Trading during last 30 minutes of day...explain please!?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Trading during last 30 minutes of day...explain please!?

Post by Daffy » Tue Sep 27, 2011 11:57 pm

Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day? And I mean really explain it to me. None of this profit-taking, day-trading stuff, as that can't be it otherwise we'd all be profit-taking and day-trading. There has to be a more logical explanation as to why the market always sells off like it does during the last 15-30 minutes, but I don't know how to explain this phenomenon to other people. We saw it again today where the DOW tanked 190 points during the last hour after being up over 300 around 2:15 PM EST. I've been retired the past year so I've had a lot of time to watch the market closely. This happens so often. What is going on?

xerty24
Posts: 4827
Joined: Tue May 15, 2007 3:43 pm

Post by xerty24 » Wed Sep 28, 2011 12:19 am

It runs up like crazy sometimes too, just yesterday for example. What do you want by means of an explanation? Sometimes there are more sellers than buyers, sometimes the opposite. If we could tell you why in advance for any particular day, well, that information would be worth a nearly infinite amount of money so don't expect anyone to give you an answer.

User avatar
grayfox
Posts: 4932
Joined: Sat Sep 15, 2007 4:30 am

Post by grayfox » Wed Sep 28, 2011 12:28 am

Good question. It's hard enough (impossible?) to even know why the DOW is some value like 11,190.69, let alone provide rationale for the path of prices over time. Any explanation would be pure speculation. But that won't stop me me from making up a good story.

I pin the blame on High Frequency Trading. Most trading now is done by computer algorithms that buy and sell, holding stock for only a few minutes or even seconds. Now if you programmed your computer with some algorithm to make trades on milliseconds whenever the market is open, you probably would not want to keep a position after the market closes for 16 hours until the market opens the next day. Too much news can happen that would move prices which you can't trade on. The market would gap up or down, wiping out days or weeks worth of profits. So my guess is that those computers all start unwinding their positions before the end of the day, starting maybe 1/2 hour before the market closes.

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 12:42 am

xerty24 - thanks for trying to simplify my thinking, but I'm not sold that it's random and unexplainable.

grayfox - high frequency trading could be an explanation, but even if we know these high frequency trades are programmed to kick in sell orders toward the end of the day can't this information be exploited? In other words, since it happens all the time (or should I say quite frequently), and everyone therefore expects it to happen, shouldn't the sell-offs start happening sooner than the last 15-30 minutes and eventually smooth out over the course of the afternoon and perhaps the whole day? I find it hard to believe that *all* high frequency trading is executed during the same 15-30 minute time frame. If that were the case then one would think a lot more people would exploit that opportunity.

SP-diceman
Posts: 3968
Joined: Sun Oct 05, 2008 9:17 am

Post by SP-diceman » Wed Sep 28, 2011 12:43 am

(there was actually some bad news/rumors about Europe)

The open and the end of the day are typically the most active.

Traders like to make money quick. (morning)
Others position themselves based on what’s happened. (end of day)

The real fear (for a trader) is tomorrows open.
Should the market gap down you wont be able to sell a position.
(until its too late)
You need to be in the right position or hedged by the close.

The recent sell-off has also created lots of short covering.
Throw in end of the quarter window dressing and you have a mess going on.


Thanks
SP-diceman

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 12:55 am

SP-Diceman - your theory falls in the category that day-traders control market swings in the last 15-30 minutes. Has this theory been studied and confirmed somewhere as proof to these sudden and often steep sell-offs? The sell-offs during the end of day clearly outnumber days of flat activity, or even more rare the late afternoon rally. In my original post I discounted day-traders as the reason for this, but I have a feeling a lot of the replies are going to be like yours. If anyone can point me to academic studies on this I'd appreciate it.

xerty24
Posts: 4827
Joined: Tue May 15, 2007 3:43 pm

Post by xerty24 » Wed Sep 28, 2011 1:05 am

Daffy wrote:xerty24 - thanks for trying to simplify my thinking, but I'm not sold that it's random and unexplainable.
It's not entirely random and unexplainable, but nobody with a real solid and predictive explanation is going to give it you. You might as well ask for a $1B donation.
these high frequency trades are programmed to kick in sell orders toward the end of the day can't this information be exploited? In other words, since it happens all the time (or should I say quite frequently), and everyone therefore expects it to happen, shouldn't the sell-offs start happening sooner than the last 15-30 minutes and eventually smooth out over the course of the afternoon and perhaps the whole day? I find it hard to believe that *all* high frequency trading is executed during the same 15-30 minute time frame. If that were the case then one would think a lot more people would exploit that opportunity.
You're right that HFT goes on all day long, and while it's true that many of those players are trying to close their positions towards the end of the day, it's not at all clear whether those positions in totality are net long or net short (so mostly they trade with eachother to no net effect). That's why it's not obviously exploitable and why it's not an explanation.

