Head2k

Notice that DbPhoenix doesn't use Volume At Price much, but identifies the zones in a price chart (see e.g. this thread). For that purpose a chart constructed of bars dependent on amount of activity is more suitable than a time based chart. DbPhoenix uses Constant Volume Bar (CVB) charts. If volume is not available, you can as well use N-tick charts.

AFL file is just a text file. Simply rename it to .txt and open it in Notepad. Or you don't even need to rename it -- just double click the .afl file and when you are asked which program do you want to use to open the file, choose Notepad.

I haven't read all the lengthy posts in this thread, but let me state something. IMHO, a wrong mental attitude can -- and I believe it often does -- prevent a person from developing an edge. If someone trades and lives in state of fear of being wrong and is building illusions to deny this state, how effective he can be in developing a good strategy and executing properly whatever he has developed? And how can he know whether whatever he has developed is a true edge if his fears and illusions prevent him from executing it and even testing it properly?

It started to work properly again a few days ago. Thank you.

I cannot mark forums read. If I click User CP --> Quick Links --> Mark Forums Read, then I get this error message: Although "inform the administrator" in this message is clickable, if I fill the form and want to send it then I get the same error message again.

IMHO, the setup itself must contain all critera for entry. Or you may distinguish between a setup and a trigger within this setup, but again, the trigger must be defined. In other words, the decision when to take the setup and when to pass must be contained within the setup definition. If it is not, then the definition is not complete.

That's what I don't get. How can you develop something over a weekend and apply it next week? How do you know what the result will likely be?If your exits are also "developed" over a weekend and then just "applied", it is no wonder that you are not quite sure whether what you are doing is right. Perhaps MM's suggestion will work, but I think the problem is in insufficient testing. If there is not "test thoroughly on several months of data" between "develop" and "apply", then you cannot have enough confidence to stick to your plan like a machine. No walks can substitute the test, IMHO.

I don't think that it is so simple. As mentioned in another recently active thread, being a trader is a very lonely career. Teaching, in one form or another, can mean a way to socialize and to share the trader's thoughts or wisdom to raise followers, like when a master accepts apprentices to teach them his craft. The apprentice, however, must use his common sense to distinguish a real master from a swindler.

I use replay for testing and for pattern recognition training, because it is quite a difference between identifying a finished setup in a hindsight chart and identifying the setup as it unfolds. I think that replay, even at higher speed, is good for learning to identify the setup in "real time", as opposed to "hindsight". Replay can also tell you whether you were honest when you defined what is and what is not your setup in hindsight, that is whether you didn't overlook too many cases when your setup didn't work. On the other hand, I don't think that replay, particularly at higer speed, is good for trying to define a setup, if that's what you mean by learning pattern recognition. I think that a setup should be first defined after carefully studying static hindsight charts.

Well, I hesitated whether I should write this, but one of the first things I did when i decided to pursuit the trading career was trying to program an Inside Bar system. I can even say that the first half a year or even a year I generally mistook trading for programing. Now I wish I had invested that time better. IMHO, programing can serve to automate certain parts of your system or even the whole system, but only after you understand what you are doing and why. That is after you spend countless hours watching the market move and understand why and how it moves. Staring out programing this and that, tweaking and optimizing is a wasted time, IMHO. Anyway, I don't know in what phase you are as a trader and I speak only from my personal experience, so I don't claim to posses the one and only truth. But perhaps you can take it as food for thought.

That's tricky because you don't know what happened first - the break of the low or the break of the high. But it all depends on how you define your system. The conditions I wrote in the previous post don't consider an outside bar folowing the inside bar as a special case. So every time the high of the inside bar is broken the trade is triggered. You can write a condition for a stop in the same manner, so every outside bar after inside bar would mean entry and a full stop loss. Generally, the more special cases you want to incorporate, the more complicated code you wind up with.

