GREY1 P/L daily

[Grey1]

Grey1, Your EDS is one fire dude! Would it be ok to ask you to post the screen shot of the times you took the trades so we can review them and learn.

Thx

Yes gulum with one minor problem . IB after 10.30 PM closes its platform and gets ready for next day and all trades disappears . so i wont be able to post the trades tonight but will take a screen shot next time I trade. No problems what so ever. it only takes me few seconds

grey1

 

[Grey1]

Thanks for that Grey1. I really want to try and understand properly. Can I paint an example picture to see if I am starting to grasp it?....

So, using the 1% of account 'rule' you have described before, lets say we have $1000 to "go shopping".

We are trading on the 1,3 and 5 minute time frame analysis.

Lets say the 1 minute ATR of stock A is 10c
Lets say the 3 minute ATR of stock A is 30c
Lets say the 5 minute ATR of stock A is 50c

So, one can afford to $1000 divided by 50c = 2000 shares?

So, market is strong (+50?) and stock is strong (+ve from open?)

We then get (using your MACCI engine) OS in each of three time-frames (DOW or stock or both? still unsure here?).

We therefore buy 2000 shares of stock A.

However, if price continues to fall by another 10C then we sell one fifth (10c/50c) of the shares = sell 400 shares.

If price fall by another 20c (by total 30c) we sell another 800 shares.

If price falls by another 20c (total 50c fall) we sell remainder (800 shares).

Does this make sense, Grey1? Is my understanding getting close to this aspect of your teaching?

I hope you have time to reply and correct me where necessary.

Thanks as always,

Steve

That is not correct my man . You cannot dump the same number of shares, you need to dump the number of shares equivalent to the volatility as follow

Number of shares= $1000/10 = 10 000 shares for 1 min time frame
Number of shares = 1000/20 = 5000 shares for 3 min time frame
Number of shares = $ $1000/50 = 2000 shares for 5 min time frame

You get a confirmation signal in three time frames lets say BUY , BUY , BUY

You open a position of 10 000 stock X .. Trade goes against you on the 1 min time frame you then dump ( 10 000 -5000 )= 5000 shares of the stock X
Now the 3 min goes against you then dump ( 5000-2000) = 3000 shares of the stock X

Now the 5 min goes against you ,,,, You close the position . i.e you dump the rest

Now lets try another scenario,,

what if 1 min goes against you and you dumped 5000 and then stock moved in ur direction what next ? easy ,, we add what we dumped before to have 10 000 to be in full force.

This is dynamic risk analysis and can be coded in radar in two min .


Of course the $1000 shopping can be reduce to $500 or less


grey1

 

[leovirgo]

let say you have a signal confirmed in three time frames 1,3 ,5 as BUY , BUY , BUY .. you open the position based on ATR of the 1 min . The trade then goes against you in the 1 min time frame , you then reduce position size proportional to ( ATR 1 min - ATR3 min ). Now you realise that the 3 min time frame gone against you , you then reduce the pos size by ( ATR 3 min - ATR 5 min ) and finally your 5 min BUY signal goes against you which you must close ur position completely . This algorithm is the foundation of your risk management .

Grey1

One of the best advice !

For the inconsistent TF signals, such as Buy in 1min, Buy in 3min but neutral in 5min, should that buy signal be taken? As the trend is born in lower TF, is it a good signal?

Or Buy in 5min, Buy in 3min, but 1min neutral/bearish.. This one is a time frame top down signal. is this trade any better?

regards,..

 

[Grey1]

One of the best advice !

For the inconsistent TF signals, such as Buy in 1min, Buy in 3min but neutral in 5min, should that buy signal be taken? As the trend is born in lower TF, is it a good signal?

Or Buy in 5min, Buy in 3min, but 1min neutral/bearish.. This one is a time frame top down signal. is this trade any better?

regards,..

should the buy signal be taken ? YES as long as ur pos size is correct and according to the time frame you are having the confirmation .
If you understand this simple procedure your path to be a consistant and profitable trader is paved


grey1

 

[Trader333]

Hi Gray1

I have a very elementary question – what time frame should I use for VWAP?

I would assume it starts with 1 day, the day we trade, and starting from the very first Tick during pre-market.

I am aware of the formula used for Tradestation, provided by you.

But, I have noticed with my Realtick (RT) platform that the most accurate VWAP prices only occur when I use a setting of 1 Day, using a Tick Chart (any number to ticks), set from Pre-Market. I’ve read all of the threads in this BB authored by you, but could not find any specific TF, I could be wrong.

In RT I found discrepancies in the 1 Min, 5 Min, Daily, etc. by larger & larger errors as the time frame (TF) was increased. I’ve tested from 10 to 5000 ticks, on my tick chart & all provided the exact same price.

Also, if I'm having this problem, then what are the MM's & large Brokerage houses using?

Thanks!

It shouldnt make much difference which timeframe you use as the readings should be the same for a 5 minute charts as that of a 1 min which is what I use.

