Measuring Choppiness

dima777 · Aug 27, 2008

Quote from Pa(b)st Prime:

Instead of viewing my remarks as "rude' why not try to learn. In all liklihood (99.999%) I'm more experienced and profitable than you. That being said I'll give you a quick mini-thought that I get paid to give clients. A pro-bono for my "rudeness."

First: It matters little what time frame you're looking at. "Chop" is relative (like everything eh?). Chop can be defined as in a range. The market has found value at a price point and the trade is just a random auction back and forth. Sellers dry up on the lows buyers dry up on the highs. No need for a definition-a simple look at a bar chart tells all. If you're using Bollinger bands you'll see the bands narrow and move sideways. You'll see a preponderance of bars move back and forth through your 21 period ma. Doesn't matter if it's a 1 minute or a monthly. Once again it's relative to your time frame. What ends the chop?

For starters prices doing something that they didn't do before. If ES hasn't made a bigger move than 7pts one way or the other over x periods the moment the market moves 7.25 it's now done something not true to form. The profile has changed. Will there be follow through? Probably to some degree. Will there be massive follow through? There's no way of telling. All you can do is make the bet and put in a stop at a price that shouldn't be hit if the market is indeed going to expand it's usual range. I suggest you read Curtis Faith's Way of the Turtle for some simple break out formulas.

Just accept that there's no right answers because you're dealing with unknowable outcomes. There's always going to be false breakouts that do nothing more than cause short covering and new longs to appear with their buying then checked by aggressive sellers laying in the weeds higher. Just set up your trades as much as you can with some idea of probabilities and let your positive expectancy smooth out the disappointments from so many individual bummers.
More...

thanks for the detailed explanation....as far as I understand you define choppiness as moving randomly within a horizontal prize range..right...i wonder what are the other approaches..

Pa(b)st Prime · Aug 27, 2008

Quote from Thunderdog:

I think that trading indexes on breakout is unnecessarily risky and somewhat painful. Unless, of course, you're the kind of guy who likes to draw a crowd by holding your hand over a flame.

The thing with free advice is that sometimes you get less that what you pay for.
More...

Personally I RARELY initiate an index or Bond trade on a breakout although I sold the market into weakness on the Sunday before MLK day and made the easiest/quickest 70pts I've seen in years. I religiously use breakouts though to take my losses. And in commodities I use breakouts FAR more often than trying to stand in front of the trend.

I will add this though. Just the fact that most ST traders I've worked with are more prone to taking shorts on up days and longs into down days (irrespective of trend) is why most traders fail.

austinp · Aug 27, 2008

<i>"I think that trading indexes on breakout is unnecessarily risky and somewhat painful. Unless, of course, you're the kind of guy who likes to draw a crowd by holding your hand over a flame."</i>

Knowing where & when to target breakouts is the easiest index trading you'll ever enjoy. Buying strength and selling weakness... at the key tipping points.

dima777 · Aug 27, 2008

Quote from MAESTRO:

A unique price change is the event when previous((Bid + Ask)/2) <> current((Bid + Ask)/2). The formula is showing the mean expected price deviation in N steps. If the real price change is less than expected value the chop is high; however, if it is higher, than the directional move is prevailing.
More...

THANKS..that sounds a bit complex...do you have any variant of the formula which will work on the raw prices?

MAESTRO · Aug 27, 2008

Quote from dima777:

...do you have any variant of the formula which will work on the raw prices?
More...

No

vikana · Aug 27, 2008

My favorite chop-index is calculated as follows for e.g. 10 periods

(close-close(10))/sum of individual bars (O-C) range

That gives you an idea of how much of the 10bar move was noise vs. directional.

You can also experiment with squaring the distances, using absolute values etc.

dima777 · Aug 27, 2008

Quote from vikana:

My favorite chop-index is calculated as follows for e.g. 10 periods

(close-close(10))/sum of individual bars (O-C) range

That gives you an idea of how much of the 10bar move was noise vs. directional.

You can also experiment with squaring the distances, using absolute values etc.
More...

THANKS...LOOKS GREAT!

Thunderdog · Aug 27, 2008

Unless your trading is automated (something I know nothing about), I can't imagine anything being better than a visual inspection of the chart. I think this is particularly the case when you rely solely on price action and without the unsightly scaffolding of indicators, trend lines and so on. If you need a formula to tell you when price is choppy, then you may wish to rethink what you really need. Again, this is assuming you are not engaged in automated trading.

Thunderdog · Aug 27, 2008

Quote from Pa(b)st Prime:

...most ST traders I've worked with are more prone to taking shorts on up days and longs into down days (irrespective of trend) is why most traders fail.
More...

I cannot imagine why there would be a causal relationship between time frame and the boneheaded willingness to step in front of a train.

dima777 · Aug 27, 2008

Quote from Thunderdog:

Unless your trading is automated (something I know nothing about), I can't imagine anything being better than a visual inspection of the chart. I think this is particularly the case when you rely solely on price action and without the unsightly scaffolding of indicators, trend lines and so on. If you need a formula to tell you when price is choppy, then you may wish to rethink what you really need. Again, this is assuming you are not engaged in automated trading.
More...

THANKS.ACYUALLY I AM DEVELOPING FULLY AUTOMATED SYSTEM

I WILL ADD THAT I HAVE CONCLUDED THAT MY EYE CAN BE BIASED AFTER HAVING CODED MY DISCRETIONARY APPROACH..PC WILL NEVER SEE WHAT IS NOT THERE