xerty24
Posts: 4827
Joined: Tue May 15, 2007 3:43 pm

Post by xerty24 » Wed Sep 28, 2011 1:07 am

Daffy wrote:The sell-offs during the end of day clearly outnumber days of flat activity, or even more rare the late afternoon rally.
I really really doubt this. If I have a chance, maybe I'll try to dig up the intraday data to prove it. I don't know what you call a big selloff or a big rally, but I imagine that most days don't move that much one way or the other.

User avatar
grayfox
Posts: 4932
Joined: Sat Sep 15, 2007 4:30 am

Post by grayfox » Wed Sep 28, 2011 1:13 am

Actually, as I look at the S&P500 chart for yesterday, it looks like the market peaked at 14:14PM (1195.57) and then started heading south. So that was 1 hour 46 minutes before closing.

It hit a bottom at 15:56 (1171.01), then bounced up to 1175.38 at 16:00

http://www.google.com//finance?chdnp=1& ... INX&ntsp=0

Now look at VIX. It started rising at about the same time, from 35.5 at 2:30PM to 38 at 4PM

http://finance.yahoo.com/q/bc?s=%5EVIX& ... z=l&q=l&c=

So there must have been some kind of news making the rounds in the afternoon starting around 2:15 or 2:30 that made everyone cautious.

SP-diceman
Posts: 3968
Joined: Sun Oct 05, 2008 9:17 am

Post by SP-diceman » Wed Sep 28, 2011 1:16 am

Daffy wrote:SP-Diceman - your theory falls in the category that day-traders control market swings in the last 15-30 minutes. Has this theory been studied and confirmed somewhere as proof to these sudden and often steep sell-offs? The sell-offs during the end of day clearly outnumber days of flat activity, or even more rare the late afternoon rally. In my original post I discounted day-traders as the reason for this, but I have a feeling a lot of the replies are going to be like yours. If anyone can point me to academic studies on this I'd appreciate it.
I wouldn’t call it day traders, more like big money activity.
Day traders sounds like a guy playing with $1000.

Remember, when changes happen in the futures contracts
(between actual stocks vs. futures)
that creates stock arbitrage opportunities. (by the big boys)

There’s a site called Indexarb.com (I believe) that goes thru the math.

It tells you at what levels the big boys will kick their programs in.


Thanks
SP-diceman

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 1:19 am

xerty24 - thanks again for replying. If I'm reading your posts correctly, you're taking the position that basically "there is no explanation". I respect your opinion and replies to that position, and perhaps I'm seeking nirvana, but I'd like to continue the dialogue as I'm of a different opinion that something is indeed going on from 3:30-4:00PM EST due to the very noticeable patterns.

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 1:23 am

grayfox wrote:Actually, as I look at the S&P500 chart for yesterday, it looks like the market peaked at 14:14PM (1195.57) and then started heading south. So that was 1 hour 46 minutes before closing.

It hit a bottom at 15:56 (1171.01), then bounced up to 1175.38 at 16:00

http://www.google.com//finance?chdnp=1& ... INX&ntsp=0

Now look at VIX. It started rising at about the same time, from 35.5 at 2:30PM to 38 at 4PM

http://finance.yahoo.com/q/bc?s=%5EVIX& ... z=l&q=l&c=

So there must have been some kind of news making the rounds in the afternoon starting around 2:15 or 2:30 that made everyone cautious.
Don't look at it in one-day isolation. This phenomenon happens more frequently than most perceive, and it's not entirely driven by the day's news from what I can tell. There is a distinct pattern of sell-offs, usually in the last 15-30 minutes of the trading day, regardless of the days events.

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 1:31 am

SP-diceman wrote: Remember, when changes happen in the futures contracts
(between actual stocks vs. futures)
that creates stock arbitrage opportunities. (by the big boys)

There’s a site called Indexarb.com (I believe) that goes thru the math.

It tells you at what levels the big boys will kick their programs in.


Thanks
SP-diceman
If it's the big boys now, and not the $1000 day trader (not what I was originally implying, but I get your point), again why during the last 15-30 minutes? If we know this is going to happen based on the programming why isn't it exploited? Shouldn't other "big boy" investors be exploiting their peers prior to the last 15-30 minutes?

