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

zengolf2011
Posts: 108
Joined: Tue Jul 21, 2015 1:27 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by zengolf2011 » Fri Dec 08, 2017 6:54 pm

I think comments about impact of the best and worst days miss the point, which is that the market is unpredictably volatile. There are two ways to win: 1) be really, really, really lucky; or 2) ride the roller coaster and reap the long-term return.

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Fri Dec 08, 2017 6:56 pm

zengolf2011 wrote:
Fri Dec 08, 2017 6:54 pm
I think comments about impact of the best and worst days miss the point, which is that the market is unpredictably volatile. There are two ways to win: 1) be really, really, really lucky; or 2) ride the roller coaster and reap the long-term return.
That would be my thought as well. Either one is lucky or just be all in all the time to make sure you don't miss those best days which drive a large portion of the market success DESPITE taking the lumps of the worst days with it.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

Slacker
Posts: 576
Joined: Thu May 26, 2016 8:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by Slacker » Fri Dec 08, 2017 7:22 pm

zengolf2011 wrote:
Fri Dec 08, 2017 6:54 pm
I think comments about impact of the best and worst days miss the point, which is that the market is unpredictably volatile. There are two ways to win: 1) be really, really, really lucky; or 2) ride the roller coaster and reap the long-term return.
Reminds of a worthwhile "cute" comment I've heard Dave Ramsey say in the past to callers who are wary of the stock markets volatility: "You don't get hurt on a roller coaster if you don't jump off in the middle of the ride"...or something to that effect.

User avatar
abuss368
Posts: 12728
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: "Think you can time the stock market? Look at this chart first"

Post by abuss368 » Fri Dec 08, 2017 8:44 pm

We gave up thinking we were smarter than the market years ago. Thank you Jack Bogle for Vanguard and mutual funds.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + U.S. & International REITs

User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: "Think you can time the stock market? Look at this chart first"

Post by saltycaper » Sat Dec 09, 2017 1:01 am

staythecourse wrote:
Fri Dec 08, 2017 6:43 pm

Either way I was not sure why you would think that the two have anything to do with each other. Thought you had some logic to that opinion and not just a query. My bad.

Good luck.
Perhaps I was not fair in my response because we have different observations. My observation is that volatility often, but not always, appears with positive price swings and negative prices swings "close" together. If the data show that is true, it would make sense to me primarily because the market is often wrong in its predictions/expectations of future events, and is subsequently also wrong in correcting its previous miscalculation. I have no idea where my observations rank on the scale of best and worst days. It's quite possible I am thinking of middling price movements that are substantial but nowhere near the best or worst of all time.

If the data would show that to be true, then there is the question of the hypothetical market timer and what he is up to. There are day-trading market timers on one extreme and rebalancing market timers on the other extreme, but I do not have either of those in mind. I'm thinking of the person who we hear from on the forum from time to time who is reluctant to fully invest or has been waiting for a while to start or has some extra cash and is trying to time their entry. I'm not arguing that what they are doing is a good idea, I'm only questioning the extent to which having the potential to "miss the best days" may be used as an argument against them without considering whether they also have the potential to miss the worst days, which of course they do.

Within the constraints of my imaginary market timer, and within the premises of the article, if the best and worst days occur close together, I don't think he is likely to suffer as much from missing the best days as the article claims. However, if the best and worst days occur very far apart, and especially if the best days occur near other best days and worst days occur near other worst days, then it might be true that missing the best days could be a problem for him.

In the end, I think it would be determined that the premises are baloney, that a day is an arbitrary measurement, and that none of this "missing the best days" analysis really matters for demonstrating any seriously actionable point, whether against marketing timing or for it.
Quod vitae sectabor iter?

JBTX
Posts: 3862
Joined: Wed Jul 26, 2017 12:46 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by JBTX » Sat Dec 09, 2017 3:01 am

https://www.ifa.com/12steps/step4/missi ... orst_days/

Missing the worst days is actually far more beneficial than missing the best days is harmful.

