BlueHorseshoe

I don't have any interest in dissuading the author/vendor from whatever he is doing. I do have an interest in trying to prevent new or naive traders from jumping in headfirst with something like this just because they read it on a forum. The ES is possibly the most mean-reverting market going - trading range/volatility breakouts in this market does not constitute sound advice. BlueHorseshoe

Though unimportant, THIS IS NOT TRUE. Phone your broker. Ask them what filters they apply to the data they receive from the exchange before they pass it on to you. Pretty much the only people receiving raw data are HFTs, and they apply their own data cleansing methods just like your broker does. Not that this matters - it is irrelevant to the fact that the close of the bar is meaningless. BlueHorseshoe

You're still not getting it, are you? Not one of you. "Staggered" timeframes is a way to illustrate the point - nobody is suggesting that they occur in any significant manner (though differences in data might result in such discrepancies). Forget about "staggered time frames". Forget anyone ever proposed their existence. The point is very simple: The timeframe is a discrete structure arbitrarily imposed on continuous data. It's an arrogant getsure, a bit like trying to map the landscape. Consider the contour lines showing altitude on an ordnance survey map . . . When I walk outside my front door, is the ground perfectly flat for a hundred yards, and then at the point where the contour line is drawn on my map, does the ground instantaneously drop by twenty feet? Of course not. The terrain has continuous gradient; the contour lines on a map are a discrete structure we impose to try and represent this in two dimensions. If you don't "get it" from this post, then you should probably give up, because it really doesn't matter too much anyway . . . BlueHorseshoe

Go and sit at an institutional desk where there are no charts - just a ticker. How often does a bar form? Everytime the trader glances up from his crossword and notes the current price. But the issue is not whether we are all seeing the same chart. The issue is whether or not the discrete time snapshot that we all see holds special significance. It doesn't. BlueHorseshoe

By the way, you mentioned Pivots. If by this you mean a "swing" type structure (a high with lower highs on either side of it) then then this is an excellent quantitative way to define S/R levels. If you meant "daily pivots" or "floor trader pivots" which are the static lines calculated each day from the prior day's prices - well they're a complete waste of time in my opinion and experience. Hope that helps.

I would imagine, only based on historical simulation. If you're doing it with enough regularity then you could aslo track how well subsequent reality conforms to your historical projections, and arrive at some kind of 'confidence quotient' for your probability calculations. It's very difficult to take in more than a fraction of the complex tapestry of information which may or may not affect the probabilities of a particular outcome. If you're asking more specifically about the ES, it is 'mean reverting' and it is possible to find directional edges around this behaviour, especially in higher timeframes. You can typically fade any move to support or resistance using limit orders, for example, and expect to be right more often than you're wrong. You could then refine this - probabilites for fading S/R in an up/downtrend, on a second test of S/R, a third test, on high volume, on low volume, on a low volume second test following a high volume first test . . . The possibilities (and probabilities) are endless. BlueHorseshoe

I don't think that I could really help with any code for that, I'm afraid. Have a hunt around in the default indicators for the platform though - you may find something that is workable. Sorry I couldn't be more help. BlueHorseshoe

I'm not entirely clear I fully understand your question - do you mean the historical percentage win rate? (Number of Winning Trades / Total Trades) * 100 Whether your future win rate will reflect your hypothetical win rate with historical data is another matter. And percentage win rate is fairly irrelevant to net profitability; it's not good winning $10 for 90% of the time if you lose $100 10% of the time. After 100 trades you will, on average, be down by $100. The very best illustration of this kind of high win rate net loser trading that I can think of is none other than . . . Victor Niederhoffer. BlueHorseshoe

It wouldn't make TA nonsense, it would make TA less precise. I'm not sure that anyone else on the thread who supports TA supports the idea that it is precise. I think this is what SIUYA and Sun Trader are trying to explain - it's a general guide, not an exact measure. The two distinct and obvious lows of a double-bottom will always be there, whether you're looking at minute bars, range bars, tick charts . . . whatever. You know, if you want to get pedantic about it, it's completely impossible to put a buy order in at a support line anyway - a line has no thickness Of course, "support" is really a vague and ill defined area of price where something happens that prevents price from trading below it. To the tick? Not often. BlueHorseshoe