SP-diceman
Posts: 3968
Joined: Sun Oct 05, 2008 9:17 am

Post by SP-diceman » Wed Sep 28, 2011 1:46 am

Daffy wrote:
SP-diceman wrote: Remember, when changes happen in the futures contracts
(between actual stocks vs. futures)
that creates stock arbitrage opportunities. (by the big boys)

There’s a site called Indexarb.com (I believe) that goes thru the math.

It tells you at what levels the big boys will kick their programs in.


Thanks
SP-diceman
If it's the big boys now, and not the $1000 day trader (not what I was originally implying, but I get your point), again why during the last 15-30 minutes? If we know this is going to happen based on the programming why isn't it exploited? Shouldn't other "big boy" investors be exploiting their peers prior to the last 15-30 minutes?
It depends when the situation arises.
(obviously many are squaring things towards the end of the day)
More activity at the open/end make it more likely to happen then.

If SP500 futures sell at $90 and a stock index is at $100 it makes sense
to buy the futures and sell short stocks.
(I will have a guarantee of $10 at no risk whenever they reach parity)

It looks like stocks are selling, but its just filling the void.

You cant “exploit” it till it exists.


Thanks
SP-diceman

xerty24
Posts: 4827
Joined: Tue May 15, 2007 3:43 pm

Post by xerty24 » Wed Sep 28, 2011 2:24 am

Daffy wrote:xerty24 - thanks again for replying. If I'm reading your posts correctly, you're taking the position that basically "there is no explanation". I respect your opinion and replies to that position, and perhaps I'm seeking nirvana, but I'd like to continue the dialogue as I'm of a different opinion that something is indeed going on from 3:30-4:00PM EST due to the very noticeable patterns.
Daffy - why don't you tell me subjectively how often you think this happens, as a percentage of all days, say for moves bigger or smaller than 0.5%? Do you think declines are more frequent than rallies? If so, by what ratio? You make your guess, then I'll run the numbers.

l2ridehd
Posts: 256
Joined: Sat Sep 10, 2011 5:18 am

Post by l2ridehd » Wed Sep 28, 2011 4:54 am

If everyone looks at there 401K rules and many mutual fund rules, all sell and buy orders are executed at the day end. So if you make a move in your 401K or 509 plan you impact end of day stock direction. I do market timing with my 401K because of the non taxable aspect. I went all in last week and all out yesterday. Picked up a 3.6% value change in about 8 days tax differed. I am sure there are another couple million traders doing the same thing I do. And we all get end of day settlement price. However some where someone is watching all those buy and sell orders as they accumulate all day.

User avatar
SpringMan
Posts: 5357
Joined: Wed Mar 21, 2007 11:32 am
Location: Michigan

Post by SpringMan » Wed Sep 28, 2011 5:20 am

l2ridehd wrote:If everyone looks at there 401K rules and many mutual fund rules, all sell and buy orders are executed at the day end. So if you make a move in your 401K or 509 plan you impact end of day stock direction. I do market timing with my 401K because of the non taxable aspect. I went all in last week and all out yesterday. Picked up a 3.6% value change in about 8 days tax differed. I am sure there are another couple million traders doing the same thing I do. And we all get end of day settlement price. However some where someone is watching all those buy and sell orders as they accumulate all day.
Wouldn't the buy and sell orders in mutual funds or 401(k) plans be executed after the closing NAV of the day for the funds have been determined, thus not affecting the last hour of trading? I could see how orders for ETFs or individual stocks could affect the end of the day direction but mutual fund orders aren't filled until overnight.
Best Wishes, SpringMan

Snowjob
Posts: 1482
Joined: Sun Jun 28, 2009 10:53 pm

Post by Snowjob » Wed Sep 28, 2011 6:23 am

Actually yesterday can be described as an isolated incident as far as the original question was put.

Around 2.15-2.30 Eastern Time, the Financial Times over in London broke a story about how 7 of the 17 euro zone member nations wanted larger haircuts for private bond holders and it highlighted some of the uncertainty around the new rescue measures -- which had been the good news driving the market higher.

livesoft
Posts: 62294
Joined: Thu Mar 01, 2007 8:00 pm

Post by livesoft » Wed Sep 28, 2011 7:10 am

SpringMan wrote:Wouldn't the buy and sell orders in mutual funds or 401(k) plans be executed after the closing NAV of the day for the funds have been determined, thus not affecting the last hour of trading? I could see how orders for ETFs or individual stocks could affect the end of the day direction but mutual fund orders aren't filled until overnight.
No, not at all. M utual fund managers will have to buy/sell throughout the day and the next day in order to deal with the orders from their shareholders. The managers have some leeway because most funds have cash on hand as well.