Paul

 

[LevII]

should the buy signal be taken ? YES as long as ur pos size is correct and according to the time frame you are having the confirmation .
If you understand this simple procedure your path to be a consistant and profitable trader is paved

grey1

Good morning Grey1. It may help everyone understand more clearly if the question of multiple time frame confirmation could be considered with a few examples.

Taking LONG signals only (opposite for shorts) in 3 time frames 1, 3 & 5 minutes. Assuming the risk profile is OK and neither the market nor the stock are exhausted, are the following (with correct position sizing) valid signals?

1) long(1), long(3), long(5)
2) long(1), long(3), neutral(5)
3) long(1), long(3), short(5)
4) long(1), neutral(3), long(5)
5) long(1), short(3), long(5)
6) neutral(1), long(3), long(5)
7) short(1), long(3), long(5)

LII

PS, EDS seems to be working well - another winner from the G1 stable?

 

[LevII]

Grey1,

I have question on position sizing if you don't mind.

It is often the case in my approach to multiple time frame analysis that at the time a signal is produced it follows a low volatility consolidation in a stock which usually has much greater volatility. When the stock breaks out volatility returns to more normal levels.

Using ATR in such circumstances as the means of calculating position size results in a position size much larger than it ought to be according to the stock's volatility at the time the trade is running.

Other than calculating ATR over a number of different time periods or taking a stab at what it ought to be do you have any suggestions?

LII

 

[Grey1]

Grey1,

I have question on position sizing if you don't mind.

It is often the case in my approach to multiple time frame analysis that at the time a signal is produced it follows a low volatility consolidation in a stock which usually has much greater volatility. When the stock breaks out volatility returns to more normal levels.

Using ATR in such circumstances as the means of calculating position size results in a position size much larger than it ought to be according to the stock's volatility at the time the trade is running.

Other than calculating ATR over a number of different time periods or taking a stab at what it ought to be do you have any suggestions?

LII

LEV

As soon as I eliminate few technical issues from the EDS I will give it to the mighty 4.. Yourself , Ian , Blue and julian which is 80% more mechanical than the original VWAP engine .The technical issues are problem with TS and Radar as apparently they handle data differently .

As far as Volatility is concerned , there are other ways of controling the risk profile but they are resources hungry and donot justify their use. Implied volatilty is beter than historical but do we really want to make things even more slow ?

I think the way ahead is to turn every thing into a black box including the execution , then we all have a cash machine .

Grey1

 

[ianp]

LEV

As soon as I eliminate few technical issues from the EDS I will give it to the mighty 4.. Yourself , Ian , Blue and julian which is 80% more mechanical than the original

Hi grey1,

thanks very much for that


Just kidding I have never been known as mighty........ well not when it comes to trading.


I have been looking at implied volatility I have put the setting on 12 period on 1 min chart
as my maximum stop is that sufficient .

Thanks
ian

 

[evostik]

That is not correct my man . You cannot dump the same number of shares, you need to dump the number of shares equivalent to the volatility as follow

Number of shares= $1000/10 = 10 000 shares for 1 min time frame
Number of shares = 1000/20 = 5000 shares for 3 min time frame
Number of shares = $ $1000/50 = 2000 shares for 5 min time frame

You get a confirmation signal in three time frames lets say BUY , BUY , BUY

You open a position of 10 000 stock X .. Trade goes against you on the 1 min time frame you then dump ( 10 000 -5000 )= 5000 shares of the stock X
Now the 3 min goes against you then dump ( 5000-2000) = 3000 shares of the stock X

Now the 5 min goes against you ,,,, You close the position . i.e you dump the rest

Now lets try another scenario,,

what if 1 min goes against you and you dumped 5000 and then stock moved in ur direction what next ? easy ,, we add what we dumped before to have 10 000 to be in full force.

This is dynamic risk analysis and can be coded in radar in two min .


Of course the $1000 shopping can be reduce to $500 or less


grey1

Thanks Grey1 for the clarification but the reason I feel I am still missing a key point is that in the scenario you painted above, assuming the trade goes against you all the way, the total loss will be $2100 and not the $1000 we said we had to 'go shopping':

1st - closing 5000 shares at 10c loss = $500
2nd - closing 3000 shares at 20c loss = $600
3rd - closing 2000 shares at 50c loss = $1000

Sorry to be a pain but this aspect is clearly a very important one to get right - you regularly make the point that this can be the diference between success and failure.

Thanks Again,

Steve

 

[Grey1]

Thanks Grey1 for the clarification but the reason I feel I am still missing a key point is that in the scenario you painted above, assuming the trade goes against you all the way, the total loss will be $2100 and not the $1000 we said we had to 'go shopping':

1st - closing 5000 shares at 10c loss = $500
2nd - closing 3000 shares at 20c loss = $600
3rd - closing 2000 shares at 50c loss = $1000

Sorry to be a pain but this aspect is clearly a very important one to get right - you regularly make the point that this can be the diference between success and failure.

Thanks Again,

Steve

Steve

You are assuming you are stop lossing @ 10,20 and 50 Cents. This is not correct. Your stop loss is not an ATR based, It has other technical reasons. MY stop loss comes from Darvas which is part of the VWAP engine. YOU ONLY USE ATR TO ARRIVE @ CORRECT POSTION SIZING ..

