Head2k

So far it doesn't look like rejectiion in the zone around 60. Also, unfortunatelly, the blue uptrend channel is pretty creepy and, at least for me, reversals of creepy trends are more difficult to take. Anyway, if price is rejected form the zone, then I might short a retest. Then I will look for a long off the zone above 32. Or I might take a a continuation trade after a break-in into the previous range, that is after breaking the entire ~60 zone plus a bit more. If we go up, then I will look for a reversal off the zone around 85 or break-in into the range above. I marked also the 73 midpoint, but I won't do anything there since I expect price to breach the largest supply line.

Why to pay for it? If you enter on a pullback into a trend, then you in fact enter in an intra-trend range. Now if the trend is reversing, then you can expect at least a target implied by this small range to be hit. Same with reversals. If you wait for sort of a climax at S/R and then you enter on a re-test, you are in fact entering within a small range defined by the bounce between the climax and the re-test. Then you can expect a target implied by this small range to be hit, even if the trend eventually continues. I don't use these small targets for scale-outs, but for stop reduction. The only question which remains to be answered is what the target is within a range. And I guess you don't need to pay $50 to find it out.

I liked that part about different "trend degrees", Thales. While I am still a beginner, I focus a lot on knowing exactly what the situation is on several degrees (I don't use a term "degree", but rather "wave scale", "trend scale" or just "scale"). I recognize at least 4 scales important for every trade: Large Trend Scale - framework for the ranges I trade Trading Range Scale - the range I want to trade Intra-Range Trend Scale - the trend I seek a reversal of Entry Range Scale - a range within the Intra-Range Trend which I use for entry timing As you can see I am basically a range trader and I do mix scales or degrees. I use stop proportional to the Entry Range, but target proportional for Trading Range. But I have defined setups which allow for that. At target I exit only a half and I trail the rest. The scale used for trailing is the Intra-Range Trend. If the trading range is broken, then I can take a continuation trade if my criteria are met, and in such a case I have no target and I just trail. A trend trader trading just breakouts uses more scales for making trading decisions if he waits for a breakout of a more significant level than just a high or low of a trend which he plans to use for trailing his stop. OTOH, if he buys a new high in a certain trend and places his stop under the last swing low of the same trend, then he uses just one degree. But since the target is always uncertain once a range is broken, trailing a BO trade seems the best option to me. For this purpose, I consider the trend scale used for trailing to start in the BO. So I don't allow for hoping that I jumped into something bigger.

I've never said that I am a pro trader. DbPhoenix is a pro trader, I am just a beginner like you. As for the rest of your post, well, it's your journey. My point, and the point of On Virgins and The Tooth Fairy, was just to warn you from looking for reasons where none exist. Price certainly had a reason to turn at that price level. But did this reason have anything to do with 40 periods EMA displaced for 17 periods on a 5 minute chart? I mean, can you even imagine why would it? I can't.

Just coincidence. What's important is the stair-step based S/R, your white horizontal line. You could have put the whole indicator away and leave just the white line... and see "what happens". I suggest you to read this post from DbPhoenix's blog: On Virgins and The Tooth Fairy - Traders Laboratory - Professional Traders Community

Here is NYSE TICK from DTN IQfeed. Time is EST. 16:00 - 16:30 GMT is marked with green and red marker.

If you want me to compare two indicators, then you should use them on the same chart. I don't know what you mean exactly. I see that in the AmiBroker indicator you obviously use two plot functions. One with styleArea and another with styleCloud. StyleArea is a widened histogram, so the colored columns have middle at the middle (vertical axis) of a bar. StyleCloud changes color at the middle of a bar. That's why the the color changes are shifted by a half of the inter-bar spacing. If you mean something else then specify it better and use the same chart for muticharts and AB indicator. _____________________________ IIf( EMAactual < Ref( EMAactual, -1) , colorYellow, colorBlue); ______________________________ ColorBuble = IIf( Condition1, Color1, IIf( Condition2, Color2, IIf( Condition3, Color3, IIf( Condition4, Color4, 0 )))) ; Here, if no condition is met then ColorBubble will be zero.

Well, to enlighten you a bit more, knowing how to program isn't the same as knowing how to trade. To take it even further, knowing how to program a "flying carpet" is good only for realizing that a flying carpet is useless. I am not sure what a trader you are and what your experience is, but I think that time for programing comes, if ever, only when you thoroughly understand price action and you want to (partially) automate your discretionary strategy. But to start a trading career with flying carpets and EMA coulds is a mistake. Everything IMHO, of course.