My first reaction to the article was WOW! But after pondering, it is mostly an exercise in curiosity; a fun with numbers of sorts. The odds are probably pretty low that you will consistently miss the best days and but somehow stay in the worst days. It should mostly even out.

25 years ago I used to get a subscription to "the mutual fund forecaster". The main purpose was it gave results of various mutual fund, back before they were easily available on the internet. They also did some forecasting models, which I never followed but read just out of curiousity. Their long term model was horrifically bad, in retrospect. They were saying you should be in 50% cash in the early 90's. However, they also had a different model that looked at historically what days tended to be the worst, it could be calendar days, "triple witching hour" days, etc. The idea was to keep you out of the bad days. My recollection was it had modest success. It didn't beat the market, but seemed to come close to market returns, and kept you out quite a bit, the result being most of the return with a lot less risk. Of course, it could have been all luck.

While I don't believe in market timing, the fact that the worst days are far worse than the best days if anything would tend to lead to theoretical credence to market timing, assuming there was some way to identify those days (which there isn't)

smesman
Posts: 95
Joined: Thu Apr 27, 2017 7:11 am

Re: "Think you can time the stock market? Look at this chart first"

Post by smesman » Sat Dec 09, 2017 4:26 am

Here is a chart I created with the daily changes of VFINX (S&P 500).

The percentage change is calculated by dividing the closing price of the current day by the previous day:

Image

Very good days seem to occur in times of high volatility, in which very bad days also occur.

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by long_gamma » Sat Dec 09, 2017 6:22 am

Doesn't this article encourage market timing?

What if i am only in the market during best days and short the market during the worst days. It is common strategy employed by the day traders. Buy when the market breaks yesterday's high and short it when the market breaks yesterday's low. Close the position at the end of the day using market on close (MOC) order. Sit on the safe instrument like 5 year bonds on other days. Basically it is a momentum play.

Since the return distribution is fat tailed, it is better off to miss the worst days than missing the best days.
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

Ari
Posts: 490
Joined: Sat May 23, 2015 6:59 am

Re: "Think you can time the stock market? Look at this chart first"

Post by Ari » Sat Dec 09, 2017 6:35 am

This article does nothing to dissuade market timing. You're as likely to miss the best days as the worst.
All in, all the time.

lazydavid
Posts: 1751
Joined: Wed Apr 06, 2016 1:37 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by lazydavid » Sat Dec 09, 2017 9:10 am

Ari wrote:
Sat Dec 09, 2017 6:35 am
This article does nothing to dissuade market timing. You're as likely to miss the best days as the worst.
Depends on how people decide to time the market. If you randomly choose time spans to divest, or base decisions on things other than price, you might be right. However, I suspect for a substantial subset, it looks like this:

"Big drop in the market today, probably the start of a long-term downward trend, I better sell and wait it out".

"Ooh, big jump today, must be the start of recovery. Time to get back in."

This market timer has just captured the bad day and missed the good one.

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Sat Dec 09, 2017 9:30 am

JBTX wrote:
Sat Dec 09, 2017 3:01 am
it is mostly an exercise in curiosity; a fun with numbers of sorts. The odds are probably pretty low that you will consistently miss the best days and but somehow stay in the worst days. It should mostly even out.
Maybe I am not understanding. Why do people assume as a thought process that if one misses the best days they are also likely to miss the worst days and it evens out. Each day that a person misses may be the best or worst or neither. Each day missed of a good day only increases slightly the chances of missing the worst day (as one less best day out of the total days left to possibly miss which increases a neither OR a worst day).

Yes you are correct, however, that the professional market timer actually prides themselves on missing the worst days and that is how they improve returns. That is why I asked on my first post on this thread if anyone knows any valuation metrics that would help determine what a good day is in advance (same for bad day). IF looking back on the data shows NO "signs" of a good or bad coming then that ends the conversation very quickly on the usefulness on market timing. Unless one is advocating betting on just getting it lucky.