illuminati, AmiBroker has a function called Inside() which contains your cond1 and cond2. But lets stick to the conditions you wrote. In AFL, Buy is an array. If that array contains TRUE value for a particular element (i.e. for a particular bar), then you buy in that bar. So if you write buy = cond1 and cond2 and cond3, it is wrong, because you would buy within the inside bar itself, not in the next bar. The correct buy condition would be Buy = Ref( cond1 AND cond2 AND cond3, -1 ); so you would buy on a bar which follows after the inside bar. Furthermore, you need to consider that the bullish inside bar -- that is an inside bar with close > open, if I understand your definition correctly -- doesn't need to get broken in the upward direction. And you don't want to buy in such a case. To incorporate this, you need to write Buy = Ref( cond1 AND cond2 AND cond3, -1 ) AND H > Ref( H, -1 ); which says:"I want to buy if the previous bar is a bullish inside bar and if the current bar makes high which is higher than the high of the previous bar (i.e. if the current bar breaks the high of the previous bar)." To get rid of your conditions and to write everything in one line, you can use Buy = Ref( Inside() AND C > O, -1 ) AND H > Ref( H, -1 ); Now it is important to realize that Buy array contains only the information about which conditions a bar must meet so you buy on that bar. It tells nothing about the price you buy for. The price you buy for can be set somewhere in settings within the Automatic Analisis window (I would need to have a look at where exactly), or you can do it directly in the code. If you set the buy price in the code, it overrides the settings in the AA window. In the code, you must use BuyPrice array to hold buy prices for every bar. So if you want to use stop limit with stop price 1 tick higher than the high of the previous bar and limit price 2 ticks higher than the high of the previous bar, you will write tick = 0.25; // for example BuyPrice = Ref( H, -1 ) + 2 * tick; which sets you buy price always two ticks above the previous High. (A general note: If the price defined in BuyPrice array doesn't lie in the bar which you buy at, AmiBroker uses the nearest price which does in that bar.) You can define the tick value directly in the code, like I did now, or you can define it for every symbol independently. To do so you must enter it in Tick Size field in View --> Symbol Information. Then you can refer to it in the code if you use TickSize keyword. Then you would write BuyPrice = Ref( H, -1 ) + 2 * TickSize; So, to sum it up so far, you need the following two lines to define your entry: Buy = Ref( Inside() AND C > O, -1 ) AND H > Ref( H, -1 ); BuyPrice = Ref( H, -1 ) + 2 * TickSize; assuming you defined tick size in Symbol Information window. _________________________________________ Re position sizing, look up SetPositionSize function in AmiBroker help. You can set size as fixed $ amount, fixed share (contract) amount, % of your equity, and perhaps even as something else, i can't remember exactly. ______________________________________ As for ApplyStop function, I would need to study in myself first. I don't use automatic strategies so I am no expert in using related functions. Or you can define your exits without ApplyStop. Just using Sell and SellPrice arrays. But it will get certainly complicated, because you will need to separate exiting at target(s), initial stop, trailing stop, etc. and define SellPrice appropriatelly to the exit used. You will probably use a lot of conditions (study IIF function). But the principle remains the same as for the buy signal.

Attila, do you put your stop at breakeven after some time / confirmation? If you don't, perhaps you should consider it. Then you wouldn't risk a loss any more (except for rare cases) and it is much easier to let the trade run then. It is even easier to let the trade run if you have already some realized profit. Perhaps you could scale-out at a nearer target and let the rest run. Examine how much points / money you have troubles returning.

I believe that by a "square curve" a parabolic function is meant. But I believe that Thales' MM doesn't result in a parabolic equity growth. Actually it is exponential-like curve composed of little linear segments Each linear segment represents one week and each following week is twice steeper than the previous one, which resembles 2^x, not x^2. Assuming he actually doubles his position after every week and he has perfectly constant gains per contract. As for Risk of Ruin, I believe it remains more or less constant for the whole race, if one doesn't change system and if one uses the same position to equity ratio for the whole race. Which seems to be the case. In normal conditions, I guess a trader would like to decrease his RoR as his capital grows, but this is a race so an aggressive MM is used. That is also the reason I wouldn't join such a race. I am a beginner and the primary goal for me is to preserve my capital, not to become a millionaire in half a year.

As you can see in the attachement, we are in a range from 2008 now. This range can be devided into halves, and ~2018 is the midpoint of the upper half. ~1997 and ~2040 seem more important, though.

I've been sim trading for quite some time. And I really think that until you show consistent profits on the sim there is no reason to start live. On the other hand, I undersand zdo and Thales, but i must admit that it took me quite long to come to this understanding. The problem is that on the sim you know you don't risk anything. And that can lead to building bad habits and/or it can prevent you from taking the trading seriously. If you had to start trading live, you wouldn't start until you had a perfectly defined and tested system. Assuming you got over the initial naivity stage. But with the sim there is no need for that. You can just try it and see. And that's what makes the sim counter-productive. If you are lazy or you just don't realize or don't want to realize what it actually takes to develop a trading plan, then the sim provides an alternative. You don't need a plan, you can trade ideas or concepts. But this is not the way how to develop a system, not even the way how to test anything. It is just a way to keep playing and a way to delay taking the trading seriously and doing what it takes to develop and test a system.