To the 1st question: If you want the color to change in a discrete manner you can list the conditions just like you did. If you want the color change continuously over the color spectrum, or if you can join your descrete conditions into a mathematical function, then you can use this function to define the color. Then you can use ColorRGB function (find it in AFL help) to change the color. So you need to convert your conditions into mathematical functions which output values between 0 and 255 for red, green and blue. To the 2nd question: The color change doesn't happen late. If you zoom in you can see that the two averages cross usually somewhere in between of two bars. The color is defined by relation of the two averages at the last bar. Then the averages cross, but before the next bar is reached the color remains the same. Once the next bar is reached, the relation between the averages is re-examined and the color changes at that bar. AmiBroker can't change color somewhere between bars, only at a bar. To the 3rd question IIf( EMAactual < EMAactual[1] , colorYellow, colorBlue); This will always return colorBlue, because you compare EMAactual to EMAactual, that is two same numbers. So the expression in the condition is always false. I can't judge this. I am not a programmer and I never learned C. I've learned just AFL from AmiBroker help. When I was a 10 years old child I used to learn TurboPascal which was kind of an educational programing language, so I had an idea about what a loop is and so on. So the only thing I can say is that AFL is probably easy, because if one has some elementary knowledge of programing languages principles, then he/she can learn it very well only from the help included in AB.

Sorry, actually it is the other way around. Negative displacement moves the past data into the future and positive displacement moves the future data into the past. So with positive displacement you refer to future data which are not known in the actual moment, and that won't work in real time usage and it can screw up your back test. That's why I allowed only for negative displacement in the code.

Use Ref function for displacement. Plot( Ref( EMA( c, Periods ), Displacement), ... Negative displacement moves the average in the past, positive displacement in the future. Anyway, the whole code would be as follows: Plot(C,"Price",colorBlack,styleCandle); Periods = Param("Periods", 30, 2, 200, 1 ); Displacement = Param("Displacement", 15, 0, 150, 1 ); EMAactual = EMA( c, Periods ); EMAdisp = Ref( EMA( c, Periods ), -Displacement); Color = IIf( EMAactual > EMAdisp, colorGreen, colorRed); PlotOHLC( EMAdisp , EMAactual, EMAdisp, EMAactual, "Cloud", Color, styleCloud); I admire people who can trade off flying carpets.

Since Db doesn't seem to be active, I will try to answer. Db always said that TICK divergence is only the icing, not the cake. That means that you shouldn't look for TD mechanically. The market tells a story, and TICK is only a part of this story. As for the confirmation, I take exhausion and current trend in account. Taking exhauston in account means that I look for climactic action at S/R and a lack of follow-through then. Taking the current trend in account means that I expect a higher high after a higher low. And I am ready to enter off a lower high after a lower low. And if I don't get such entries I am ready for a hinge. The main thing I am looking for is effort vs. result and acceptance vs. rejection. TICK is only supplementary information on micro-scale. That means I use the help of TICK only to guess a local turning point. But before I even decide to look at the TD I must be convinced that this potential local turning point has a greater meaning than just "local", regarding both the level (S/R) and the setup (exhaustion andf / or trend reversal pattern). Anyway, if and when Db is active again, he will probably provide a better answer.

I guess I am missing the point here. DR = HHV ( H, Periods ); generates an array called DR which, for every bar (= array element), contains the highest high in "periods" bars before and including that bar. Then you take that DR array and redefine (=rewrite) it as follows DR = IIf ( PDR != High AND DR < PDR AND PDR != 0, PDR, 0 ); The DR array used inside the IIf condition is the original DR array defined by the HHV function. But by the IIf function you then re-write it and assign either PDR or 0 to its elements depending on the condition. That's what this code does. But perhaps it is not what you want it to do.

Well, that is more difficult for me because I don't know EL. But according to info you posted above HHV returns oExtremeVal of Price in Length periods with HiLo set to 1. LLV returns oExtremeVal of Price in Length periods with HiLo set to -1. HHVBARS and LLVBARS return oExtremeBar, I believe. Because oExtremeBar outputs the number of bars ago the extreme value occurred and HHVBARS and LLVBARS outputs how many bars occurred since the extreme was reached, which is the same. Just notice that HHV, LLV, HHVBARS and LLVBARS functions return arrays. So your line EV = Extremes( H , Periods , 1 , DR , EB ) ; would be in AFL as follows: DR = HHV ( H, Periods ); EB = HHVBARS ( H, Periods );

The exact transcription of your EL code would be DR = IIf (PDR != High AND DR < PDR, IIf ( PDR != 0, PDR, 0), 0 ); That is one IIf function inside another one. The only difference is that in IIf function "else" must be defined and I set DR to zero in "else" case. In another words, I believe your EL code doesn't say what DR is if the conditions aren't met. In this AFL code it says that DR will be set to zero then. Anyway, I guess the code would be more efficient like this: EL: if DR <> H AND DR < PDR AND PDR <> 0 then DR = PDR ; AFL: DR=IIf (PDR != High AND DR < PDR AND PDR != 0, PDR, 0); In another words, you can put all the expressions into one condition. Now to your previous post and array functions vs. loops. I think in most cases you won't recognize the difference in speed. So if you happen to understand loops better than array functions, you can of course use loops.