The managers must be aware of all the submitted orders for the day and know what they need to do. I guess it is possible that if there are quite a lot of orders to buy shares of the fund in the earlier hours that the managers will need to do something in the last hour if they have not been doing something earlier in the day.

Basically, news can cause the market to drop in the last hour. If Bernanke or Geithner or the Fed issues a statement, it will affect the markets.
Wiki This signature message sponsored by sscritic: Learn to fish.

User avatar
Ice-9
Posts: 1306
Joined: Wed Oct 15, 2008 12:40 pm
Location: Rockville, MD

Post by Ice-9 » Wed Sep 28, 2011 8:10 am

Scrolling through Google's chart for VTI in recent days...

9/27 sell off at end of the day
9/26 rise at end of the day
9/23 relatively flat
9/22 slight rise at the end of a day of slight downward movement
9/21 sell off at the end of the day
9/20 more gradual sell off during latter part of the day
9/19 big rise at end of the day
9/16 relatively flat
9/15 gradual rise throughout the day
9/14 sell off at the end of the day
9/13 relatively flat
9/12 rise at the end of the day

From this admittedly small, recent sample, it appears random to me. From personal experience, whenever I trade an ETF, I always fear crazy activity later in the day. Sometimes it moves exactly the way I don't want it to, sometimes it moves in my favor, and sometimes it closes close to the price of my transaction. I not convinced it isn't random.

User avatar
mas
Posts: 1461
Joined: Tue Feb 20, 2007 12:54 pm

Post by mas » Wed Sep 28, 2011 8:46 am

Daffy wrote:If we know this is going to happen based on the programming why isn't it exploited? Shouldn't other "big boy" investors be exploiting their peers prior to the last 15-30 minutes?
I think this gets to the heart of the matter. Yes, big sophisticated investors would exploit this if it truly was a pattern... Then it would cease to be true, simply because of their actions.

You haven't really presented any convincing evidence that what you think happens actually happens. How long has this been happening? What percent of days are down in the last 30 mins? How big is the average drop?

I basically agree with several other posters. The last 30 minutes is probably more volatile than mid day trading (due to "big boys" closing out their positions), but I don't think that the direction is predictable.

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 11:55 am

xerty24 wrote:
Daffy wrote:xerty24 - thanks again for replying. If I'm reading your posts correctly, you're taking the position that basically "there is no explanation". I respect your opinion and replies to that position, and perhaps I'm seeking nirvana, but I'd like to continue the dialogue as I'm of a different opinion that something is indeed going on from 3:30-4:00PM EST due to the very noticeable patterns.
Daffy - why don't you tell me subjectively how often you think this happens, as a percentage of all days, say for moves bigger or smaller than 0.5%? Do you think declines are more frequent than rallies? If so, by what ratio? You make your guess, then I'll run the numbers.
Based on casual observations (nothing scientific, just daily watching of the markets) during the past year I would say sell-offs occur 70-75% of the time during the last 15-30 minutes of the trading day. So based on how you asked that would be roughly 7 out of 10 trading days, or 2.33:1 ratio. I'm not gonna guess on % declines or anything like because that's not what I'm trying to figure out.

If you can track this data down (a) I would be impressed, and (b) I would love to be proved wrong so I can ease my mind about late afternoon volatility.

User avatar
ddb
Posts: 5509
Joined: Mon Feb 26, 2007 12:37 pm
Location: American Gardens Building, West 81st St.

Re: Trading during last 30 minutes of day...explain please!?

Post by ddb » Wed Sep 28, 2011 12:01 pm

Daffy wrote:Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day? And I mean really explain it to me. None of this profit-taking, day-trading stuff, as that can't be it otherwise we'd all be profit-taking and day-trading. There has to be a more logical explanation as to why the market always sells off like it does during the last 15-30 minutes, but I don't know how to explain this phenomenon to other people. We saw it again today where the DOW tanked 190 points during the last hour after being up over 300 around 2:15 PM EST. I've been retired the past year so I've had a lot of time to watch the market closely. This happens so often. What is going on?
Before looking for an answer, perhaps you should gather some data to prove your claim that the market "seems to always tank during the last 30 minutes of the trading day".

Just ballparking recent days of last-half-hour returns of S&P 500 Index since 08/18/2011, which is the last available day where Google Finance shows half-hour price movements on their historical chart. For this exercise, "a little" means less than 1% move, "a lot" means more than 1% move, and flat means less than 1-point move in index value.

