Jonbig04's Log

[jonbig04 10]

Thanks to all of you. I was up pretty late last night working on step 2. I don't know why I went out of order, but I quickly realized I wasn't going anywhere until I get step one taken care of.

 

The problem I'm trying to get around right now is simply recognizing trend. When the market opens how do I determine the trend (of course there are 90 ways to do this) reliably and quickly. For example the NQ rose a little on open then dropped 20 points and rose 20 points and now appears to be in a downtrend. What I need is a specific way to measure the prevailing daily trend. I think I've gotten past my momentum problem, but of course that doesn't help with showing which way the trend is likely to go for the day. The NQ looks to be in a downtrend (with slowing momentum) but I need to know exactly what constitutes a trend and opposed to just what it looks like.

 

I will look into that gann fann as well as the 3 higher lows (highs) and see if they have been reliable in the past.

 

 

Another thing I'm unsure of is whether to post the specifics of what I'm doing on here. If I knew if was just you all (the people tha have been helping me) I would mind, but I'm a little leary about posting exactly what I'm doing for the whole world to see. I'm not sure why exactly.

[BlowFish 10]

Some ideas for quickly determining trend. Probably the best....is price making HH HL? or LH LL? That can quickly tell you up down or sideways.

 

A quick and dirty but still quite handy way is to use yesterdays 'floor pivot' (H+L+C)/3 or yesterdays mid point or yesterdays POC or yesterdays close or todays open or or or or. Are we above or below? This can be used with other bars than daily too.... Hour bars or 240 minute bars are perfectly fine...whatever you llike for the time frame you are interested in.

 

Of course there are good old moving averages. Not my cup of tea but might suit you. You could eve use tradeguiders 'diamond system' that is simply a 5 period SMA and was reverse engineered here. (The diamonds are coloured depending on slope and which side of the average the bar closes on).

 

The problem is likely to be eliminating methods rather than finding them.

[jonbig04 10]

Yes there are so many. I'm going so start with the HH,HL,LH,LL. Thanks

[daedalus 10]

Just want to say keep at it! I am 22 and an NQ trader and i've had a rough path to get where I am. I haven't made it, but I think i've got an idea that can work and i'm working at it day after day and starting to see a glimmer of hope at the end of the tunnel.

 

KEEP AT IT. I can't tell you how many thousands of hours i've spent going through charts, backtesting, trying new ideas, just keep doing it. ITS WORTH IT.

 

I had my "EUREKA" moment last week and I can tell you its worth it in the end.

 

Cheers mate!

[jonbig04 10]

Thanks bro, how long have you been at it?

[daedalus 10]

Just over 2 years in futures... Started with the ES, now trading the NQ soley.

 

The ES just backs and fills way to much for my style of trading. The NQ moves and if its gonna go it runs for a bit straight line which allows me to quickly reduce risk and lock in profits in. The same setups in the ES would tend to backfil and stop you out too quickly, etc.

 

NQ trades nice and smooth most of the time.

 

So whats the system you're working on here? What problems are you encountering?

 

I'd like to try and help!

[jonbig04 10]

Ok I think I am getting somewhere. When I say getting somewhere I don't mean I've necessarily figured out the problem or solved it, just that I'm one step closer.

 

I think (I'm still testing, I wish there was some kind of computerized way to do this) I've found a way to identify the trend direction, and its strength. This goes a long way towards helping me with 1 and 2. firstly because I of course have to know the direction of the trend, secondly I need to knows its strength so I can assess whether to try to get in with it. This is especially important when trying to trade retracements as hey can of course turn into reversals a good percentage of the time.

 

Not to get too specific, I'm first watching price only. When price does X (a certain number of times) I then look at a certain momentum indicator. If all is well I wait for a retracement (yet to be defined) and enter hopefully near the bottom (yet to be defined) if and only if the momentum is still on my side. If the probability of a reversal is too high I don't enter. I wait for the reversal or more momentum (from PA and the indicator) to show me the trend is solid for the time being.