Either way, If one is going to assume that they are likely to miss bad days as often as good days then why even market time. The only reason to do it would be if you were confident you could miss the bad days MORE then you miss the good days on a magnitude change level, i.e. missing a 4% drop, but by accident missing a 3.6% rise.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: "Think you can time the stock market? Look at this chart first"

Post by staythecourse » Sat Dec 09, 2017 9:35 am

saltycaper wrote:
Sat Dec 09, 2017 1:01 am
My observation is that volatility often, but not always, appears with positive price swings and negative prices swings "close" together. If the data show that is true, it would make sense to me primarily because the market is often wrong in its predictions/expectations of future events, and is subsequently also wrong in correcting its previous miscalculation.
Okay now that makes sense thinking the clustering of best days and worst days are clustered as the markets "correct" itself. I would have thought the market prices one day have nothing to do (little to do) with the markets next day. So I guess we need to see the serial correlation of day to day returns of the market then see the same serial correlation of best days followed by worst days. Interesting.

I feel like asking for Simplegift to help us out. I feel like he should have a chart for this. :D

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Yes, you CAN time the stock market SOMETIMES

Post by livesoft » Sat Dec 09, 2017 9:53 am

So I see from the follow-on discussion that some folks are seeing what has been reported before: Namely that best and worst days often (not always) occur close together in time. Furthermore, sometimes best days occur before worst days and sometimes worst days occur before best days.

The linked article at the start of this thread just highlights the old missing best days story. The missing worst days is even better for an investor. And guess what happens if the investor misses both best and worst days? The investor is better off than if they didn't miss any days. Here is a link to an article showing these results from January 2016:
http://theirrelevantinvestor.com/2016/0 ... t-of-days/

Is there anything actionable here? I think so. Instead of staying in the market through thick and thin, there are signals here which have not been discussed by any of these articles. One should see why they are not discussed: If one found a way to make money easily, then they wouldn't publish their method.
Last edited by livesoft on Sat Dec 09, 2017 10:01 am, edited 2 times in total.
Wiki This signature message sponsored by sscritic: Learn to fish.

User avatar
rustymutt
Posts: 3729
Joined: Sat Mar 07, 2009 12:03 pm
Location: Oklahoma

Re: "Think you can time the stock market? Look at this chart first"

Post by rustymutt » Sat Dec 09, 2017 9:54 am

Great reminder of one of investing's hard lessons for many to understand. Buy and hold will get you there if you're steady and just let time and compounding returns do their job. It worked for me, and will work for anyone whose serious about investing, and not gambling away their wealth, as if Wall Street was a casino.

:moneybag
I'm amazed at the wealth of Knowledge others gather, and share over a lifetime of learning. The mind is truly unique. It's nice when we use it!

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by long_gamma » Sat Dec 09, 2017 10:15 am

livesoft wrote:
Sat Dec 09, 2017 9:53 am

http://theirrelevantinvestor.com/2016/0 ... t-of-days/

Is there anything actionable here? I think so. Instead of staying in the market through thick and thin, there are signals here which have not been discussed by any of these articles. One should see why they are not discussed: If one found a way to make money easily, then they wouldn't publish their method.
Volatility targeting works on the similar principle. When realized volatility starts increasing, cut down the position size and move into safer assets.
Back test of SP500 from portfolio visiualizer
http://bit.ly/2jjciY6

Image
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Re: Yes, you CAN time the stock market SOMETIMES

Post by livesoft » Sat Dec 09, 2017 10:28 am

^And the Volatility portfolio was only 0.5% a year extra. For many readers of this forum that's not worth it. Of course, many posters are paying an extra 0.5% or more in expense ratios, taxes, advisor fees, and behavioral things. This is probably why many people would just say "It's all in the noise."
Wiki This signature message sponsored by sscritic: Learn to fish.