08/18 up a little
08/19 down a little
08/22 down a little
08/23 up a little
08/24 up a little
08/25 down a little
08/26 up a little
08/29 up a little
08/30 down a little
08/31 up a little
09/01 down a little
09/02 flat
09/06 up a lot
09/07 up a little
09/08 down a little
09/09 up a little
09/12 up a lot
09/13 up a little
09/14 down a lot
09/15 up a little
09/16 up a little
09/19 down a little
09/20 down a little
09/21 down a lot
09/22 up a lot
09/23 flat
09/26 up a little
09/27 down a little

28 data points: 12 up a little, 9 down a little, 3 up a lot, 2 down a lot, 2 flat. Looks very similar to a normal distribution with "flat" at the center.

So, in more than a month of data, we have two "down a lot" days among a mixed bag of small movements in the last half hour of trading. Where's the beef?

Methinks you fall into the typical behavioral trap of remembering the bad more than you remember the good.

- DDB
"We have to encourage a return to traditional moral values. Most importantly, we have to promote general social concern, and less materialism in young people." - PB

xerty24
Posts: 4827
Joined: Tue May 15, 2007 3:43 pm

Post by xerty24 » Wed Sep 28, 2011 12:22 pm

Daffy wrote:Based on casual observations (nothing scientific, just daily watching of the markets) during the past year I would say sell-offs occur 70-75% of the time during the last 15-30 minutes of the trading day. So based on how you asked that would be roughly 7 out of 10 trading days, or 2.33:1 ratio. I'm not gonna guess on % declines or anything like because that's not what I'm trying to figure out.
The data I have on hand covers most of 2008-2010 inclusive. I looked at the market return in the final 30 minutes, 3:30pm to close, as reflected by the etf SPY. Here were the results:

Code: Select all

avg    0.00%
sigma  0.61%
max   +3.35%
min   -3.53%
count    705
Of the data, 15% (104/705) fell outside the +-1 sigma interval, indicating a rise or fall of more than 0.6%. Of these larger moves, 53 were rallies and 51 were declines.

There's really nothing to see here. It's all just superstition (looking for God or seeing faces on Mars), and the fact that big moves, especially big downward moves when you're betting half your retirement long, are more painfully memorable than other outcomes. But they're just as common as the rallies and, if anything, these bigger moves are less likely than one would expect a priori (compared to the normal distribution).

User avatar
ddb
Posts: 5509
Joined: Mon Feb 26, 2007 12:37 pm
Location: American Gardens Building, West 81st St.

Post by ddb » Wed Sep 28, 2011 12:30 pm

xerty24 wrote:
Daffy wrote:Based on casual observations (nothing scientific, just daily watching of the markets) during the past year I would say sell-offs occur 70-75% of the time during the last 15-30 minutes of the trading day. So based on how you asked that would be roughly 7 out of 10 trading days, or 2.33:1 ratio. I'm not gonna guess on % declines or anything like because that's not what I'm trying to figure out.
The data I have on hand covers most of 2008-2010 inclusive. I looked at the market return in the final 30 minutes, 3:30pm to close, as reflected by the etf SPY. Here were the results:

Code: Select all

avg    0.00%
sigma  0.61%
max   +3.35%
min   -3.53%
count    705
Of the data, 15% (104/705) fell outside the +-1 sigma interval, indicating a rise or fall of more than 0.6%. Of these larger moves, 53 were rallies and 51 were declines.
Uh...I like your data set better than mine :)

Same results, though. Definitely nothing to see here.

- DDB
"We have to encourage a return to traditional moral values. Most importantly, we have to promote general social concern, and less materialism in young people." - PB

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 12:49 pm

Ah! Excellent. Thank you xerty24 and ddb. All along it appears I've been duping myself (or as you say looking for God or seeing faces on Mars... :shock: ). Also the comment about "the typical behavioral trap of remember the bad more than you remember the good" could very well be true. On the bright side, it's good to know I'm "typical".

Thanks everyone for a good discussion and for setting me straight. As you guys have been saying, "there's nothing to see here anymore. Move along."

User avatar
Daffy
Posts: 152
Joined: Tue Nov 06, 2007 4:05 pm

Post by Daffy » Wed Sep 28, 2011 1:35 pm

xerty24 wrote:
Daffy wrote:Based on casual observations (nothing scientific, just daily watching of the markets) during the past year I would say sell-offs occur 70-75% of the time during the last 15-30 minutes of the trading day. So based on how you asked that would be roughly 7 out of 10 trading days, or 2.33:1 ratio. I'm not gonna guess on % declines or anything like because that's not what I'm trying to figure out.
The data I have on hand covers most of 2008-2010 inclusive. I looked at the market return in the final 30 minutes, 3:30pm to close, as reflected by the etf SPY. Here were the results:

Code: Select all

avg    0.00%
sigma  0.61%
max   +3.35%
min   -3.53%
count    705
Of the data, 15% (104/705) fell outside the +-1 sigma interval, indicating a rise or fall of more than 0.6%. Of these larger moves, 53 were rallies and 51 were declines.

