Dogpile

This is in market profile section because it will relate back to Jerry's series of posts. This chart by Brett Steenbarger shows that in 2007, the S&Ps were a net loser during the pit session and that all of the gains were made in the overnight session. I am not saying this is a strategy going forward -- merely pointing out that in 2007, if you were to only do long-side trades of the market, you were fighting an uphill battle and needed good intraday location in order to make money. This fact simply underscores Jerrys posts -- which were in part about finding location relative to that days volume weighted price.

yah, I see how this can be a little confusing. the longer tail actually means less conviction in this context. in other places, it shows a period of 'higher timeframe' activity. whenever you see strong 'range expansion off opening price' --- be on alert. this is especially the case if you 'gap out of value'. the vast majority of days, price will trade back to opening price. so when it doesn't do this -- and instead 'dogpiles' away from opening price -- check to see how much activity there was on other side of opening price. also, see this video on trading the first 30 minutes of the day: http://www.cciclub.com/marketvu/linda-raschke-30mins.html

The idea is just to gauge conviction off Opening Price. Extremely strong trend days often show Opening Price as Low or High of Day (Open-Drive). Next is open-test-drive, small quick test but still strong conviction. Next is open-rejection-reverse, test one way and drive the other -- good conviction. The more action around opening-price, the less conviction --- you can call them both 'open-test-drive' -- and just grade the conviction based on how big the tail is, IMO... stopped? I don't see that. it shows that price traded down off opening price and then back above in period D. Period E tests opening price one more time and then launches up. it doesn't stop in period E. maybe you can elaborate if I mis-interpreted what you are saying.

I am left according to this test. I can't make it go clockwise yet. I keep going back to seeing the foot as if I am above the woman looking down. The foot is clearly going counter-clockwise to me and I can't reverse it.

updating this chart. the freefall continues... the 'rate of change' of estimate cuts sure smells like a recession:

another route is to declare a name for a variable like this (on 60-min chart): vars: aa(0), bb(0); if time=1030 then aa=high; if time=1030 then bb=low; ---- then you can reference aa or bb in further code. or for 30-min chart: vars: aa(0), bb(0); if time=1030 then aa=highest(h,2); if time=1030 then bb=lowest(l,2); etc... ----- depending on further code, sometimes may have to write as: if time=1030 then aa=high else aa=0;

one quick correction. my chart above shows Tuesday being a 'low to high' day as designated by the green bar at the end of the day in the indicator 'H or L Made First'. I have watched this situation where the 24-hour chart goes 'high to low' and the pit session chart trades 'low to high' --- which is it? In this situation, I default to the 24-hour chart. This didn't matter at the time of the entry yesterday as either chart showed the same thing in terms of the set-up -- the correction is really for the set-up for today. Here is the EasyLanguage Code I am using for this indicator at present:

hmm.. I don't have an elegant solution so I will venture an unelegant one... open one of the indicators that does work in Radarscreen and re-save it as a new name. check to see if the new name is now available to be imported into Radarscreen. next, copy/paste the code from the new indicator you want to import into the re-saved document (the one you just re-named) and re-verify it (F3). Now see if its available? this is a non-programmers approach... gl.

This being the 3rd month of this thread, thought I would try to generate some systematic rules. I will try to 'score' the quality of the Taylor set-up. Taylor is really just a method to try to seek good 'location' for a trade. This is similar to Market Profile in a lot of respects. Pg 88 of 'Markets In Profile' (Jim Daltons excellent book) could just as easily be applied to George Taylors method: "Remember the subtlety in what we do: we don't forecast, we assess the risk of our positions. We seek to... establish positions that provide favorable risk/reward characteristics." I have used 4 primary rules to try to do something Taylor-like. 1. What is the 'high to low'/'low to high' pattern of the preceeding 2-4 days 2. Is there a HV or LV (Violoation of previous day high/violation of previous day low) 3. What is the 'location' of price relative to the 15-min 20-period Exponential Moving Average 4. What are key support resistance prices from previous days (highs/lows and PVP's) Finally, there is one thing that is kind of subjective that you have to watch: Is the market trending strongly off opening price? (called 'Open-Drive' in Market Profile). Todays Taylor Set-Up: 1. There were 2 days that traded 'High to Low' 2. We had a 'Low Violation' 3. We were trading FAR below the 15-min 20-ema 4. We had key support in 1461 area and price traded into 1462.50 this morning. *We did not have 'open-drive' -- strong trending action off opening price... Thus, we had an excellent 'Taylor BUY Set-Up.' This can also be interpreted as: 'The risk of being short was high due to bad location' 'The reward vs risk for being long was good'