[jonbig04 10]

Right now I'm in the process of studying the actual retracement. One thing I'm doing (not the only thing) is going through chart after chart and numbering with a pen every retracement (and reversal) I see. I then can assign specific attributes to each one and hopefully find some common denominators (and hopefully attributes exclusive to reversals). Now I'm focusing on the retraced amount as a percentage from the trend. So basically I'm looking at these 50 or so RETs (now shorthand for retracement) and calculating the difference between the point it left the trend and the point at which is started back towards trend; then converting that into a percent as to make apples to apples comparisons between RETs, reversals and just small fluctuations in price. This will help me do a couple things hopefully. If a RET has gone way beyond the typical percentage it may actually be a reversal, and I can decided which fluctuations are in fact RETs I can profit on or if they are too small. I havent done this much math since....ever. haha.

[sevensa 10]

Another way of looking at the retracements is with Fibonacci ratios. Many consider it a trend reversal once it go past 62% where 38% and 50% is considered retracements. I don't really use it, but you might find it useful and saving you some time.

[jonbig04 10]

Thanks for the info. I've seen them before and essentially it is what I'm doing, I just would rather determine all the information myself. May sound strange but I don't want to attempt any shortcuts and definitely don't want to become indicator dependent. If I detect a pattern myself and find out later that an indicator does the same thing then I'm sure I will use it, but for now I would rather do most (I am using one indicator for momentum) of the work myself. If that makes sense. thanks.

[daedalus 10]

I trade retracements into a prevailing bigger trend. What platform are you trading with?

[jonbig04 10]

I trade retracements into a prevailing bigger trend. What platform are you trading with?

 

Hey, sorry I didn't see your earlier post.

 

I'm using NT w/zenfire if that's what you mean?

[daedalus 10]

Ok, well i rely heavily (and solely) on the ValueChart indicator to tell me when go situations prevail to take retracement trades.

 

I found a NinjaTrader version on this website: http://www.ninjatrader-support.com/vb/showthread.php?p=39817

 

Post #8, but i think you have to register. Throw that up on your charts... i think it just might revolutionize your trading. Its the best indicator i've ever come across and is VERY powerful for retracement trading.

[jonbig04 10]

Ok, well i rely heavily (and solely) on the ValueChart indicator to tell me when go situations prevail to take retracement trades.

 

I found a NinjaTrader version on this website: http://www.ninjatrader-support.com/vb/showthread.php?p=39817

 

Post #8, but i think you have to register. Throw that up on your charts... i think it just might revolutionize your trading. Its the best indicator i've ever come across and is VERY powerful for retracement trading.

 

 

Awesome, I will check it out. Thanks for your help.

[jonbig04 10]

I'm finding a lot of interesting info. It's slow work, but the results have been pretty interesting.

 

Basically I'm finding a few things that happen 50%-60% of the time before, or during a RET. I haven't yet cross checked them together, but I soon will. I guess I'm just glad that there seems to me a method to all the madness when you really sit down, concentrate, and manually and methodically sift through all the info that's there in the PA/V for you to see. Maybe I'm getting somewhere.

 

Maybe not. haha

 

 

Is there someway I can create my own indicator without being a computer super genius...based on VOL and PA. I.E. price does this with vol doing this along with this etc and indicator lights up...something like that?

[IT-Trader 10]

Firstly I've been reading your blog with much interest. The frustration of banging your head against a wall for months on end is something I know well and am still working through. So much is written about the psycological aspects but it seems to get forgotton that the best discipline in the world won't help you win if you have a negative expectancy system.

Developing and defining that system explicitly is a huge challenge.

Anyway, this is your journal. I just wanted to say that you're not alone fighting this. On to the purpose of my post...

 

Is there someway I can create my own indicator without being a computer super genius...based on VOL and PA. I.E. price does this with vol doing this along with this etc and indicator lights up...something like that?

 

In short, yes but.....

It depends on exactly what you want to look for.

The work breaks into two parts, the first is writing the 'logic' out to define exactly what you want to look for. Then the second part is coding it up.

