NexusFi: Find Your Edge


Home Menu

 





ACD trading By Mark Fisher


Discussion in Traders Hideout

Updated
      Top Posters
    1. looks_one mfbreakout with 195 posts (185 thanks)
    2. looks_two Fat Tails with 92 posts (260 thanks)
    3. looks_3 sysot1t with 21 posts (2 thanks)
    4. looks_4 forrestang with 20 posts (4 thanks)
      Best Posters
    1. looks_one silver99 with 4.7 thanks per post
    2. looks_two Fat Tails with 2.8 thanks per post
    3. looks_3 dandxg with 2 thanks per post
    4. looks_4 mfbreakout with 0.9 thanks per post
    1. trending_up 353,378 views
    2. thumb_up 598 thanks given
    3. group 119 followers
    1. forum 541 posts
    2. attach_file 163 attachments




 
Search this Thread

ACD trading By Mark Fisher

  #11 (permalink)
 dandxg 
Denver, Colorado
 
Experience: Intermediate
Platform: Ninja Trader
Trading: ES, CL, anything that makes $$
Posts: 178 since Nov 2009
Thanks Given: 164
Thanks Received: 77


Fat Tails View Post
If the break-out occurs during the night session, you simply missed it. The market opened exactly at the pivot level R2, so a large part of the move was over.

Now the pivot levels are calculated from the range of the prior day. As yesterday was a NR7 day, the pivot levels are still valid, but they are quite narrow. In this case I look at the average daily range of the last 10 days as a reasonable target. Because the move started during the European session I would look at the ETH range. The ETH range target was already reached prior to the open.

What was behind? Today is a large treasury auction, so prudent market participants shifted their money from treasuries to stocks to protect themselves. During the auction the large market makers support the auction, so there is no flows from treasuries to stocks.

Chart attached.

Average daily range is basically ATR correct? Good points and thanks again. Edited follow. Would you use the same 10 days for stop. I ask because sometimes, CL, other side of OR is $$$.

Reply With Quote

Can you help answer these questions
from other members on NexusFi?
NexusFi Journal Challenge - April 2024
Feedback and Announcements
Better Renko Gaps
The Elite Circle
Futures True Range Report
The Elite Circle
Deepmoney LLM
Elite Quantitative GenAI/LLM
My NT8 Volume Profile Split by Asian/Euro/Open
NinjaTrader
 
  #12 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102


dandxg View Post
Average daily range is basically ATR correct? Good points and thanks again. Edited follow. Would you use the same 10 days for stop. I ask because sometimes, CL, other side of OR is $$$.

The average daily range is just the arithmetic average of the range for the last n days. The average true range uses the formula modified by Welles Wilder. If yesterday's close is within today's range there is no difference.

For the ETH sessions they are nearly always identical. However for the RTH sessions there may be a difference. For example today's ADR(RTH) uses the low of today's RTH session (1276), the ATR uses yesterday's close (1270.50) instead.

The vola bands on the chart refer to ETH and are built from a 3-day and a 10-day ADR.

Reply With Quote
Thanked by:
  #13 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102



ines View Post
Regarding my calculation of A&C values - just have a look at this post #108 of Forex Factory Board called
"The Logical Trader" - Thread to develop manual trading using insights from this book

I would post the links but I do not have permission on this board.

Regards
Ines

It is a recent thread, thank you for the information. The link is here:

"The Logical Trader" - Thread to develop manual trading using insights from this book @ Forex Factory

Let me copy your formulas for A and C values here

aValue = (0.10 *((ATR10Days +ATR30Days)/2));
cValue = (0.15 *((ATR10Days + ATR30Days)/2));

I do no think that they match the value published by Mark Fisher. I have another project to locate these breakout values, based on the daily noise level (moves from the open that failed as a breakout). I also use the 3 day rolling pivot range on my 60 minute chart to locate major support and resistance. Both NoiseBands and RollingPivots can be found in the download section of this forum for NinjaTrader.

The indicators of the thread are MetaTrader indicators, I cannot read them. Maybe you could post some charts and explain, how you use the opening range and the A and C values.