RadAudit
Posts: 3005
Joined: Mon May 26, 2008 10:20 am
Location: Second star on the right and straight on 'til morning

Re: "Think you can time the stock market? Look at this chart first"

Post by RadAudit » Sat Dec 09, 2017 10:32 am

livesoft wrote:
Fri Dec 08, 2017 5:29 pm
bogleheads.org has often had discussions of missing the best days and worst days. Here is one from 2010:
I'm reminded from Peter Pan that this has all happened before and it will all happen again. But, this time ...

Some things just bear repeating. If only for the new arrivals.
FI is the best revenge. LBYM. Invest the rest. Stay the course. - PS: The Calvary isn't coming, kids. You are on your own.

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: Yes, you CAN time the stock market SOMETIMES

Post by long_gamma » Sat Dec 09, 2017 10:38 am

livesoft wrote:
Sat Dec 09, 2017 10:28 am
^And the Volatility portfolio was only 0.5% a year extra. For many readers of this forum that's not worth it. Of course, many posters are paying an extra 0.5% or more in expense ratios, taxes, advisor fees, and behavioral things. This is probably why many people would just say "It's all in the noise."
I see. I tend to look at sharpe ratio. More than 50% improvement in sharpe ratio, along with improved performance is nothing to sneeze at.
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by livesoft » Sat Dec 09, 2017 10:45 am

^And the extra 0.5% in performance is for a 100% allocation to equities, so in a more "normal" portfolio that is allocated to equities and bonds, the extra performance would be even less. For example, a portfolio allocated 50:50 would only see a 0.25% extra performance which can also show up (or disappear) just from own more/less foreign equities or more/less small-cap equities.

Most people on this forum would not think an extra 0.25% a year could be attributed to market timing. They would attribute it to noise and thus not significant and something to just sneeze at.
Wiki This signature message sponsored by sscritic: Learn to fish.

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by long_gamma » Sat Dec 09, 2017 10:58 am

I am not arguing for market timing. Since the article itself is based on the past performance, it is better to point out short comings of selective statistics.

Since you want to compare it to balanced portfolio of stocks and bonds, how about comparing vol. targeting has much lower SD and drawdown than the balanced index.

http://bit.ly/2jjciY6
Image
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

bantam222
Posts: 87
Joined: Sun Nov 23, 2014 5:57 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by bantam222 » Sat Dec 09, 2017 11:21 am

This is interesting but I’m not sure it proves or even suggests you can’t time the market.

Where’s the chart of what your returns are if you miss the x worse days in the market. Wouldn’t a market timer just argue that their “timing abilities” allow them to miss more off the bad days than good days?

This article shows market timing can have big downside but does not show the potential upside. I assume most people that time the market are after that upside potential

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by livesoft » Sat Dec 09, 2017 11:22 am

bantam222 wrote:
Sat Dec 09, 2017 11:21 am
Where’s the chart of what your returns are if you miss the x worse days in the market. Wouldn’t a market timer just argue that their “timing abilities” allow them to miss more off the bad days than good days?
Those charts are in the link I posted.
Wiki This signature message sponsored by sscritic: Learn to fish.

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: Yes, you CAN time the stock market SOMETIMES

Post by staythecourse » Sat Dec 09, 2017 11:25 am

livesoft wrote:
Sat Dec 09, 2017 9:53 am
So I see from the follow-on discussion that some folks are seeing what has been reported before: Namely that best and worst days often (not always) occur close together in time. Furthermore, sometimes best days occur before worst days and sometimes worst days occur before best days.

The linked article at the start of this thread just highlights the old missing best days story. The missing worst days is even better for an investor. And guess what happens if the investor misses both best and worst days? The investor is better off than if they didn't miss any days. Here is a link to an article showing these results from January 2016:
http://theirrelevantinvestor.com/2016/0 ... t-of-days/

Is there anything actionable here? I think so. Instead of staying in the market through thick and thin, there are signals here which have not been discussed by any of these articles. One should see why they are not discussed: If one found a way to make money easily, then they wouldn't publish their method.
Am I really get dense or am I just not understanding. Your article shows what ALL have known for over 10-20 years that missing the worst days is better then missing the best days. Heck, FA use that as the big push for market timing. Nothing new here.