There's really nothing to see here. It's all just superstition (looking for God or seeing faces on Mars), and the fact that big moves, especially big downward moves when you're betting half your retirement long, are more painfully memorable than other outcomes. But they're just as common as the rallies and, if anything, these bigger moves are less likely than one would expect a priori (compared to the normal distribution).
Actually, one last thought. What if you were to throw out all days where the market was flat or down? In other words, take a look at just those days where the market had gains, and then tell me what happens during the last 30 minutes of trading. Perhaps what I'm seeing is sell-offs on just days when the market is up. Anyway to rerun your test to see what happens on that kind of sample?

Edit: BTW, coincidentally it happened again today, even though the market was already down once the 3:00PM hour rolled around....LOL.

Image
Last edited by Daffy on Wed Sep 28, 2011 3:23 pm, edited 3 times in total.

Dandy
Posts: 5287
Joined: Sun Apr 25, 2010 7:42 pm

Post by Dandy » Wed Sep 28, 2011 1:43 pm

It doesn't always tank but many times is much more volitile. There are a lot of institutions managing money. Mutual Funds, hedge funds, pension funds, etc. These institutions often wait till the last minute to place trades and they are often very large trades. They want to get the latest information before placing a trade.

Our Institutional trading desk would often get swamped with trades at 3.55 pm and would pay dearly for any trade errors. Finally, the fund portfolio managers required their specific permission for trades over a certain amount that were called in late. This late institutional trading seemed to be even greater before weekends/holidays.

ftobin
Posts: 1022
Joined: Fri Mar 20, 2009 3:28 pm

Post by ftobin » Wed Sep 28, 2011 2:33 pm

Leveraged ETFs rebalancing definitely contribute to the movement's you're seeing.

hicabob
Posts: 2690
Joined: Fri May 27, 2011 5:35 pm
Location: cruz

Post by hicabob » Wed Sep 28, 2011 6:01 pm

It worked again! Alright from now on, instead of being all Bogleheaded, we buy spy at 3:59EST then sell it the next day at 10:00 EST and become bazillionaires in a few weeks 8-)

umfundi
Posts: 3361
Joined: Tue Jun 07, 2011 5:26 pm

Post by umfundi » Wed Sep 28, 2011 8:41 pm

So there must have been some kind of news making the rounds in the afternoon starting around 2:15 or 2:30 that made everyone cautious.
Well, if that news made everyone cautious, it must have been known to everyone. What was that "some kind of news"?

These postdictions are just as much garbage as the predictions. For weeks now, I have heard on the radio every day:

Prices are up, "Investors are optimistic about the Greek debt crisis", or;

Prices are down, "Investors are pessimistic about the Greek debt crisis".

Get a beer (or are we the Martini crowd?), sit on your deck, and contemplate whether that flock of starlings will next veer left, or right. Once you figure that out, you can then work on predicting how much. And, why?

Tomorrow the markets will be up or down. And the Germans will maybe vote on another component of the Greek bailout. And there will be perfectly plausible postdictions of what happens.

Please, why can't you tell me ahead of time?

Keith
Déjà Vu is not a prediction

riskonoff
Posts: 225
Joined: Tue Aug 23, 2011 8:22 am

Re: Trading during last 30 minutes of day...explain please!?

Post by riskonoff » Wed Sep 28, 2011 8:44 pm

Daffy wrote:Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day? And I mean really explain it to me. None of this profit-taking, day-trading stuff, as that can't be it otherwise we'd all be profit-taking and day-trading. There has to be a more logical explanation as to why the market always sells off like it does during the last 15-30 minutes, but I don't know how to explain this phenomenon to other people. We saw it again today where the DOW tanked 190 points during the last hour after being up over 300 around 2:15 PM EST. I've been retired the past year so I've had a lot of time to watch the market closely. This happens so often. What is going on?
High Frequency Trading Front Running. Just as Facebook knows everything, HFT knows everything.

umfundi
Posts: 3361
Joined: Tue Jun 07, 2011 5:26 pm

Post by umfundi » Wed Sep 28, 2011 9:06 pm

Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day?
Who cares why?