I have also worked on the Logical Trader approach and coded an indicator, which automatically displays pivot range, opening range (period selectable), and A and C values. The charts attached show the price action of yesterday for TF and CL. I am not convinced that the method, which dates back to the 1990's can be traded today without modification. These are the parameters to be modified

-> duration of opening range, as markets have become faster
-> calculation of A and C values should be based on statistical data from the preceding 50 or 100 days

Attached Thumbnails
Click image for larger version

Name:	TF 06-11 (5 Min) 12_04_2011.jpg
Views:	1077
Size:	106.9 KB
ID:	36480   Click image for larger version

Name:	CL 05-11 (5 Min) 12_04_2011.jpg
Views:	824
Size:	99.4 KB
ID:	36481  
Reply With Quote
  #14 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102


ines View Post
>>>> Maybe you could post some charts and explain, how you use the opening range and the A and C values.



The indi is made for forex, therefore no opening and no closing - it just works properly on the 5M chart . My tradingsessios start at 08:00 and I close day-positions latest at 18:30. See blue EU Session close.. This is also the time for me to calculate the macro indicator numberline - see table below..
The eurochart of last Friday - shows a GAU = Good A up Setup. In a trending market it works well - but in Range Market I fade the A/C Values - btw. my favored setups are FAU = Failed A up and FAD = Failed A down - because low risk stops.
1. Trading Session 08:00 - 12:00/13:00
2. Trading Session 14:00 - 18:30



Open Range 30M from 8:00 - 8:30. (Frankfurt = GMT+1)

I use this OR for all 4 majors - I prefer Frankfurt opening instead of London opening. It works fine for me – maybe because I trade mainly EURUSD.
Earlier I used for setup good Aup entry time confirmation from 2x5M candle close above Aup.
In recent years with more dynamic / volatile price action - I skipped this entry condition.
Most difficult for me is mentally - staying in a positive trade on a trend day until the target is reached – that requires sometimes more passion I have.



Hope it helps - regards
ines

PS: I forgot to mention - I calculated the daily Pivot Range based on the data from 8:00 yesterday to 7:55 today in order to include asia night moves.
I know Fish says the OR should be placed in accordance to the domestic market of the traded product - but I prefer my approach for various reasons.
1) 4 Majors are highly correlated
2) I like to compare the 4 against each other durig my trading sessions - therefore I need the same framework ( times )
3) Most top-selling markt is London

Thanks for posting. Some similarities and differences:

I calculate pivots from 5:15 PM Est to 5:00 PM Est (based on the New York Close).

Also have my European Open for the FOREX market at 7:00 AM GMT. That is when volume picks up. NinjaTrader has session templates so I use three sessions: 5:15 PM Est to 7:00 GMT, 7:00 AM GMT to 8:00 AM Est and then 8:00 AM Est to 5:00 PM Est.

For the breakout, I currently rely on noise bands. The noise is a multi day average of the smaller of (High - Open, Open - Low). For example if EURUSD counted from the London open first makes its daily low 30 pips below the open and then makes its daily high 60 pips above the open, 30 pips is the noise. The indicator calculates the average noise over the last M and N days and displays this as a noise band.

I am only taking a breakout against the pivot during the US session (not shown), and if the target band had been reached during the European session. The daily range target bands are based on the average daily ranges as shown in the upper left corner of the chart.


Reply With Quote
Thanked by:
  #15 (permalink)
 dutchbookmaker 
NYC
 
Posts: 187 since Dec 2010


dandxg View Post
Has anyone noticed that is trading ACD, the issue of what to do when a significant narrow range day gets taken before the opening range is even set? It makes for a confusing dilemma.


It is a volatility break idea developed during a trending market. To me the reason The Logical Trader is still a good book to read is because it teaches the difference between running a system vs random trades with no context.

Like Fish says in one of those nymex videos....it was blah blah up..so I had to take it.
My best swing trade yoy so far has been a bull flag on gold even though I've been waiting to buy LEAPS puts on GLD for 2 years. If I see a flag, I take it, unless it is some total and utter rubbish. Swing trades I trade trends on flags and that is it. That is my "system" that filters out what to look at for me.