I don't know EL, I never used it. But your transcription into AFL seems OK. AFL is also easy, once you get the difference between arrays and nubmers. And i would recommend using preferably array functions, as they are usually faster than loop code. As for the Extremes function, according to documentation you get 2 values from it. 1. The value of the extreme 2. Where the extreme occured In AFL you need two functions for that. HHV for the value and HHVBARS for the number of bars since the high occurred. LLV and LLVBARS for lows. Browse AmiBroker Help -> AFL chapter -> Categorized list of AFL functions -> Lowest/Highest

In AFL coding it is essential to distinguish between a number and array. High is an array of all highs in the chart. if operator works only for comparing numbers and returns also a number (boolean). On the other hand, IIf operates with arrays and returns also an array. It is like running if operation for every array element. ODR = IIf (DR != High, PDR, 0); This line says: Take every array element in DR array and compare it to adequate (i.e. with the same index) element of array of Highs. If the compared array elements are different (!=), then the adeqate element of ODR array (i.e. an element with the same index as the compared elements of DR and High arrays) shall be equal to an adegate element of PDR array. Else it will be zero. Obviously ODR will be an array after this operation. EDIT: Just to show the difference, I will write how to use if operator to get the same result as the line above gives. for (i = 0; i < barcount; i++) { if (DR != High) ODR = PDR; else ODR = 0; }

Interesting. Price data seem to be the same but the indicator differs. How is the indicator calculated? Or is it provided directly by the data providers? Do you backfill tick-by-tick data? Does backfill contain information about bid and ask on tick-by-tick basis? Maybe the indicator can be calculated properly only on real-time data. I am just guessing, of course.

AmiBroker is just a charting and testing software. You need to plug in data from some provider or import them from ASCII file (like csv for example). Check AmiBroker website for compatible data providers. I am sorry but I can't provide info on other programs, since I am a beginner and AmiBroker was my first choice. I don't have much other experience. I only tried NinjaTrader but I didn't go deep enough.

i use AmiBroker for charting and it has a very transparent and comprehensible data management. You can create and name different databases and choose their folder save location, then the data are stored in special files within the database folder. You can create new databases from .csv (or similar) files, you can export data into .csv (though you need to download a tool for the export). Then you can replay the data. In replay you set a step and speed. The step can be as low as the base time interval of your database, but not lower than 1 second (because time stamp in AmiBroker has 1 second resolution, even for a tick-by-tick database).

Discipline is sticking to your plan and strategy. In order to be able to define discipline, you must have a thoroughly defined strategy which allows for consistent application. Then, discipline means doing what you planned to do and not doing what you didn't plan for. The better, the more precisely defined strategy you have, i.e. the better you know what you should do, the easier is to maintain discipline, i.e. the easier is actually doing it. If you are not really sure what you are looking for in order to enter or exit a trade, then it is imposible to evaluate discipline. I thanked collegetrader for his posts because I think many aspiring traders seek problems in discipline or emotions while their problem lies in the fact that they don't know what exatly they are looking for, i.e. in their strategy and/or tactics. They seek problems in discipline while discipline can't be even defined in their case. Another factor is confidence in what you are doing. Some people might take some mechanical strategy found on the internet, a strategy which is objectively defined, but they fail in sticking to it. Then the problem is not that these people wouldn't know what they should look for, but that they don't know why they should look for it. They don't understand the strategy and without understanding they have no trust in it. It is much easier to maintain discipline if you know why you do what you do (and why you don't do what you don't). It is much easier to trust something which you developed and fine-tuned yourself, because then you know exactly all the whys. And of course, for confidence it is better if you base your strategy on some sound logic than on pure statisctics and/or back-test optimization. Because the logic is the piece you understand and trust, not the statistics. The statistics serve to test and confirm the logic. They are a tool to build your trust in the logic. But they can't become the confidence-founding logic themselves. So my advice on how to "get" discipline is to develop a precisely defined strategy which allows for consistent, reproducible application in order to be able to even define the discipline.

Out of curiosity I googled "historical tick data" and several companies popped up. The first one in the list offers CL since 1987.

Regarding bar intervals, or better to say different scales used for a trading decision, I use three of them. The largest scale serves for determining zones of interest, that is S/R. The intermediate scale serves for determining trend between the zones of interest and its changes. And the smallest scale serves for observing behavior at the very test where I am looking for the entry. To be concrete, for NQ I determine S/R on 10k CVB, then I use 15s - 1m to watch trend and swings, and finally I watch 1 tick chart to find triggers (that is for a decision whether to enter on a particular test). As for trading live vs. sim, I believe one should learn everything he can on a sim. The reason why your behavior changes when you swith to live is that you are still not sure enough about what you are doing. At least that's my case. On a sim you can trade while you are not 100% sure about all your actions and it even doesn't look strange to you. But when you come to the live acount, every your uncertainity gets magnified, be it about entries or exits. If you are not sure what your plan is it forces you to work on a more detailed one (while on a sim you might settle for using "intuition" in that particular case). If you know what your plan is but you break it, then you don't trust your plan enough, perhaps because you haven't tested it enough or because you think it is not good enough. In another words, switching to live account stresses all the problems you have with your strategy, problems which were not so obvious when you traded on a sim. When you identify these problems, then determine their origin and reason and switch back to sim to work on them. Then repeat. At least that's what I am currently trying to do.

Non-Farm Employment Change forecast -119k vs. actual -11k is a difference which sends us back to the yesterday's range. R 1806, 1812-14 S 1792, 1786-8, 1776