Find an ETF that mirrors the market. Sell short 35 minutes before the close, buy 5 minutes before the close.

If your observation is correct, you are all set. I predict you will lose your shirt.

Keith
Déjà Vu is not a prediction

Atilla
Posts: 1237
Joined: Tue Feb 09, 2010 7:44 pm

Post by Atilla » Thu Sep 29, 2011 9:17 am

So - on days the market is up big, buy an inverse ETF around noon and sell it before market closing.

Let's look at the price of SDS at noon today and again at market closing... :lol:

User avatar
Epsilon Delta
Posts: 7430
Joined: Thu Apr 28, 2011 7:00 pm

Post by Epsilon Delta » Thu Sep 29, 2011 10:01 am

umfundi wrote:
Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day?
Who cares why?

Find an ETF that mirrors the market. Sell short 35 minutes before the close, buy 5 minutes before the close.

If your observation is correct, you are all set. I predict you will lose your shirt.

Keith
If you think he will lose his shirt you should be willing to take the other side of the bet and win big. I would wager that the expected loss is zero, apart from expenses and with an unknown standard deviation.

umfundi
Posts: 3361
Joined: Tue Jun 07, 2011 5:26 pm

Post by umfundi » Thu Sep 29, 2011 11:29 am

Epsilon Delta wrote:If you think he will lose his shirt you should be willing to take the other side of the bet and win big. I would wager that the expected loss is zero, apart from expenses and with an unknown standard deviation.
You are correct. The transaction itself is probably a zero-sum game. I was thinking that expenses would quickly erode your stake.

I'd rather be a day trader's broker than a day trader. :)

Keith
Déjà Vu is not a prediction

natureexplorer
Posts: 4191
Joined: Thu Sep 03, 2009 10:52 am
Location: Houston

Post by natureexplorer » Thu Sep 29, 2011 11:49 am

Pattern imagination at work.

gulliver
Posts: 93
Joined: Wed Aug 05, 2009 12:54 pm

Post by gulliver » Thu Sep 29, 2011 1:25 pm

This doesn't really answer your question but...quite a while ago I worked on a project for a back office on wall street and when we implemented I spent a fair amount of time hanging around the equity desks and I learned the following:

1. traders like to talk about sports
2. when volume is light and the averages aren't moving much, traders are grumpy and spend a lot of time thinking about what other traders are thinking about. traders hate trading into a dead market. one of them referred to it as "working in the dark."
3. when volume is high and the indexes are jumping, traders are happy and they trade a lot.

My conclusion was that there's a herd mentality which more or less feeds on itself. Not sure if that's quantifiable in a way that can make money though I'm sure a lot of people have tried!

User avatar
ddb
Posts: 5509
Joined: Mon Feb 26, 2007 12:37 pm
Location: American Gardens Building, West 81st St.

Post by ddb » Thu Sep 29, 2011 3:04 pm

Nice final 30 minutes today!
"We have to encourage a return to traditional moral values. Most importantly, we have to promote general social concern, and less materialism in young people." - PB

umfundi
Posts: 3361
Joined: Tue Jun 07, 2011 5:26 pm

Post by umfundi » Thu Sep 29, 2011 3:45 pm

ddb wrote:Nice final 30 minutes today!
Dang! There goes that theory!
herd mentality
Be nice! It's called "momentum".

Keith :P
Déjà Vu is not a prediction

squirm
Posts: 1426
Joined: Sat Mar 19, 2011 11:53 am

Post by squirm » Fri Sep 30, 2011 12:01 am

I'd like to know why the markets are even open all day. Most trading takes place in the first and last hours anyway. I blame all those high frequency trading boxes for all the trading during the last hour.

User avatar
OAG
Posts: 1061
Joined: Sat Mar 03, 2007 11:54 am
Location: Currently Central Ohio, USA

Post by OAG » Fri Sep 30, 2011 5:19 am

Some (sell off) could be attributed to Mutual Funds selling to cover redemptions requested that day. Same could be said for Banks and Insurance companies to cover required ratios. Maybe just BIG LARGE holdings doing asset allocation for the day. Pure specuation on my part.

Why aren't markets open all day every day 24/7? Seems "after hours" trading extends the day. Also Markets open in part of the world at 8 PM EST and final closing for the day is in NY at 4 PM EST. So I guess one could say they are actually open all but 4 hours daily excluding weekends and holidays.
OAG=Old Army Guy. Retired CW4 USA (US Army) in 1979.

User avatar
Majormajor78
Posts: 910
Joined: Mon Jan 31, 2011 9:13 pm
Location: Chicagoish

Re: Trading during last 30 minutes of day...explain please!?