ORB is nonsense though at this point. It is pit stuff.
If you are going to commit to it, you should be looking to expand it to overnight range breaks, volatility breaks on asia/europe..ect.

Moving pivot avg will work as well as the random line theory thread on this board.

The ultimate goal should be to read the tape/order flow like Tudor Jones as a cotton trader and PTJ mentor in the cotton pits. Fish even admits that.

Reply With Quote
  #16 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102


dutchbookmaker View Post
It is a volatility break idea developed during a trending market. To me the reason The Logical Trader is still a good book to read is because it teaches the difference between running a system vs random trades with no context.

Like Fish says in one of those nymex videos....it was blah blah up..so I had to take it.
My best swing trade yoy so far has been a bull flag on gold even though I've been waiting to buy LEAPS puts on GLD for 2 years. If I see a flag, I take it, unless it is some total and utter rubbish. Swing trades I trade trends on flags and that is it. That is my "system" that filters out what to look at for me.

ORB is nonsense though at this point. It is pit stuff.
If you are going to commit to it, you should be looking to expand it to overnight range breaks, volatility breaks on asia/europe..ect.

Moving pivot avg will work as well as the random line theory thread on this board.

The ultimate goal should be to read the tape/order flow like Tudor Jones as a cotton trader and PTJ mentor in the cotton pits. Fish even admits that.

There are some points with which I do not agree.

(1) There is an indication prior to the open, whether volatility could be up or down. This includes prior range, news, implied volatility and the difference between price and value (pivot range)

(2) Opening range breakouts may well work, if you know at what distance from the open they typically fail. You can take the failure trade, and if it fails, you can trade the breakout. Just need to determine the best distance from the open for that trade. I think we should do an indicator that calculates the statistical distribution of moves from the open for each distance that price has already travelled. Then determine the maximum profit potential. Not easy though.

(3) Rolling pivot ranges work quite well, if you know how to read them. They convey information on potential volatility, trend and act as support/resistance, in particular the balance point.

Reply With Quote
  #17 (permalink)
MissEuroCent
Innsbruck
 
Posts: 7 since Apr 2011
Thanks Given: 0
Thanks Received: 8


Fat Tails View Post
It is a recent thread, thank you for the information. The link is here:

aValue = (0.10 *((ATR10Days +ATR30Days)/2));
cValue = (0.15 *((ATR10Days + ATR30Days)/2));

I do no think that they match the value published by Mark Fisher. I have another project to locate these breakout values, based on the daily noise level (moves from the open that failed as a breakout). I also use the 3 day rolling pivot range on my 60 minute chart to locate major support and resistance. Both NoiseBands and RollingPivots can be found in the download section of this forum for NinjaTrader.

The indicators of the thread are MetaTrader indicators, I cannot read them. Maybe you could post some charts and explain, how you use the opening range and the A and C values.

I have also worked on the Logical Trader approach and coded an indicator, which automatically displays pivot range, opening range (period selectable), and A and C values. The charts attached show the price action of yesterday for TF and CL. I am not convinced that the method, which dates back to the 1990's can be traded today without modification. These are the parameters to be modified

-> duration of opening range, as markets have become faster
-> calculation of A and C values should be based on statistical data from the preceding 50 or 100 days

My A and C values do not match Fish's values. But they include a dynamic component - the daily ATR's. 30 for longterm and 10 for short term. My daily bias is based on Fish's Numberline concept acc. 12 Days NL and acc. 30 Days NL . Furthermore I use the Trade and Facde concept - given by the PDF file (Link above).

regards ines

PS: My original member name is banned - maybe because the empty postings. That 5 Posting rule is useless.

Reply With Quote
  #18 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102


MissEuroCent View Post
My A and C values do not match Fish's values. But they include a dynamic component - the daily ATR's. 30 for longterm and 10 for short term. My daily bias is based on Fish's Numberline concept acc. 12 Days NL and acc. 30 Days NL . Furthermore I use the Trade and Facde concept - given by the PDF file (Link above).

regards ines

PS: My original member name is banned - maybe because the empty postings. That 5 Posting rule is useless.