 

Lets pick something simple like a volatility breakout.

The 'logic' is say

- find the average range over the last 10 bars

- if this bars range is 2x average range then flag this bar as 'volatile'.

 

Then you have the fun of coding it! ...which depends on exactly which package you use etc.

 

You can also add criteria such as "...and closes on its high/low" or "...and on above average volume".

 

To start with you can use something like Sierra Charts so you can code this stuff up using Excel which makes life easy for relatively simple ideas. Unfortunately it quickly gets complicated, defining things we can easily see in 'cold computer terms' is often waaaay more difficult that it first seems.

I've been doing just this kind of thing lately (the simple stuff) to highlight bars that meet certain critera. I'm far from an expert in indicator writing but I'm happy to help if I can.

 

ITT.

Edited August 16, 2008 by IT-Trader

[BlowFish 10]

I'm finding a lot of interesting info. It's slow work, but the results have been pretty interesting.

 

Basically I'm finding a few things that happen 50%-60% of the time before, or during a RET. I haven't yet cross checked them together, but I soon will. I guess I'm just glad that there seems to me a method to all the madness when you really sit down, concentrate, and manually and methodically sift through all the info that's there in the PA/V for you to see. Maybe I'm getting somewhere.

 

Maybe not. haha

 

 

Is there someway I can create my own indicator without being a computer super genius...based on VOL and PA. I.E. price does this with vol doing this along with this etc and indicator lights up...something like that?

 

Good work! You will probably find you have to back track a bit and needless to say will have other hurdles in your path but sounds like you are working purposefully towards a goal which is great!

 

There are a couple of volume indicators that do what you want. If you search here you will find one that works similar to tradeguiders. You might search for 'hawkeye volume indicator'. To be honest you will probably find you don't really need them - though a simple moving average of the volume might help visually. I think you use Ninja right? Those two might be TS only. Ninjatrader has a 'better volume indicator" by the guy over at e-mini watch. You can get a version at the NT website. Never used it but sounds like what you want.

 

The problem is unlikely to be identifying 'high' or 'low' or 'declining' volume but interpreting it in context. Be sure to ask yourself 'what do I want from this indicator' and 'is it providing this'.

 

Cheers,

Nick.

[jonbig04 10]

Firstly I've been reading your blog with much interest. The frustration of banging your head against a wall for months on end is something I know well and am still working through. So much is written about the psycological aspects but it seems to get forgotton that the best discipline in the world won't help you win if you have a negative expectancy system.

Developing and defining that system explicitly is a huge challenge.

Anyway, this is your journal. I just wanted to say that you're not alone fighting this. On to the purpose of my post...

 

 

 

In short, yes but.....

It depends on exactly what you want to look for.

The work breaks into two parts, the first is writing the 'logic' out to define exactly what you want to look for. Then the second part is coding it up.

 

Lets pick something simple like a volatility breakout.

The 'logic' is say

- find the average range over the last 10 bars

- if this bars range is 2x average range then flag this bar as 'volatile'.

 

Then you have the fun of coding it! ...which depends on exactly which package you use etc.

 

You can also add criteria such as "...and closes on its high/low" or "...and on above average volume".

 

To start with you can use something like Sierra Charts so you can code this stuff up using Excel which makes life easy for relatively simple ideas. Unfortunately it quickly gets complicated, defining things we can easily see in 'cold computer terms' is often waaaay more difficult that it first seems.

I've been doing just this kind of thing lately (the simple stuff) to highlight bars that meet certain critera. I'm far from an expert in indicator writing but I'm happy to help if I can.

 

ITT.

 

 

It seems like it would be complicated. I have not the faintest clue about writing, but it would be nice to let the computer do some of the recognizing, but I suppose it will be good for me to do it all manually first, i just hope everything doesn't slip away in the time it takes me to figure out where everything is at. I just discovered replay on ninja which is pretty badass. I'm sure I will be using that tomorrow. Thanks for the help.

[jonbig04 10]

Good work! You will probably find you have to back track a bit and needless to say will have other hurdles in your path but sounds like you are working purposefully towards a goal which is great!