Post by Majormajor78 » Tue Oct 04, 2011 3:20 pm

Daffy wrote:Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day? And I mean really explain it to me. None of this profit-taking, day-trading stuff, as that can't be it otherwise we'd all be profit-taking and day-trading. There has to be a more logical explanation as to why the market always sells off like it does during the last 15-30 minutes, but I don't know how to explain this phenomenon to other people. We saw it again today where the DOW tanked 190 points during the last hour after being up over 300 around 2:15 PM EST. I've been retired the past year so I've had a lot of time to watch the market closely. This happens so often. What is going on?
How about today? My Vanguard extended market ETF VXF jumped over 5.5% in the last 42 minutes. If the market could predictably "tank" in the last 30 or so minutes then there would be traders getting filthy stinking rich off the phenomenon. Volitility swings both ways.
"Oh, M. le Comte, it is only a loss of money which I have sustained... nothing worth mentioning, I assure you."

User avatar
Dan Moroboshi
Posts: 866
Joined: Sat Jul 07, 2007 7:10 pm

Re: Trading during last 30 minutes of day...explain please!?

Post by Dan Moroboshi » Tue Oct 04, 2011 4:18 pm

Majormajor78 wrote:
Daffy wrote:Can someone explain to me why the market seems to always tank during the last 30 minutes of the trading day? And I mean really explain it to me. None of this profit-taking, day-trading stuff, as that can't be it otherwise we'd all be profit-taking and day-trading. There has to be a more logical explanation as to why the market always sells off like it does during the last 15-30 minutes, but I don't know how to explain this phenomenon to other people. We saw it again today where the DOW tanked 190 points during the last hour after being up over 300 around 2:15 PM EST. I've been retired the past year so I've had a lot of time to watch the market closely. This happens so often. What is going on?
How about today? My Vanguard extended market ETF VXF jumped over 5.5% in the last 42 minutes. If the market could predictably "tank" in the last 30 or so minutes then there would be traders getting filthy stinking rich off the phenomenon. Volitility swings both ways.
I knew this thread would be bumped after the market close. I'm just gonna leave this here:

Image

billjohnson
Posts: 620
Joined: Thu Nov 04, 2010 10:41 pm

Re: Trading during last 30 minutes of day...explain please!?

Post by billjohnson » Tue Oct 04, 2011 4:38 pm

Daffy wrote:Can someone really explain it to me.
Most of these pops/drops center around EU bailout rumors. Sometimes it scares the shorts...sometimes the longs....depending on the rumor. Today it was the the shorts who got cold feet.

Minot
Posts: 422
Joined: Sat Jan 29, 2011 2:35 pm

Post by Minot » Tue Oct 04, 2011 6:07 pm

Well, it wiped out over $5000 worth of the tax loss harvesting I did at about 3:30 p.m. (12:30 here on the West Coast).

Minot

User avatar
Oicuryy
Posts: 1236
Joined: Thu Feb 22, 2007 10:29 pm

Post by Oicuryy » Tue Oct 04, 2011 8:03 pm

Daffy wrote:If it's the big boys now, and not the $1000 day trader (not what I was originally implying, but I get your point), again why during the last 15-30 minutes? If we know this is going to happen based on the programming why isn't it exploited? Shouldn't other "big boy" investors be exploiting their peers prior to the last 15-30 minutes?
Markets move at the open and near the close because exchanges announce opening and closing order imbalances. Prices move as needed to clear the imbalance.

http://answers.yahoo.com/question/index ... 644AAR82nX
http://www.tradingmarkets.com/.site/sto ... -78086.cfm
http://www.nyse.com/pdfs/fact_sheet_nyse_open_close.pdf
http://www.nyse.com/equities/nysearcaeq ... 05155.html

Ron
Money is fungible | Abbreviations and Acronyms

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Post by yobria » Tue Oct 04, 2011 8:46 pm

Proving once again - as soon as you think you've found a predictable market pattern, the market will be happy to prove you wrong.

Nick

natureexplorer
Posts: 4191
Joined: Thu Sep 03, 2009 10:52 am
Location: Houston

Post by natureexplorer » Tue Oct 04, 2011 10:36 pm

Looks like international equities are not following the US lead.

User avatar
Rosebud
Posts: 386
Joined: Sun Feb 21, 2010 11:31 pm

Post by Rosebud » Wed Oct 05, 2011 4:28 pm

Here's an article from Reuters on yesterdays massive "melt-up," that attempts to explain yesterday's late increase.

http://www.reuters.com/article/2011/10/ ... WA20111005

Post Reply