http://rick.bookstaber.com/

I got the earnings data from a research service I get -- I entered the data myself to create the chart. Note that Q3 had tons of write-downs so Q3 earnings were really weak -- this is why I am using forward earnings estimates and not reported results. Forward estimates reflect forward expectations and thus changes in forward estimates reflects changes in expectations. Here is 1 more look. Note that Q4 earnings have been marked down from what was double-digit growth (+10 to +12%) down to 0% as of today. Q1 2008 has started to come down but still stands at +7%. Thus, you have a situation where either: 1) Q4 estimates have been marked down to beatable levels or, 2) 2008 earnings expectations remain too aggressive and will need to be marked down. Many are discussing statistics to show we are not in recession, ie, employment/GDP etc.... I am saying the more important thing to watch is what profits are doing. As an investor, who cares if you are in recession or not when earnings are falling this hard? given what has happened to earnings over last few months --- 2008 earnings estimates look pretty aggressive at this point -- and may need to come down. the extent to which is the big question. The global economy has been booming. Begs the question of what happens to earnings if global economy does not boom in 2008. It could get ugly. But that remains to be seen. I will just continue to monitor the situation.

Within the indicator code is an option as to which type of window you can open that indicator to.... You likely don't have the box checked that enables that particular indicator to work in Radarscreen. Go to 'File' - 'Open Easylanguage Document' and choose the indicator you would like to make available to Radarscreen Right click the mouse within the EL document and choose 'Properties' Then click on the 'Applications' tab and check the box for 'Radarscreen' when you have done this, click 'F3' and document will save and re-verify and you are now done and can close the document and open it in radarscreen.

Reported GDP by the government was a healthy +4.9% last quarter. Meanwhile, S&P 500 forward earnings estimates have been in freefall lately. Note the stability all year long until late August, when estimates began to trend down. The market made a new high while estimates began to fall and then the bottom came out of estimates in November.

I wrote a long and detailed response and then proceeded to erase it by accident. The net of it was that this was a great week to be a Taylor/Market Profile disciple. I have annotated a chart and hopefully it makes some sense as I don't want to write out my entire post again. Maybe someone else can take a crack at a detailed write-up of this week.

here is an example from today. I entered ES when it appeared price failed just under VWAP. I did this with a market order before the 'confirmation' of the reversal price triggered. I set a stop-market order at the computed reversal price to enter YM as a way to add to my initial order. I totally screwed up on the YM exit as you can see. The key here was not this indicator -- but it was nice in the heat of battle to have a computed price to go with. Just an example.

but of course. this is a constant work in process. you need a rule in the code to indicate whether the computed reversal price is off the lowest low or highest high -- you can see I set this to be where the midpoint of the last bar is relative to the range of the last 6 bars. if anyone can think of a better way to do this, I am all ears. clearly, you cannot use this indicator by itself -- you must incorporate many other factors -- ie, higher timeframe pattern, vwap etc.. whatever you use... but this indicator gives a price as an idea for an entry that you can take or leave.

any coding buffs know how to run a continually updating linear regression line off the adv-decl line since 1st bar of day? I changed the code to make it work off the first bar of the day on regular 'price'... but my adv-decl line runs as an 'indicator' keying off the difference of Data1($Adv) & Data2 ($Decl). thus, is there a simple way to tell this indicator to run off another indicator rather than re-coding the entire thing to reference the difference of data1 & data2?? thx in advance. Here is my current code (which is currently the canned 'Linear Reg Line' TS indicator with a small adjustment to run off the first bar of the day):

