hey jem,
check this out:
http://www.elitetrader.com/vb/showthread.php?s=&postid=108105#post108105
p.s. read the top line.
unique view of the market
- Thread starter Gordon Gekko
- Start date
check this out:
http://www.elitetrader.com/vb/showthread.php?s=&postid=108105#post108105
p.s. read the top line.
Sarasota...
Never ceases to amaze me how many speech police / thought police we have around these days. Thanks for your condolences, but I'd rather you send money, if you really mean it! Sarasota, lighten up...Drink a beer....Flirt with a pretty girl...Take the kids out to play Putt Putt. It's the weekend man!
Gordon does have some interesting things to say. It just seemed as though he might have been bored on Saturday morning. Post seemed kinda "left field" to me. The first response by the infamous FasterPussycat just struck me as funny.
I'm basically in the rs7 camp: Keep It Simple. I tend to disagree with what I perceived to be the concept of the thread. I want to go whichever way the money is flowing, which would not be a "unique view of the market", IMHO.
Peace.....Hope you have good trades come Monday. ;-)
BSAM
i'm sick with a cold from hell actually.. anyway, when i started this thread, i wasn't declaring it to be "the way" to trade. i'm unsure of the concept myself, but it has at least sparked my interest.
as for insults and jerky remarks, i don't mind them one bit. i've even been known to utter a few myself. note to others: just don't let jerky remarks get in the way of speaking your mind, asking questions, and throwing out new ideas.
I suspect that the "unique view of the market" concept is treasonous to some types of traders (and that is why they rabidly oppose it). If everyone uses different bar lengths and bar phases (the start-times of bars), then it would lessen the herding or flocking behavior that some types of traders depend on for profits.
For example, you often hear comments about QQQ and e-Minis being so tradable. Perhaps this is because they have a large number of day traders all looking at the same intraday charts and same/similar indicators. With enough traders all following the same trade setups (which creates its own self-reinforcing confirmation of the setups) one would see nice little tradable flows as the flock of traders jumps on each new shared vision of a setup. Faster traders would profit over slower/nervous traders and traders with better exit discipline would also do better in flocking-based trading.
But, if everyone uses a different "Different Outlook Chart", then the market would have fewer tradable flocks of traders. With less shared vision of where the market is and where it seems to be going, fewer traders would see the same setup at the same time. The result is a less-coherent pattern in the market price action because everyone is doing their own thing. Definitely dangerous!
<b>Overfitting Different Outlooks</b>
As an aside, the "Different Outlook Chart" approach is likely to get the less-mathematically inclined traders into big trouble. Letting some automagical optimizer find the "best" definition of bar timeframe and phase WILL lead to overfitting. It will be too easy for novice traders to decide that 4.325276272 minute bars starting on prime-numbered seconds is "best" (or at least it was with last week's data). The more parameters a system has, the harder it is to find statistically defensible "optimal" values for those parameters. The reason "Keep It Simple" is so popular is that it is too easy to overfit the parameters/rules of a more complex trading system -- creating a system that backtests great and trades like crap.
<b>But Don't Worry</b>
So, should traders worry if software and data vendors offer more flexible charting/data series? I bet not. The desire to synchronize with other people is strongly built in to the human brain and into the markets.
I predict that in a world of "Different Outlook Charting," the traders on ET would be sharing tips like "I use 6.25 minute bars with 33 second start intervals." The more people that use a given definition, the more confirmation of the setups based on that definition. Other traders would use automagical optimizers that would latch on to any popular definitions of "Different Outlooks." Again, the self-reinforcing nature of the situation would attract other flocking traders to create convergence of a shared definition for "Outlook."
It really is like the penguins on the iceberg. "Different Outlook Charting" means that traders no longer have a shared vision of the situation -- its as if the penguins could no longer all see each other on the iceberg. But the desire to stick together is strong -- the penguins/traders would come up with mechanisms (such as ET) to communicate and recreate a shared vision.
Thanx GG, for a very interesting thread,
-Traden4Alpha
For example, you often hear comments about QQQ and e-Minis being so tradable. Perhaps this is because they have a large number of day traders all looking at the same intraday charts and same/similar indicators. With enough traders all following the same trade setups (which creates its own self-reinforcing confirmation of the setups) one would see nice little tradable flows as the flock of traders jumps on each new shared vision of a setup. Faster traders would profit over slower/nervous traders and traders with better exit discipline would also do better in flocking-based trading.
But, if everyone uses a different "Different Outlook Chart", then the market would have fewer tradable flocks of traders. With less shared vision of where the market is and where it seems to be going, fewer traders would see the same setup at the same time. The result is a less-coherent pattern in the market price action because everyone is doing their own thing. Definitely dangerous!
<b>Overfitting Different Outlooks</b>
As an aside, the "Different Outlook Chart" approach is likely to get the less-mathematically inclined traders into big trouble. Letting some automagical optimizer find the "best" definition of bar timeframe and phase WILL lead to overfitting. It will be too easy for novice traders to decide that 4.325276272 minute bars starting on prime-numbered seconds is "best" (or at least it was with last week's data). The more parameters a system has, the harder it is to find statistically defensible "optimal" values for those parameters. The reason "Keep It Simple" is so popular is that it is too easy to overfit the parameters/rules of a more complex trading system -- creating a system that backtests great and trades like crap.