You don't get any information from the exchange, I'm afraid. To do so is very expensive. You get your data from your broker, and your broker has "done things to it" to make it "better" for you (like filtering bad ticks). My broker might do things slightly differently to yours. I once ran a comparisson between data from a UK Spread Betting company and Tradestation - the differences were huge - but then the SB company made their own markets . . . Although that happens with Spot FX, doesn't it? But the real issue isn't whether we're all seeing a different chart, the issue is that the close, relative to the high, low, open, prior close etc, is arbitrary and meaningless. Even if we are both looking at the same chart it is meaningless. Why? Because as I said before the chart is imposing an artificial and discrete time structure on a continuous data set. As long as you are aware of this and how it does or doesn't affect what you do, then it's probably not worth getting overly hung up about. Those who frequent the 'candlestick corner' may need to be a little more concerned. BlueHorseshoe

What platform do you use? BlueHorseshoe

And yet, speculators get paid to facilitate the transfer of risk. There is an inherent tension between the need to limit risk and the need to maximise profit. Are you CERTAIN that stop losses are the best way to resolve this tension? BlueHorseshoe

Hi Patuca, The high or low of a given bar could be different, just like the close could be different. So an idea like "I'm going to place a buy stop at the high of the prior bar" becomes a bit meaningless in terms of order entry (which is not to say that it wouldn't work nonetheless, just not because the prior high is some magical price that everyone is looking at). Similarly, "I'm going to buy on the close when the RSI(of the close) goes under 80" attributes meaning to the closing price that is nonsensical. HOWEVER . . . Once you start looking at things like trendlines, significant data points become harder to dodge, whatever way you stagger your timeframes. A virgin data point (such as yesterday's low tick) is there on a daily chart, an intraday chart, and any staggered picture of intraday data. Perhaps because your five min bar chart is staggered three minutes after mine it will appear in a different bar, but it will still be there and obvious on both are charts because price never went below it. So we could both draw a support line using it. This applies (with decreasing efficiency) to any short term high or low - if it is higher than the prices surrounding it then it will be the high of a bar, however you stagger your timeframes. IN SUMMARY . . . As far as using TA goes, SIUYA seems to put it most succinctly when he says: " its a road map not a set of universal rules or laws " BlueHorseshoe

The difficulty I have with this is that you're talking about relevance to specific instances, rather than making statistical generalisations, which is what indicators are intended to assist with. I have no idea whether the Euro will go higher or lower in this specific instance. Generally, though, I know that across a large enough sample size it will conform to certain expectations (provided that it continues to behave in the future as it has in the past). Because of the way market data is distributed (kurtosis, skew, fat tails, jump processes, behavioural shift etc) these models for generalising expectations do not tend to work as well as they would in a maths classroom. They often become frustratingly inaccurate. But they do still work, provided that they are based on a sound underlying concept. Most importantly, the concept needs to come first, and then you can start searching for the best tool (indicator) to do the job (summarize the information that you want). BlueHorseshoe

WARNING: Volatility Breakouts are not a viable strategy for the ES in any timeframe. It's too heavily mean reverting. Trying to promote this as a valid concept for the ES suggests an ignorance of the way this market behaves, probably because the author has not carried out any sort of basic data analysis for this instrument. On a swing basis from a daily chart you can even get a decent return on an account by fading volatile breakouts in the ES. I cannot be more categorical about this: do not try and trade volatility breakouts in the ES. BlueHorseshoe

I'm not sure about that . . . Draw a support line - use the lows, not the closing prices - that's good old fashioned TA and perfectly valid. Indicators based on closing prices are more questionable. My main objection though is Candlestick and similar charting methods - pure nonsense. Range bars are also another matter; though the close of a range bar is meaningless, it can have utility in terms of order entry because it's maximum distance from the open can be anticipated at the open. BlueHorseshoe

Hi, Just plot the text to the bar before the one you are currently plotting it to. Then it will be to the left. There is also something more complicated you can do to "pad" text to reposition it - let me know if you would like an example of this. BlueHorseshoe

Iceberg Capital? Sounds like you could end up having your account frozen . . . :rofl:

Don't be impressed by all the weird and wonderful lines, boxes, and digits . . . Be impressed by the fact that he's evidently managed to leverage the platform's functionality to provide him with the exact information he needs to trade. Charting software is just a tool - how cool or different it looks is unimportant - what matters is how you use it to create a picture of the market that allows you to profit. BlueHorseshoe

Hi Bakrob, Thanks for replying. RE: unfilled market orders . . . any order that is not a limit order will be executed at market. So when you see a boat load of market orders hit the bid, say, then rather than knowing that a lot of short sellers are committed to get in on the move, how do you know that a bunch of stops weren't just triggered? RE: filled on the bid or the ask . . . I still maintain that this is a very partial picture. The relationship/relative quantities between limits and stops/market orders on both sides of the spread has got to be far more meaningful than just up and down volume. It doesn't matter how many hundreds of sellers get very enthusiastic with market orders if the bid is 20k deep, for example. RE: your chart . . . thanks for posting this - it's really fascinating. What you do is clearly intelligent and well thought through, and far smarter than anything I have managed to put together with the same data. If it's working for you then there is no need for us to continue this conversation. And completely ignoring volume delta etc, the general approach perfectly fits with my understanding of how to trade the ES intraday, which is encouraging for me. On a side note, TradeStation 9.2 looks great - I really should remember to hit that 'UPDATE' button the next time it flashes up! Cheers, BlueHorseshoe