hi guys... I have been working on something similar so thought I would chime in... rather than have the indicator follow price as in a close relative to a moving average, personally I find it more useful to compute a price in advance that signals a 'reversal' if touched. Thus you can see a 'mechanical trigger point' in advance rather than having to wait for a bar to close. I realize this is really only a minor difference in what you are doing but thought I would throw this in anyway. Linda Rashke just did a presentation using her own indicator here: https://lbrgroup.com/images//BloombergChicagoNovember2007/bloomberg.pdf her indicator paints the bar depending on where the current bar closes relative to the trailing 16 bar high/low. thus, if in a downtrend -- the bar would be red and switch to green on any close that is above the 16-bar low +2.5 avgtruerange's --- she uses: avgtruerange(9), btw. I have played around these settings and like 1.5 true ranges better and like computing a price in advance rather than waiting for the bar close -- you can use buy and sell stops to enter when you have the computed price in advance. Here is an example:

since you are entering on a little momentum, you want it to move right away... thus if it doesn't, you don't HAVE to wait for the swing low to trigger your stop. if it does move right away in your favor, you can move your stop up fast. I often move my stop quickly to the opening price of the bar that triggered me in or to the low of the bar that triggered me in -- with logic being that the momentum you joined in on needs to carry or else you will exit for scratch or rounding-error type of loss. another thing is if you do get the initial push up, you can take a piece off like 1/4 of the position and now keep a slightly wider stop to ride out some noise and the gain on the 1/4 position offsets the small loss on the 3/4 of the position for again, a scratch. you essentially get a free ride -- a free call option with asymmetric payoff relative to risk. you occassionally get chopped up a little no matter what you do anyway. this type of method works for me because you can still catch a big move -- higher timeframe reward if you are right about the higher timeframe technical pattern with lower timeframe risk if you are wrong. another option is to take a small initial position on a limit with a wide stop and wait for the parabolic to trigger to complete the position and then take your stop up to something near the swing low.

this is just an idea to consider because believe me, I know your pain. that is sick stop out. been there. this kind of shit made me switch to something mentally easier --- entering on stops. this entry method allows you to stick a stop in at the swing low because you are entering AFTER the swing low. enter on buy stop at the parabolic... or write your own indicator to do something similar.

I love the buy/sell stop for entry but many traders I talk to don't seem to use it much. You occassionally join in on a big buy/sell program trade -- catching it early on and when it carries, you have a 'zero-pain' trade on your hands. But it also helps discipline you as you have a natural stop loss point that is created when you trigger into the position.

won't last you 10 months but showering cash on strippers in Vegas is a solid option to consider.

agree that ES is a good one to trade. example: ES has done a big move up and does a blow-off to 1435.50 and instantly reverses with just a few hundred contracts trading at 35.50. This is generally a good pivot high now and you can actually place a stop at 35.50 and feel ok about it, IMO. The move to 35.50 was an extreme that probably had some panic to it by shorts and gunned stops to that level. It might very well go test 35.00 or 35.25 -- but 35.50 should hold. This seems to be a less common situation for YM and ER2 -- and NQ for that matter. that spike to '13009' on YM that looked like stop-gunning to the nth degree might actually run stops again above 13009 --- just for the pure fun of it. Russell is worse, IMO. (This situation can and does of course happen on ES --- but its much less frequent.)

curious if many of you use buy and sell-stop entries as an entry technique a significant amount of the time? one nice aspect to it is that you may not be triggered into a few bad (losing) trades because price won't move into your stop order. another nice aspect is that you will always have a good idea for a stop-loss level if the buy/sell stop triggers you in -- ie, you can always use the a recent bar high/low that just occured prior to the momentum that triggered you into your position as a stop.

Comment regarding watching the late afternoon action to 'read' the tape. On Monday, Nov 19th -- the market attempted to break lower late in the day. On Wednesday, Nov 21st -- the market also attempted to break lower late in the day. There were lots of differences that I won't go into. But watching the volume distribution late in the day was an excellent way to see in real-time what was happening (tape-reading by watching the volume distribution build). Monday: Price attempted down and then volume spiked at 1437.50. 55k contracts built at this price very late in the day. Big money institutions, the 'higher-timeframe' players that you should be aligning yourself with -- were saying, we like 1437.50 a LOT on this particular day. Wednesday: Price moved down and no volume showed up. It just fell and fell. Higher-timeframe institutions were not around or not interested and this led to a very nice move down. Knowing nothing else, (there were big differences in the days anyway), you could have seen how one meant to either close your short and/or go long. The other meant, stay with your short and play for a trend (until volume showed up low in the range).