<b>But Don't Worry</b>
So, should traders worry if software and data vendors offer more flexible charting/data series? I bet not. The desire to synchronize with other people is strongly built in to the human brain and into the markets.
I predict that in a world of "Different Outlook Charting," the traders on ET would be sharing tips like "I use 6.25 minute bars with 33 second start intervals." The more people that use a given definition, the more confirmation of the setups based on that definition. Other traders would use automagical optimizers that would latch on to any popular definitions of "Different Outlooks." Again, the self-reinforcing nature of the situation would attract other flocking traders to create convergence of a shared definition for "Outlook."
It really is like the penguins on the iceberg. "Different Outlook Charting" means that traders no longer have a shared vision of the situation -- its as if the penguins could no longer all see each other on the iceberg. But the desire to stick together is strong -- the penguins/traders would come up with mechanisms (such as ET) to communicate and recreate a shared vision.
Thanx GG, for a very interesting thread,
-Traden4Alpha
hmmm Traden4Alpha, i noticed you have begun calling these "different outlook charts." hmmm
i think i see an empty bag and a cat..but i could be wrong.
i think i see an empty bag and a cat..but i could be wrong.
Yes, and its a very pretty, but somewhat dangerous cat.Originally posted by Gordon Gekko
hmmm Traden4Alpha, i noticed you have begun calling these "different outlook charts." hmmm
i think i see an empty bag and a cat..but i could be wrong.
Of course, anyone used to playing with markets, should be comfortable with playing with fire. I wonder if some others will say that they see an empty can and a whole mess'o worms
Get well soon,
-Traden4Alpha
well, OK. Too much flexibility may concern some traders. Well, I remember the same kind of concerns regarding decimalization.
I'd like to remind you something though.
candle bars are NOT the reality of the market. (I would argue there is no 'single reality' but that's another topic).
the candles 1, 5 15mn or whatever are only a representation of the price action. it does not matter what size you go for, the market action DOES NOT CHANGE because you select 5mn or 30mn bars !
However, it will make a difference on your decision. So that means it's a filter.
Filtering is good to help you make decisions.
Filtering is bad because you may miss important information.
Straight to the point, changing the filter won't show you a trend early. It may filter some trends and help you focus on one.
tweaking the bars might be interesting, it's comparable to changing parameters of an indicator. the bars are an indicator !
and just like any indicator this is in no way a recipe to easier profits, sorry. thinking about it that way is, again, seeking for the holy grail.
the closest chart to 'reality' is the tick chart. if you want to understand how the market works, follow in realtime a tick chart of the nasdaq futures (or ES whatever).
After a while you should get it.
Don't get me wrong I do use bars too, not tick charts. But it's not the most important thing. it's just a filter. and you want to filter because you can't trade all the active trends in a market at any given time. filtering helps you focus on one. don't make a big deal about it.
worse, you may overoptimize your view for a particular market condition and be off when another condition comes in.
of course, if you use the flexibility and remain flexible, then you can do whatever you want.
once you understand indicators (including bars) are not the most important thing.
tntneo
I'd like to remind you something though.
candle bars are NOT the reality of the market. (I would argue there is no 'single reality' but that's another topic).
the candles 1, 5 15mn or whatever are only a representation of the price action. it does not matter what size you go for, the market action DOES NOT CHANGE because you select 5mn or 30mn bars !
However, it will make a difference on your decision. So that means it's a filter.
Filtering is good to help you make decisions.
Filtering is bad because you may miss important information.
Straight to the point, changing the filter won't show you a trend early. It may filter some trends and help you focus on one.
tweaking the bars might be interesting, it's comparable to changing parameters of an indicator. the bars are an indicator !
and just like any indicator this is in no way a recipe to easier profits, sorry. thinking about it that way is, again, seeking for the holy grail.
the closest chart to 'reality' is the tick chart. if you want to understand how the market works, follow in realtime a tick chart of the nasdaq futures (or ES whatever).
After a while you should get it.
Don't get me wrong I do use bars too, not tick charts. But it's not the most important thing. it's just a filter. and you want to filter because you can't trade all the active trends in a market at any given time. filtering helps you focus on one. don't make a big deal about it.
worse, you may overoptimize your view for a particular market condition and be off when another condition comes in.
of course, if you use the flexibility and remain flexible, then you can do whatever you want.
once you understand indicators (including bars) are not the most important thing.
tntneo
if we take what you say to be true, if nothing else, could it give you better fills?
for example, say in the past you traded the same daily bars like everyone else, but now you trade .923 daily bars. if in the past you put a stop under a day's low, you may have gotten a bad fill since there may have been a lot of other stop orders there too. if you place your order below the relative .923 daily bar, maybe you are avoiding the crowd, which could give you better executions.
All of this is better discussed in the frequency domain.Originally posted by Gordon Gekko
if we take what you say to be true, if nothing else, could it give you better fills?
for example, say in the past you traded the same daily bars like everyone else, but now you trade .923 daily bars. if in the past you put a stop under a day's low, you may have gotten a bad fill since there may have been a lot of other stop orders there too. if you place your order below the relative .923 daily bar, maybe you are avoiding the crowd, which could give you better executions.
nitro