Why Can't I Trade with the Trend

Quote from hank rollins:

i don't believe your formulations are applicable to the issue at hand.

applying basic statistical measures like the serial correlation coefficient or the goodman independent time series test to most any market series, a trend is not in evidence--even in those cases were a MA indicates trending behavior. the tests show randomness in the data regardless of a "visual" illusionary trend.

You can either apply the accepted tests to the market data, or continue to believe whatever you wish.


I would love to see any data that refutes the above--- it would sure make my life easier :)
best wishes,

:)
More...

Hi hank,

What were the data vars you tested? Were they the right vars for determining "trend". I guess, one first needs to ask themselves exactly what it is they need to measure. Is the var applicable. What, if any, multivariate conditions exist. Which of them are essential (primary). Which are redundant or previously influenced. IMO, establish the primary.

Was your data continuous, or discrete ...by (those silly) "bars"?

Was the scattered distribution skewed... clustered... were there outliers... was it at all non-linear... etc. If you encountered any of these conditions, the correlation coefficient maybe irrelevant.

The measure I gave is correct, amongst other applicable versions. Calculating statistics though...correctly, is tricky. And believe me, I'm only a studying practioner. :)

One of my recent posts (probability thread) links to a good stats learning program for wrapping one's mind around stats, if anyone's interested.

Ktm'r
 
Quote from ktmexc20:

Hi hank,

What were the data vars you tested? Were they the right vars for determining "trend". I guess, one first needs to ask themselves exactly what it is they need to measure. Is the var applicable. What, if any, multivariate conditions exist. Which of them are essential (primary). Which are redundant or previously influenced. IMO, establish the primary.

Was your data continuous, or discrete ...by those silly "bars"?

Was the scattered distribution skewed... clustered... were there outliers... was it at all non-linear... etc. If you encountered any of these conditions, the correlation coefficient maybe irrelevant.

The measure I gave is correct, amongst other applicable versions. Calculating statistics correctly is tricky. And believe me, I'm only a studying practioner. :)

One of my recent posts (probability thread) links to a good stats learning program for wrapping ones mind around stats, if anyone's interested.

Ktm'r
More...


thanks. i'll reply shortly.

:)
 
There are two variables here: price and time.

A formula like y = mx + b simply describes the graph, using x and y axes, of a straight line in terms of the y (e.g. price) for each x (e.g. time). The constant m is the slope (or "trend") of this line.

The expression "dy/dx" refers to the slope of a curve at a point, which some "trend traders" might find suggestive when looking at their charts.

In the TA under consideration, price action is a discrete series, dependent on one variable, without serial correlation.
 
Quote from hank rollins:

even in those cases were a MA indicates trending behavior. the tests show randomness in the data regardless of a "visual" illusionary trend.


:)
More...

LOL those must be some tests !
 
Quote from NickelScalper:

...There are two variables here: price and time....
More...

Price...why price. Is price primary or influenced? Might be something to consider...?

Things that make you go ...hmmm. :)
 
Quote from NickelScalper:

The reason trendies run into trouble is that they look at price as dependent only on time.
More...

I'm amazed when someone that believes so strongly that trends do not have any value in trading puts themselves out as knowing what lurks in the minds of those that do.

Price is perfect and not dependent on anything. Markets are traded in transactions (contracts, calls, puts or shares) not time.
 
Quote from ProfLogic:


...Price is perfect and not dependent on anything. Markets are traded in transactions (contracts, calls, puts or shares) not time.
More...
Hi Prof,

Respectfully disagree! Price (as in: change in price) is a by-product. The effect of the cause...so to speak. Kind of like all that nasty CO emitted from gas guzzlers. :)

I know... you were baiting me,lol.
 
You must log in or register to reply here.
Back
Top