"Edge" is a strange word, though I often use it myself. In an idealistic sense, an edge can never be lost because it is a fundamental and immutable true decription of some aspect of market behaviour . . . If you think there is such a description is impossible and markets always change . . . well that is just another meta-theory and you could exploit it as an edge. As Terry Eagleton put it 'not having an ideology is an ideology'. Enough of that though! In the less general sense that you're using the word (with "edge" pressumably more similar to "strategical advantage"?) then I wonder whether what you say is true? Surely in any large and liquid market it is not a question of specific traders evolving or perishing . . . Isn't it more useful to think of other traders in terms of "types"? And if you do this, isn't it easy to imagine that a certain type is always re-populated as its old members perish or evolve? A more concrete example: those who buy breakouts in the ES on 5min charts are not the same traders who bought into these same moves five years ago. Those doing this five years ago perished or evolved. But someone else is doing it today, and I can still take the other side of their trade. All of this kind of feeds back into the topic in my mind as follows: As the lemming's numbers depreciate, so too must the liquidity into which you can lean a money management approach. So whether a class of participants who wise up and are, or are not replaced, is critical to position sizing. If the proliferation of lemmings across any specific instance of a bunch of markets is random, and lemming extinction is a function of time, then diversification makes sense: the more you diversify the more chance you have of taking the opposite side of a population of lemmings in an ascendent phase . . . Christ that got wordy! BlueHorseshoe

You can't see unfilled market orders, and once they fill you don't know at what price they were actually intended to execute. Moreover, the significance of market orders can only be measured in relation to resting limit orders, and they're slippery and ephemeral things. Unless you're able to piece everything that's going on together at very high speed (no offence, but I think that's probably beyond the capability of any human brain, yours included), then it is hard to see how you could divine anything of significance from this information. I could easily be wrong though. BlueHorseshoe

No sophisticated larger market participant is using limit orders that you can detect (unless you have some seriously advanced algorithm for reconstructing icebergs, estimating PIQ etc) from the bid and ask. The only time they will show their hand in this way is when it's the last leg of a massive order and they no longer need to care. If they tipped their hand as clearly as you suggest then every retail trader would just front run them from the DOM. BlueHorseshoe ps I'm not trying to suggest that your entry setup doesn't work - stepping in front of fragmented net positive/negative order flow can work - it's just not for the simplified reason you suggest.

You'll always buy 'a tick up' - that's just paying the spread and happens EVERY time you use a market order. You can't get a market buy order filled at the best bid because nobody can sell to you on a limit at the best bid. Slippage is where other market orders fill all the available limits on the other side of the book, that price tier collapses, and your order interacts at an even worse level. Normal market order - you pay the spread (minimum one tick). Slippage - you pay the spread and then some . . . BlueHorseshoe

Because a limit order may not get filled - for a limit order to be filled, someone has got to cross the spread with a market order. If you bid 100 limit, and the best ask is 102 (ie a 2 point spread), then for your order at 100 to be filled, a new seller has to enter the market and be willing to sell at 100, crossing the spread and stepping in front of the limit seller at 102. Then your orders will be matched and, assuming that you are the only bid at 100 and nothing else changes, the spread would widen to the next highest bid (99, say) by 102. A new limit seller might enter at 101, more limit buyers might come in at 100 or 101, or a massive buy market order may hit, and everybody selling at every price below 110 get taken out in a millisecond. If there are a whole load of other buyers ahead of you in the queue at 100 (they placed their orders first), and only a few sellers cross the spread with market orders to trade at that price, then you won't get filled. What does all this mean to you? If you're "desperate" to get filled then you'd use a market order. The upside is you're guaranteed to get filled. The downside is that you will have to pay the spread (and possibly slippage). If you're only prepared to buy or sell at a specific price then you'd use a limit order. The upside is that you'll make the spread if you get filled. The downside is that you might not get filled. Moreover, you'll always be filled on a losing trade, but you won't always be filled for a would-be winning trade (a loser must, by definition, trade through your limit price). I strongly recommend that you spend some time getting a decent grasp of all this, otherwise you may not be able to execute effectively, regardless of how good your strategy is. Regards, BlueHorseshoe