With 1R (1 tick range) chart you get almost the same precision calculated from much less data points than if you use 1 tick chart. I have observed that running computation on time charts or larger range charts is generally less precise than using CVB or tick charts. The reason is obvious. Bars on CVB and tick charts form in relation to activity, while in time bars the activity within the bar is much less likely to be evenly distributed. So if one wants to calculate developing value areas from less data points but still with high level of precision, he should use CVB (or tick) charts.

Sevensa is often rough, but perhaps it can be helpful (though I am not sure if that's his intention). If you honestly answer to yourself the questions he asks, you might get a better insight at what it is that you are looking for and why. And maybe make some conclusions. Or not. _________________ AFAIK, DbPhoenix started trading off 1 tick chart some time ago. But it is not the only chart he uses. He does his prep on CVB charts (finds S/R), then he watches 30 s or 1 min chart for overview of intraday moves and then he uses 1 tick to watch action in the areas of his interest. AFAIK he used 5 s chart before he started to use 1 tick. Unfortunatelly, DbPhoenix hasn't been active on TL for the last 3 months. But you can find his insights in his blog and in Wyckoff Forum. Regarding definition of PA trading, I think the core meaning is to base your trading decisions on watching demand / supply changes at key areas. You determine S/R and you watch traders' behavior when price gets there. How they push to get through, how hard sellers and buyers try and what they achieve. And your task is to determine the point of victory of one side, that is the moment when the other side is done, exhausted, and gives up. And to determine the key zones of S/R, a PA trader usually uses AMT (auction market theory). Or at least that's how I see it.

I haven't seen a laptop with two VGA connectors (but I am not an expert in this field), but it is not rare that laptops have one VGA and one HDMI connector. So if you have at least one monitor which allows for HDMI input, then it shouldn't be a problem.

If you want to learn how to trade price action, then study Wyckoff Forum. Start with stickies there.

If you want to test the indicators first, maybe you could try it with AmiBroker Demo version. I am not really sure what restrictions the demo (unregistered) version has, but I guess you can import price/volume data from .csv file and for sure you can test your custom indicator. But I think (not sure) that the demo doesn't support bar intervals under 1 minute.

I found out that I draw the best (and even objective) trend lines if I have always a channel in mind. I never draw a trend line alone, I always mark a channel.

That's nonsense. Though you don't need MP itself (as indicator) to make money, the MP -- and particularly the Auction Market Theory which it is based on -- is still far superior than all the MA Ribbon Could and other the stuff which you play with.

IMHO it is essential to look also at a larger scale than the range which you want to trade or which you wait for a breakout of. Take this picture. Would you say that waiting for an upside BO of the lower rectangle would be a good strategy? And would you say that after testing the upper rectange you just have to exit everything at the bottom of the lower one? Now imagine that this would be an intraday chart and imagine a trend following daytrader. Classical system of 2 contracts: one scalp and one runner. Such a trader would enter on the PB to the upper rectange, scalped the first contract somewhere to the midpoint or bottom of the lower rectange, then moved stop to BE and waited for trend continuation. And why couldn't this be applied on a larger scale (or TF, if you like) than daytrading? Not forcing you anywhere, just food for thought. EDIT: From the post which Thales posted I can see that the CL case didn't fit to what I describe in this post, but I guess it doesn't matter that much.

Actually, in this concrete example with CL there is one thing to notice. It is of course hindsight, but perhaps you could examine such a phenomenon so you could be prepared for it. You can see that price is range-bound between 72.5 and 73.5. Then it pokes up to 74, but it is rejected. Questions you could search answers for are: Wasnt't the poke a retracement on a larger scale? It is not visible on the chart, but wasn't 74 a low of the previous stair-step within a larger trend? Wasn't it a larger supply line? And even if it wasn't anything, doesn't a sharp rejection above a range increase probablity of a BO in the opposite direction, particularly if this direction corresponds with the larger scale trend? And if there was a higher probablity of a BO, wouldn't it be better to just scale out at S and give the trade some room? And what room? (I must admit that I according to my current rules wouldn't carry anything through such a bounce either. But someone maybe would, particularly someone who scales out it in more steps.)