Is that clear sir ?

thanks

grey1

 

[evostik]

Steve

You are assuming you are stop lossing @ 10,20 and 50 Cents. This is not correct. Your stop loss is not an ATR based, It has other technical reasons. MY stop loss comes from Darvas which is part of the VWAP engine. YOU ONLY USE ATR TO ARRIVE @ CORRECT POSTION SIZING ..

Is that clear sir ?

thanks

grey1

Grey1.

Your assumption about 'my assumption' is correct - I think the 'light' has just started to come on!!!!

You look at your adaptation of Darvas boxes in each of the three timeframes to arrive at the three stoploss levels - right?

Thank you,

Steve

 

[Grey1]

Grey1.

Your assumption about 'my assumption' is correct - I think the 'light' has just started to come on!!!!

You look at your adaptation of Darvas boxes in each of the three timeframes to arrive at the three stoploss levels - right?

Thank you,

Steve

yes sir

 

[Grey1]

TWO EDS signal after which the screen froze.

up $ 961.12 I am going also to leave the Trades so you guys can see the rational behind the trade.

Donot forget EXPD was down $6 yesterday

 

[Nastrader]

MAS,

The most accurate way of measuring VWAP is tick data from open to close excluding the pre market as this wont be affecting the long term volume profile. There is a variation in vwap reading on many platforms such as real tick and bloomberg for variety of boring reasons.

The main thing is MPD bands and how to use them .

Grey1

I understand now, since MM's/Institutions don't pay attention to pre-market data. Yes, Tick Data gave me the most accurate VWAP. I'll take some VWAP data from different TF's & report them later today., for those that are interested, & the data will be pertinent to RealTick.
Thanks Gray1

 

[gulam]

TWO EDS signal after which the screen froze.

up $ 961.12 I am going also to leave the Trades so you guys can see the rational behind the trade.

Donot forget EXPD was down $6 yesterday

Grey 1, dude the JOYG trade was incredible, but why did you get out so early even before the pull back? Any reason for that? Or is the EDS system designed to look for only 20 cents profit?

 

[Nastrader]

NAS,

The most accurate way of measuring VWAP is tick data from open to close excluding the pre market as this wont be affecting the long term volume profile. There is a variation in vwap reading on many platforms such as real tick and bloomberg for variety of boring reasons.

The main thing is MPD bands and how to use them .

Grey1

With your verification of better accuracy using Tick data, I used ADBE as an example (in an attached file, using 50 to 10,000 Ticks & 1 Min to 10 Mins) to list the data in different time frames, using Realtick. Note, the VWAP data may be different with the use of other platforms. The end result was that using 1 Min Charts compared to 500 Tick Data, & both set to 1 Day provided VWAP data within 0.03 C of each other. But prior to approximately 9:45 am (ET), the data had larger differences (0.07 C) due to the gapping stock this morning (yes VWAP should be used during consolidation - Gapping Stk should be the worst case for data comparison). Next, I need to find a way to apply the MPD bands in Realtick.

Thanks Again Gray1

 

[Nastrader]

It shouldnt make much difference which timeframe you use as the readings should be the same for a 5 minute charts as that of a 1 min which is what I use.

Paul

Paul, I just did a mini study on VWAP TF's. See my post # 217 today, I've also attached the same excel attachment below). I found the difference between 500 Tick data & 1 Min data to be 0.03 C. But the difference between 1 & 5 Mins was 0.27 C, using Realtick. So with Tick data being the most accurate, there is a possibility that with another 0.03 C error, the error in data could be as much as 0.30 C using 5 mins charts.. This error could be much less using Tradestation or eSignal, and others (I'm not sure).

Now eSignal may start it's data from the 1st tick (pre-market) - I think, which to me means that even though the data is in Mins., the root data could be in Ticks, thus making your 5 Min. data pretty accurate. I'm not sure about Tradestation. The only way is to verify 5 Min. data with 100 or 500 Tick Data for comparisons, which you've probably done.

(Edited) You are right, I just found an article for eSignal, where it states that after the 1st 1/2 hour VWAP is usually less than 2 cents between 1 and 5 min charts, I've attached the article.

Thanks for your reply

 

[leovirgo]

Today I bought PG as it was the main driver for Dow's move up. Went long when it retested VWAP and exited when exhaustion engine said Exit Long.

When Dow was exhausted, shortsold NIHD as it was above upper MPD. This stock has been in down trend since 27 july. Covered just above VWAP.

I should say Muchas Gracias Grey111111111111111111111

 

[Grey1]

Today I bought PG as it was the main driver for Dow's move up. Went long when it retested VWAP and exited when exhaustion engine said Exit Long.

When Dow was exhausted, shortsold NIHD as it was above upper MPD. This stock has been in down trend since 27 july. Covered just above VWAP.

I should say Muchas Gracias Grey111111111111111111111

heheh leo now you are learning to beat me in my French lol

Up $ 1416.89 Muchass Gracias Mr market catch us if you can but you cannotttttttttt