What you, nor anyone else, has shown is:
1. Is there at clustering of best days or worst days. No the last chart does not show that. it shows that clustering of the best and worst are around the highest 30 S.D. OBVIOUSLY having one of the largest movements in either direction (up or down) is going to push that 30 day period of volatility through the roof. Again that chart shows in relation to volatility on one axis and YEARLY return on another. The argument on the table was NOT that, but does the best and worst days cluster around within days of each other. Thus if you miss the best days who cares if you avoid the worst at the same time. That chart does not show that nor was it supposed to show that. So that question is still not answered.
2. The statement basically saying there are signals and they just choose not to publish because they don't want it getting out is EXACTLY what a FA would say. Basically, saying, "Trust me there is a secret sauce that we do that NO ONE else does". To believe that is naive. Again, I am sure someone has looked at the best and worst days to see if there are any valuation metrics that were used to predict on an ex ante basis in attempt to take advantage of it. My opinion, is if ALL the major data is out there with the most high tech computers in world and a LARGE financial incentive to find it and NO ONE has thus far probably means there isn't. To assume the opposite and some guy sitting in their office is hiding it and no one has figured out is a dangerous view point.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: Yes, you CAN time the stock market SOMETIMES

Post by long_gamma » Sat Dec 09, 2017 11:52 am

staythecourse wrote:
Sat Dec 09, 2017 11:25 am

What you, nor anyone else, has shown is:
1. Is there at clustering of best days or worst days. No the last chart does not show that. it shows that clustering of the best and worst are around the highest 30 S.D. OBVIOUSLY having one of the largest movements in either direction (up or down) is going to push that 30 day period of volatility through the roof. Again that chart shows in relation to volatility on one axis and YEARLY return on another. The argument on the table was NOT that, but does the best and worst days cluster around within days of each other. Thus if you miss the best days who cares if you avoid the worst at the same time. That chart does not show that nor was it supposed to show that. So that question is still not answered.
Those are not yearly returns. They are daily returns, similar to OP topic.

Definition of vol. clustering from Mandelbrot (1963 paper)
What is Volatility Clustering? Back in the early 1960s, financial researcher Benoit Mandelbrot observed a phenomenon that occasionally and irregularly affects stock markets, which he called “Volatility Clustering.” Mandelbrot noted that when it comes to stock market volatility, large changes tend to follow large changes, of either sign, and small changes tend to follow small changes. In other words, volatility begets volatility, and such periods of large market swings are called a Volatility Cluster.
staythecourse wrote:
Sat Dec 09, 2017 11:25 am
Again, I am sure someone has looked at the best and worst days to see if there are any valuation metrics that were used to predict on an ex ante basis in attempt to take advantage of it.
Vol. targeting (One can call realized volatility as valuation metric) is applied in most of the momentum funds
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

staythecourse
Posts: 5869
Joined: Mon Jan 03, 2011 9:40 am

Re: Yes, you CAN time the stock market SOMETIMES

Post by staythecourse » Sat Dec 09, 2017 12:24 pm

long_gamma wrote:
Sat Dec 09, 2017 11:52 am
staythecourse wrote:
Sat Dec 09, 2017 11:25 am

What you, nor anyone else, has shown is:
1. Is there at clustering of best days or worst days. No the last chart does not show that. it shows that clustering of the best and worst are around the highest 30 S.D. OBVIOUSLY having one of the largest movements in either direction (up or down) is going to push that 30 day period of volatility through the roof. Again that chart shows in relation to volatility on one axis and YEARLY return on another. The argument on the table was NOT that, but does the best and worst days cluster around within days of each other. Thus if you miss the best days who cares if you avoid the worst at the same time. That chart does not show that nor was it supposed to show that. So that question is still not answered.
Those are not yearly returns. They are daily returns, similar to OP topic.