Modifying the Concept

It makes certainly sense to include a dynamic component, and the daily range or average true range is a point to start with. 10 and 30 days should work, except for the beginning of the year, when the 10 day ATR may just reflect the low volatility of the holiday season.

My plan is to modify the indicator (as posted above) to use different concepts for A and C levels.

-> one using implied volatility (this is only possible for instruments where there is a volatility index readily available, which is calculated from options
-> one using the noise bands

The noise bands should typically represent a fraction of the ATR. You mentioned that you use 10% for the A levels and 15% for the C levels. But your calculations start from the edge of the opening range, whereas the noise bands start right away from the open.

Open Questions

(1) One thing I did not get so far: Why are the C levels generally further away from the opening range than the A levels? Does anybody know, what triggers the different cases

(a) A level smaller than C level: for example 6E, 6B and most instruments
(b) A level equal to C level : for example 6A, 6C, NG
(c) A level larger than C level: for example ES, FESX, NQ, FGBL, GC, ZW

(2) If I look at the A levels I want to find out the threshold for a failed breakout. So in terms of expectancy there should be a local extremum maximizing the potential outcome until the exit of the trade, which could be anything between the high (long setup) respectively low (short setup) and the close.

This can be achieved by only using daily data, if the threshold is counted against the open instead of the opening range. Just find the mode of the value SUM(high/2 + close/2 - threshold - slippage) with high> threshold for all long setups, where the threshold is modified in 1-tick steps to get the distribution of results.

In a second step a filter could be applied, whether the breakout was supported by the pivot range, against the pivot range or breaking through the pivot range, where the pivot range stands for yesterday's value.

(3) A concept for the C level would be similar, just one would need to add a filter requiring a failed A. So again a problem of finding the mode of a cumulative distribution, but this time filtered, as only the failrues of A are selected. This cannot be done with daily data, as the daily data will not tell us, which of the two events, breakout to the upside or brealout to the downside took place first.

(4) I understand the number line concept, but have one question. Why should I look at anything 30 business days ago? This assumes a fixed cycle with a length of 6 weeks. But there is no such thing as a cycle with a fixed length. See charts attached.

Attached Thumbnails
Click image for larger version

Name:	CL 05-11 (Daily) 10_11_2010 - 13_04_2011.jpg
Views:	726
Size:	91.2 KB
ID:	36584   Click image for larger version

Name:	6E 06-11 (Daily) 08_11_2010 - 13_04_2011.jpg
Views:	549
Size:	94.5 KB
ID:	36587   Click image for larger version

Name:	FDAX 06-11 (Daily) 12_11_2010 - 13_04_2011.jpg
Views:	428
Size:	90.3 KB
ID:	36588  
Reply With Quote
Thanked by:
  #19 (permalink)
MissEuroCent
Innsbruck
 
Posts: 7 since Apr 2011
Thanks Given: 0
Thanks Received: 8

>>>>>(1) One thing I did not get so far: Why are the C levels generally further away from the opening range than the A levels? <<<<<<

According to Fish - there is no C level for shares.

Definition on page 240:
C values A specific number of ticks that is used to determine the distance to a Point C up (C ^) or a Point C down (C v). C values will vary depending upon the commodity or stock traded. In commodities, C values are different than A values (see A values). In stocks, A and C values are the same for each stock.


For all other instruments with C - there is just ONE A per day - let's say an intraday reversal occurs after an good A up - so the next valid ACD pattern would be the C down. If the session close is below C down - the macroindicator Numberline value would be -4.

As far as I remember the book and videos - A is never bigger than C.

Servus
Ines

Reply With Quote
  #20 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102



MissEuroCent View Post
As far as I remember the book and videos - A is never bigger than C.
Servus
Ines

This is not the case. You just need to check the examples in the book or look at the last values published by Mark Fisher. See examples above.

Reply With Quote
Thanked by:




Last Updated on December 18, 2022


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts