How to determine the narrowest point based on the Bollinger

How do you know when a trend has started? Ended? This forum is for discussions about trend indicators and signals.
Post Reply
oem7110
Roundtable Knight
Roundtable Knight
Posts: 381
Joined: Wed Jul 12, 2006 9:33 pm

How to determine the narrowest point based on the Bollinger

Post by oem7110 » Mon Nov 05, 2007 11:48 pm

Does anyone have any suggestions on how to determine the narrowest point for range based on the Bollinger Band?
Range:=BBandTop(C,20,S,2)-BBandBot(C,20,S,2)

As the Bollinger Band constricts closer and closer together, price will usually be expanded. Does anyone have any suggestions on how to determine the narrowest point for Bollinger Band?
The difficulty is how to select the narrowest point between the widest points on both sides as the band expands and reach its own peak,
since there are a lot of zigzag beside the narrowest point.
Does anyone have any suggestions?

Thanks in advance for any suggestions
Eric

ADMP
Roundtable Fellow
Roundtable Fellow
Posts: 60
Joined: Wed Jul 04, 2007 9:04 am
Location: Germany

Post by ADMP » Tue Nov 06, 2007 6:06 am

Hi oem,

can you explain a bit more. What is the definition of the narrowest point?

Alex

Honeycomb
Full Member
Full Member
Posts: 16
Joined: Fri Apr 13, 2007 2:03 pm
Location: Portland, OR

Post by Honeycomb » Fri Dec 14, 2007 12:16 pm

the Bollinger bandwidth oscillator would probably be helpful for this, but I can't exactly tell what you're asking.

Roger Rines
Roundtable Knight
Roundtable Knight
Posts: 1946
Joined: Wed Oct 06, 2004 10:52 am
Location: San Jose, CA

Post by Roger Rines » Fri Dec 14, 2007 7:30 pm

Sometime this is a matter of testing from a trial and error process. For example, if I were challenged with this task, I would create a simple range band set around the data using some similar median price value, like the median price used to create the baseline for the differential of the Bollinger Bands you are wanting to understand, that you would then attach an expanded value like a range calculation of prices on either side of the median price value.

With that simple bandwidth process generating both above and below price values, you now have a tool that can work either within and outside of the BBands that might aid you in finding at which value of your expansion process you find the BBands moving inside of the Simple Bands you created.

By knowing at what value the BBand move inside of the Simple Bands, you'll have a benchmark to base your BBand contractions against. By having a benchmark, you can then test and maybe even find a broad optimized value across a basket of markets that can then be used as a trigger to know when to launch an order.

Of course, once you get that far, you might think about what happens when the bands are really far apart to see if that gives you other opportunities.

efficiency
Senior Member
Senior Member
Posts: 27
Joined: Wed Jun 02, 2004 9:17 am
Location: Omaha Nebraska

Post by efficiency » Mon Dec 24, 2007 7:02 pm

Likewise, I can't grasp what the OP is asking.

The MetaStock formula for Bollinger bandwidth is:

4x(std(C,20))/mov(c,20,S)


20 days (roughly one month) was John Bollinger's default. It's not carved in stone nor is 20 days statistically significant.

Using the preceding formula will yield a graphic depiction of expansion and contraction. A good place to start (ala' John Bollinger's "squeeze") is the narrowest point in 6 months.

There really is no "usually" about it. A contraction is constrained to NO movement, hence there "must" be eventual range expansion. The direction still remains to be seen; and any "breakout" can of course be a head fake. Thus the need for filtering.

The narrowest bandwidth in 6 months coupled with an ADX < 20 would suggest a candidate worth watching.

Post Reply