Definition of vol. clustering from Mandelbrot (1963 paper)
What is Volatility Clustering? Back in the early 1960s, financial researcher Benoit Mandelbrot observed a phenomenon that occasionally and irregularly affects stock markets, which he called “Volatility Clustering.” Mandelbrot noted that when it comes to stock market volatility, large changes tend to follow large changes, of either sign, and small changes tend to follow small changes. In other words, volatility begets volatility, and such periods of large market swings are called a Volatility Cluster.
staythecourse wrote:
Sat Dec 09, 2017 11:25 am
Again, I am sure someone has looked at the best and worst days to see if there are any valuation metrics that were used to predict on an ex ante basis in attempt to take advantage of it.
Vol. targeting (One can call realized volatility as valuation metric) is applied in most of the momentum funds
No the methodology of the study was done with daily returns, but the axis is yearly returns. That is why visually it looks like clusters. There are around 200 trading days/ year and the "clustering" would look MUCH wider apart if the x axis was spread out even quarterly instead of each year. There is a reason they chose to have it represented as they did and it was not by accident. If the x axis was shown quarterly, for example, then the clustering wouldn't look so clustered. Just like the returns would be wider if the x axis was semiannual vs. quarterly. Some visual trickery for sure. Sort of like the chart showing the seemingly benign upward trend in the market since 1926 to current of $1 with very little volatility as opposed to representing the same graph by every month which shows how much up and down the market really goes.

I don't really care about what Mandelbrot defines as clustering. I ONLY care what happens in real life. The question is what happened since (let say his definition in 1960's) to current. Naively, I would think the easiest way would be to find out how many days between the best/ worst 100, then 50, then 20, then 10 days. If there is a pattern there should be shorter duration of days as the volatility increases from 100, to 50, to 20, to 10.

Agreed, if it does show volatility as the indicator (that is IF the 30 days volatility the are looking at does NOT Include the volatility of the best/ worst day itself thus distorting that calculation itself) that would be useful. Then, of course, you still need to know 1. What measure would you use ?S.D., 2. What precise number do you get out? 3. What number do you get back in? 4. What is the impact of taxes, costs of trading, etc...? , 5. How much of an advantage in return is it over the simple buy-hold approach.

Personally, I am not a buy/hold investor because I want to be. I REALLY do want to use active management in some form to improve my returns, but just have not seen any unbiased and good data to show there is a reproduceable approach over the default of just buy/ hold.

If someone figures it out PM me!! :D
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

long_gamma
Posts: 226
Joined: Mon Sep 14, 2015 4:13 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by long_gamma » Sat Dec 09, 2017 1:30 pm

staythecourse wrote:
Sat Dec 09, 2017 12:24 pm

No the methodology of the study was done with daily returns, but the axis is yearly returns. That is why visually it looks like clusters. There are around 200 trading days/ year and the "clustering" would look MUCH wider apart if the x axis was spread out even quarterly instead of each year. There is a reason they chose to have it represented as they did and it was not by accident. If the x axis was shown quarterly, for example, then the clustering wouldn't look so clustered. Just like the returns would be wider if the x axis was semiannual vs. quarterly. Some visual trickery for sure.
You are right, they are not daily returns. Here is the monthly return chart of SPY from Reuters platform. Visually it does cluster together.

Image
staythecourse wrote:
Sat Dec 09, 2017 12:24 pm


Agreed, if it does show volatility as the indicator (that is IF the 30 days volatility the are looking at does NOT Include the volatility of the best/ worst day itself thus distorting that calculation itself) that would be useful. Then, of course, you still need to know 1. What measure would you use ?S.D., 2. What precise number do you get out? 3. What number do you get back in? 4. What is the impact of taxes, costs of trading, etc...? , 5. How much of an advantage in return is it over the simple buy-hold approach.
Volatility targeting is based on investor risk tolerance. It is dynamic in the sense, when the realized targeting rises above the pre-defined level then the position is reduced and when it get back below that pre-defined level position size goes back to normal (or can be levered up). These are effective in tax advantaged account. Cost of trading is higher of-course, but for investor who does not have tolerance for draw-down that is the price to pay.
http://www.lazardnet.com/docs/sp0/10211 ... search.pdf

Auto-correlation of SPY from that article.

