GG on your 9.23 bar you would see one low and then on the previous 9.23 bar you would probably see either the same low or a slightly lower low. Now you could chose to use your current 9.23 and hope that the market would also go through the previous low. Or you could look for a pattern on your 4.47 minute chart that would give you a chance to get in early and hope to revisit the lows and get follow through through past the recent 9.23 low. Or you could wait for the previous low and place a stop one tick below. Or you could wait for pentration of the low and and find a pull back on the 1.27 minute chart with the 24.6 volume and moon cycle weighted moving average. Its all ball bearings. It was thoughts like these that led me to lunderstand Van Tharps statement that we can only trade our beliefs about the market and we use t/a to help us filter. As someone said before we can use standard t/a to see what the crowd is thinking or we make a jurik indicator to get the lag out. However, I always wonder would good getting the lag out is. Could I just use a shorter time frame or faster moving average?
unique view of the market
- Thread starter Gordon Gekko
- Start date
I look at 10 days of intraday 5 minute candlesticks and hold winners overnight.
whew, now that i have my 100th post out of the way! (see the "suggest a tagline" thread, it was a doozy 
) i wouldnt've even noticed except i think inandlong pointed out when i was at 99! anyway, here is my 101st:
i have thought a lot about that question, i mean since years ago, not just since you posted it here.. many people use a 5 min or 1 min bar and so you can expect pivot points around the top of the minute or the end of the 5-minute period when traders make their decisions. i have found this to be somewhat reliable (though i dont use it as a trading indicator). so instead of using a weird 0.9 bar or 4.33333, use a normal 1 or 5 minute bar except move it back 5 seconds (or 10 seconds or 100 milliseconds or whatever). maybe this is what was posted about "rolling" bar charts, but frankly i didnt understand what that post was saying, as it looked like he was describing regular 5 min bars. anyway, you can empirically determine how far to "phase" it back by analyzing the data and your trading signals.
that is one of those things you cannot backtest for because tick data doesnt contain milliseconds (and sometimes doesnt contain seconds either).
one other thing to keep in mind is that many traders data feed has some kind of delay (including yours/mine) and it also takes time for the trader to make a decision, enter the order, etc. (hence the empirical testing mentioned above) the trades that occur quickest at the end of the bar will be computer programs in general. i.e. they occur more like clockwork. but human trades will come pretty much at random. some computer trades will come at random also, as stops are hit and other indicators trigger trades.
if you *really* want to mix it up and think like a maverick, start a new bar at local minimas or maximas where the trend changes direction. this is just the beginning, and the possibilities are limitless. (better get started!
)
obviously, all the charts i talked about here would require a custom charting program.
one final word, these are some of the thoughts that ran through my head but IMO it's all moot since you can just look at a tick chart and see the data in its native form.
) i wouldnt've even noticed except i think inandlong pointed out when i was at 99! anyway, here is my 101st:i have thought a lot about that question, i mean since years ago, not just since you posted it here.. many people use a 5 min or 1 min bar and so you can expect pivot points around the top of the minute or the end of the 5-minute period when traders make their decisions. i have found this to be somewhat reliable (though i dont use it as a trading indicator). so instead of using a weird 0.9 bar or 4.33333, use a normal 1 or 5 minute bar except move it back 5 seconds (or 10 seconds or 100 milliseconds or whatever). maybe this is what was posted about "rolling" bar charts, but frankly i didnt understand what that post was saying, as it looked like he was describing regular 5 min bars. anyway, you can empirically determine how far to "phase" it back by analyzing the data and your trading signals.
that is one of those things you cannot backtest for because tick data doesnt contain milliseconds (and sometimes doesnt contain seconds either).
one other thing to keep in mind is that many traders data feed has some kind of delay (including yours/mine) and it also takes time for the trader to make a decision, enter the order, etc. (hence the empirical testing mentioned above) the trades that occur quickest at the end of the bar will be computer programs in general. i.e. they occur more like clockwork. but human trades will come pretty much at random. some computer trades will come at random also, as stops are hit and other indicators trigger trades.
if you *really* want to mix it up and think like a maverick, start a new bar at local minimas or maximas where the trend changes direction. this is just the beginning, and the possibilities are limitless. (better get started!
)obviously, all the charts i talked about here would require a custom charting program.
one final word, these are some of the thoughts that ran through my head but IMO it's all moot since you can just look at a tick chart and see the data in its native form.
bumping this thread because of this thread:
http://www.elitetrader.com/vb/showthread.php?s=&threadid=11008
http://www.elitetrader.com/vb/showthread.php?s=&threadid=11008