questions on counter trend system

Quote from travis:

... I just wanted to know how you could measure (short frame range, rsi...?) the fact that price keeps on bouncing against a given level, rather than moving for long runs, one way or the other...
More...
Your question is best analysed in the frequency domain, not the time domain. Traders insist on doing their analysis in the time domain because the frequency domain tools are rarely available, and perhaps as a result, are less instintive and less understood.

I recommend you make a study of the autocorrelation of a signal.

http://en.wikipedia.org/wiki/Autocorrelation

It is a time-domain tool that will likely give you the measure you want, or at least start you thinking along the right lines as to other possible tools that may give you what you want. However, you should think about using frequency domain tools too.

If you take anything from this post, is that you free yourself from the time domain when subjecting a time-series to analysis.

nitro
 
Quote from travis:

In the picture I attached you will see two types of trading days. On the first example day the price goes straight up, and on the second example it goes up, bounces against resistance and comes back down, and back and forth.

Neither day is a trend day, or a breakout day, so they could qualify as range days. Yet, I only make money on the second type of day, because I have a counter trend system, that makes money when the price bounces against resistance or support, and the more it bounces the better it is.

I am trying to find a way to measure and define what exactly is the difference between day 1 and day 2. I think that maybe a way would be to measure intraday range. Whereas on both days the daily range is just the same, a range measured every ten minutes might differ. But this is not good enough still.

Does anyone have any suggestions?
More...
===========
T-ravis;

Yes, even if you don't want to understandably show the stocks/derivatives you trade, you should find something similar[but real, concrete, symbol] to post.

SPY is up-trending;
AMR is down trending.

Wisdom is profitable to direct. More context is needed.
Same thing Steve Nison said in candlechart article in current Futures mag.[stocks, futures, options]

And another elite-Trader gave you a similar hint on -am;
afternoon comment.And another on 2 system comments.

:cool:
 
travis, I think the question you are trying to ask is how to know in advance if a particular day is going to be a trend day or a range day. Presumably, you'd activate your system on the range days and deactivate it on the trend days. Is that right? If so, I think you already know the answer: to predict whether it's going to be a trend or a range is as difficult as to predict the price itself. More precisely, the odds of a range to odds of a trend is around 75:25, but your anti-trend system will probably have the size of an average losing trade about 3 times the size of an average winning trade. Of course, the opposite is true for the trend-following systems. So, while both types of systems have a legitimate premise, both of them are equally challenging.
 
Hello Travis,

Maybe you could discern something useful about intraday market structure from one of the opening range breakout systems (Crabel, Fisher, closingtrend.com etc.). Such systems may give you an idea on how to interpret the first part of the day to make a statistically likely guess at what will happen during the last part, may be able to invert or modify signals normally used to predict a trend day - predicting a non-trending narrow range...

smitty
 
Mail2smitty, congratulations, as you are one of the few people who actually understood exactly what I was talking about. Your idea sounds just right, because I want to understand the nature of the choppiness, the zigzagging of the day as it evolves. I am not trying to predict today. I am just asking to measure in some way how it zigzags. Then I will see if, as I think, I can make assumptions about the rest of the day, based on what I measured on the first part of the day.

My question is this now. How do I implement your (complex) method on excel? My system is entirely automated on excel. It trades 5 different futures. Even just using the RSI doesn't seem very simple to implement on excel...this is not to say that my system is simple, otherwise I couldn't be trading 5 systems in just one file at the same time.
 
The primary difference between the two charts is

volatility.....

the second chart is significantly more volatile than the first.

Using a cumulative volatility measurement you could obtain a way of determining just how different the two markets are as time goes on..

All you have to do is determine a baseline. When your local volatility exceeds (or fails to exceed) that baseline, you get a signal....

steve
 
Quote from travis:

My question is this now. How do I implement your (complex) method on excel? My system is entirely automated on excel. It trades 5 different futures. Even just using the RSI doesn't seem very simple to implement on excel...this is not to say that my system is simple, otherwise I couldn't be trading 5 systems in just one file at the same time.
More...

You can get most of the common indicators and such like, for Excel here

http://www.ta-lib.org/

It might help a bit.
 
The answer to the original question is rather complex but to put it simply the combination of phasing of underlying cycles determine if the price is going to advance or bounce around or decline in a particular time segment.

I use the term cycle in a broad sense if it bothers you use the words underlying influence.

In order to find periods when a particular condition will likely be present one can create a structure where (for the sake of practicality) underlying influences are substituted by commonly used indicators like oscillators, moving averages and volatility bands.

There are conditions when everything is in sync (it would be beneficial for one to know the nature of these indicators since each of these discern a different aspect of price movement ) and this is when prices move rapidly and times when the range of indicators are garbled - for lack better word - or when they are at extremes, and these are conditions when choppy action is taking place until the various influences fall into a pattern when rapid movement is possible again.

We are talking of one time scale now but this goes on in every time scale and at times these conditions in the different time scales line up in such a way that a particular move is inevitable.

There is so much liquidity and tension in some markets that even in the minute scale one can see clearly how the different influences shape price action.

I use a background similar to what I described above in scalping Crude Oil and ER2 successfully off charts that are shorter then the 1 minute time interval and it helps me immensely to sense when to fade or not to fade the market.
(I started a thread in the Energy Futures area headed: Some observation on scalping Crude)

One could say that to isolate conditions when everything in the background is supporting what one wants to do is looking for a needle in a haystack. It is true, but today we have the tools and speed to do it. :D

Regards,

GC
 
You must log in or register to reply here.
Back
Top