Image

How much of an advantage? I posted the chart above comparing the two.
"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson

JBTX
Posts: 3862
Joined: Wed Jul 26, 2017 12:46 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by JBTX » Sat Dec 09, 2017 2:54 pm

staythecourse wrote:
Sat Dec 09, 2017 9:30 am
JBTX wrote:
Sat Dec 09, 2017 3:01 am
it is mostly an exercise in curiosity; a fun with numbers of sorts. The odds are probably pretty low that you will consistently miss the best days and but somehow stay in the worst days. It should mostly even out.
Maybe I am not understanding. Why do people assume as a thought process that if one misses the best days they are also likely to miss the worst days and it evens out. Each day that a person misses may be the best or worst or neither. Each day missed of a good day only increases slightly the chances of missing the worst day (as one less best day out of the total days left to possibly miss which increases a neither OR a worst day).

Yes you are correct, however, that the professional market timer actually prides themselves on missing the worst days and that is how they improve returns. That is why I asked on my first post on this thread if anyone knows any valuation metrics that would help determine what a good day is in advance (same for bad day). IF looking back on the data shows NO "signs" of a good or bad coming then that ends the conversation very quickly on the usefulness on market timing. Unless one is advocating betting on just getting it lucky.

Either way, If one is going to assume that they are likely to miss bad days as often as good days then why even market time. The only reason to do it would be if you were confident you could miss the bad days MORE then you miss the good days on a magnitude change level, i.e. missing a 4% drop, but by accident missing a 3.6% rise.

Good luck.
I am assuming that the days missed will essentially be random, it could be a great day, or bad day, or both. However, I am going to bet, based on nothing but a hunch, (and also hinted at by another poster above) that the best days and worst days are often clustered together. When the market is tanking, it isn't unusual for the market to bounce back up, at least part of the way (look up dead cat bounce!). So I am going to guess the odds are greater than random that if you miss some of the best days, you miss some of the worst days too. But that is just speculation. I have no hard data to base that upon.

livesoft
Posts: 62251
Joined: Thu Mar 01, 2007 8:00 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by livesoft » Sat Dec 09, 2017 3:11 pm

Money that I have invested in bond funds ALWAYS misses the worst days of the stock market. By definition.
Wiki This signature message sponsored by sscritic: Learn to fish.

User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: "Think you can time the stock market? Look at this chart first"

Post by saltycaper » Sat Dec 09, 2017 3:51 pm

After data is obtained for best and worst days, please re-run model for best and worst fortnights. Also for every e hours the market is open.
Quod vitae sectabor iter?

pkcrafter
Posts: 13001
Joined: Sun Mar 04, 2007 12:19 pm
Location: CA
Contact:

Re: "Think you can time the stock market? Look at this chart first"

Post by pkcrafter » Sat Dec 09, 2017 11:48 pm

Guess I don't get it. If you market time you are certain to miss some of those best days. The message seems clear--stay invested.


Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.

MrPotatoHead
Posts: 429
Joined: Sat Oct 14, 2017 10:41 pm

Re: "Think you can time the stock market? Look at this chart first"

Post by MrPotatoHead » Sun Dec 10, 2017 11:08 am

The self heralded great market timer Bob Brinker made the decision to get 100% out of the market after the October 1987 crash. He actually went from 100% equities before the crash to 100% cash in January of 1988 with the DOW at 2015. He tepidly re-entered with a partial allocation in Feb of 1989 after the market had recovered and was at 50% equity allocation with the DOW now at 2342. He then vacillated in various allocations allocation between 40% and 95% until he finally returned to a 100% equity position in January of 1991. So essentially he sold low and bought in high with several adjustments up and down in terms of percentages to ring the tax man's bell. And that folks is allegedly the best of the market timers. LOL...

Post Reply