trend following - quo vadis?
- Thread starter man
- Start date
i think nassim taleb with his empirica fund did basically that: buying far OTM spreads and wait for the fat girls. i am not sure about the approach. paying theta all day long is a tough pill to swallow. it is definitely quite different from "normal" trend following, simply because vola and time play an important role in addition to delta ...
i am not an english native and i have not got the joke on your handle ... so sorry.
okay. first of all i thank everyone for the discipline here. actually i think i was the only one with nasty talking ...
second thing. i am truly on the quant side of things. i guess it would be called "edge"-world here. my thing is to put things in coded form. then they have to show how good they truly are. before that they are just ... ideas. and most are crap.
having said that i truly respect discretionary traders, simply because their brain is the most powerful cpu out there when it comes to creative and complex pattern recognition on a very limited set of time series. i think this is what takes place here: yes the graphs shown make clear where the decision points are, but they depend on the trend lines and these are (correct e if i am wrong) not automatically produced by coded formulas. so, though i do not doubt by any means the quality of the talk, i have to make a step aside, since i am playing another game.
nevertheless i will take the chart as advised and look at it once in a while. maybe it inspires my other work ...
usually i am reluctant to do this kind of thing, since the graphs are often biased. take bollinger bands. by pure definition almost all prices are within the bands, yet that does not mean a thing when it comes to trading. price will remain within the band, yet not necessarily mean revert.
similar here. look at the chart posted before:
http://www.elitetrader.com/vb/attachment.php?s=&postid=741703
see the upward trend line that started somewhere at 22:59 (horizontal axis). look at the end of trend line at the beginning of 20:01. it is clear that this trend line could only be drawn after you know the end at the beginning of 20:01. but all the horizontal green and red lines seem to "know" the trendline before it manifested. if you tried to computerise this and let us say we start the beginning of 22:59 and i wait for the first new low (let us assume this could be easily done) i would detect it at the end of section 22:59. if i make the trend line now by connecting it with my initial low=start at the beginning of section 22:59, i come up with a very different trend line for section 27.59, with red and green horizontals on significant higher level.
what i am trying to say is that charts of this kind might look convincing, yet they are (at times) flawed by our mind's willingness to disregard details.
and note that we are tallking about frequent trading here, thus it matters a lot where the exact locations of the decision lines are.
i truly think that some individuals trained their mind in a way that they can trade based on graphs of the presented form. most can't. grats to those of you who do well with it.
resuming: it is not as easy as the graph tries to tell ... nevertheless: good discussing this. open for comments.
peace
second thing. i am truly on the quant side of things. i guess it would be called "edge"-world here. my thing is to put things in coded form. then they have to show how good they truly are. before that they are just ... ideas. and most are crap.
having said that i truly respect discretionary traders, simply because their brain is the most powerful cpu out there when it comes to creative and complex pattern recognition on a very limited set of time series. i think this is what takes place here: yes the graphs shown make clear where the decision points are, but they depend on the trend lines and these are (correct e if i am wrong) not automatically produced by coded formulas. so, though i do not doubt by any means the quality of the talk, i have to make a step aside, since i am playing another game.
nevertheless i will take the chart as advised and look at it once in a while. maybe it inspires my other work ...
usually i am reluctant to do this kind of thing, since the graphs are often biased. take bollinger bands. by pure definition almost all prices are within the bands, yet that does not mean a thing when it comes to trading. price will remain within the band, yet not necessarily mean revert.
similar here. look at the chart posted before:
http://www.elitetrader.com/vb/attachment.php?s=&postid=741703
see the upward trend line that started somewhere at 22:59 (horizontal axis). look at the end of trend line at the beginning of 20:01. it is clear that this trend line could only be drawn after you know the end at the beginning of 20:01. but all the horizontal green and red lines seem to "know" the trendline before it manifested. if you tried to computerise this and let us say we start the beginning of 22:59 and i wait for the first new low (let us assume this could be easily done) i would detect it at the end of section 22:59. if i make the trend line now by connecting it with my initial low=start at the beginning of section 22:59, i come up with a very different trend line for section 27.59, with red and green horizontals on significant higher level.
what i am trying to say is that charts of this kind might look convincing, yet they are (at times) flawed by our mind's willingness to disregard details.
and note that we are tallking about frequent trading here, thus it matters a lot where the exact locations of the decision lines are.
i truly think that some individuals trained their mind in a way that they can trade based on graphs of the presented form. most can't. grats to those of you who do well with it.
resuming: it is not as easy as the graph tries to tell ... nevertheless: good discussing this. open for comments.
peace
Oh MAN... haha....
I am a pseudo-quant. I have a regular dayjob in Credit Derivatives. I built a modelling app that use 4 stochastic process (Interest Rates, Default Times, Recovery Rates, & GARCH). Furthurmore, I was given permission to bring on a PhD employee that could speak the same language that I am accustomed.
I only recently adjusted to discretionary trading. I've got an automated version that really CLEANS up the ultra micro T&S/DOM. It is fully mechanical and is a Regular Expression with only 6 states. I use price as the transition function, therein, a weakness. It is very difficult to assess volume in real-time and on the dom. The T&S/DOM was easy for the mechanical but difficult because of the speed and amount of information my brain can handle. It is difficult to pick out around 1200 ticks from average set of nearly 42k tick on ES data feed. Your statement about mean reverting is not really applicable. Pull up an annual fractal chart on any index. There's no alpha value there. It is like exchange rates. They need not revert, They can simply continue to drift away. During the day, the PhD quant who I only talk dayjob (never trading) with understands why people are so dependent on models. They are looking for something to predict, therein lies the most crippling problem with trading.
The chart you see has nothing to do with predicting. It is containment and decisions at containment points. Matter of fact, I make new decision points after the fact which is what makes this type of trading so stress free... I do hope that you take the time to really look at the chart as many ways as possible and hopefully you will find something that helps. I just use the channels for context. Most of the time, I just draw in my red and green lines after the fact since I know that at some point they will require that I need to take action at the very same point in order to continue profit taking. When required, I "T" over. This is singular action, not exit and enter, just reverse.
Kindest Regards,
G33M4K Type A (non-BEginner)
no.
)