 

There are a couple of volume indicators that do what you want. If you search here you will find one that works similar to tradeguiders. You might search for 'hawkeye volume indicator'. To be honest you will probably find you don't really need them - though a simple moving average of the volume might help visually. I think you use Ninja right? Those two might be TS only. Ninjatrader has a 'better volume indicator" by the guy over at e-mini watch. You can get a version at the NT website. Never used it but sounds like what you want.

 

The problem is unlikely to be identifying 'high' or 'low' or 'declining' volume but interpreting it in context. Be sure to ask yourself 'what do I want from this indicator' and 'is it providing this'.

 

Cheers,

Nick.

 

 

That makes a lot of sense. Really I was looking for a way to enter my own discoveries into the computer, creating my own indicators if that makes sense. Who knows though. I will check out the ones you mentioned.

[jonbig04 10]

Does T/A keep anyone else up at night. Its all thats in my head.

[smwinc 20]

With indicators, remember the overlying fact is that you DON'T need them. They are something which really are designed to service the account-cycling non-profitable trader.

 

That being said, they have a purpose and many a trader use specific tools for a purpose.

 

My opinion of indicators is that you use them to perform a calculation you understand and could do yourself, but it would be difficult and time-consuming.

 

I.e. In my trading I use the VWAP, PVP and Value Area frequently. It didn't "drastically change my Profit/Loss" but I like them. It quantifies certain aspects of my trading.

 

I get a little worried when I hear people recommend an indicator as being able to drastically change your trading - even if they aren't selling something.

 

Maybe I'm a little old school - but I think you should be able to make money consistently with a) market depth & b) a static chart. If your add-on indicators to a chart were to go down, you shouldn't feel like your suddenly standing in a crowd naked.

 

Your biggest advantage so far is your age. Just imagine if you spent 5 hours a day studying the price action of the NQ - 10 years later at the extremely old age of 30 (yes, dripping with sarcasm) after roughly 12,500 hours staring at the market depth you would be inhaling bids and exhaling offers, something like Neo out of the matrix. Do you really think you would still be needing a MACD and a Stochastic to be telling you when it's time to buy and sell? I don't think so. Really spend the time understanding how a market works and what is causing the smallest price movements, the tick by tick action. Everything else will come after that. Just focus on mastering the basics.

 

Trading is very simple, but damn hard. Be brilliant at the simple things.

[IT-Trader 10]

I just wanted to clarify a separation between indicators and 'highlighters' for want of a better term. (Tradestation calls them 'show_me'.)

 

I think Jon was asking for a 'highlighter'.

ie. 'Show me all instances where the range of a bar is greater than average'.

 

I don't class that as an indicator, its shorthand. Its letting the machine highlight something I'm looking for manually which speeds up the process.

That kind of stuff is very easy to code.

 

However I would agree that the use of RSI/Stochastics/MACD etc all come under the indicator banner and while inidcators can be an aid to some, taking a trade just because the 'wiggly line turned up' is not likely to be successful.

[metalhead 10]

That's a great start, because it shows you're actually thinking about it (better than most). I highly recommend you read The Phantom of the Pits. Don't let the price (free) fool you, you'd have a hard time beating the wisdom for any less . This may help you with position managment, helping you take an idea and make it as profitable as possible.

 

Push yourself to quantify your entries/exits. You will struggle when you add real money to the equation if you don't have something exact to fall back on.

 

do you have another link to that book? That one says "account suspended."

[jonbig04 10]

I agree 10,000%

 

I think "indicator" is just a loaded word. In my view an indicator just indicates something. As long as you understand why and how its doing what its doing, its just a time saver.

 

My opinion of indicators is that you use them to perform a calculation you understand and could do yourself, but it would be difficult and time-consuming.

 

 

This is my point of view exactly.

[jonbig04 10]

 

after roughly 12,500 hours staring at the market depth you would be inhaling bids and exhaling offers, something like Neo out of the matrix

 

 

 

 

I love this....