BollingerBands
- Thread starter columbus
- Start date
Dear columbus,
In the posts so far, position of current close with reference to band high & low ( Bollinger Band % Position) was considered as the criteria for entry & exit.
Please throw some light on the Bollinger Band width aspects also. ( % width = diff. of high & low / average of high & low)
When the width is less, the price % change before the next reversal is too small to make any trading sense (sideways movement). At the same time, low widths may give a possibility of a breakout.
When the width is more, the volatility is more, resulting in the possibility of giving higher returns at a higher risk.
Please offer your comments on the width aspect also.
Thanks.
murthymsr
In the posts so far, position of current close with reference to band high & low ( Bollinger Band % Position) was considered as the criteria for entry & exit.
Please throw some light on the Bollinger Band width aspects also. ( % width = diff. of high & low / average of high & low)
When the width is less, the price % change before the next reversal is too small to make any trading sense (sideways movement). At the same time, low widths may give a possibility of a breakout.
When the width is more, the volatility is more, resulting in the possibility of giving higher returns at a higher risk.
Please offer your comments on the width aspect also.
Thanks.
murthymsr
The Bollinger Bands is a calculation of a standard deviation. In essence, whatever TF you have it plotted on, it becomes a moving SD for that TF. The formula for an SD cannot be shown from a word processor, as the formula is more involved than just what a word processor would present.
Here's some statistical facts about SD's. With "normal" SD's there is a 5% chance of any given 2 sigma to be touched and a 2% chance for a 3 sigma to be touched. Considering in use the BB's for trading, they become a moving SD. That being the case the calculations increase for the probability of the extremes to be hit. I did a more in-depth study of this and found the chances for the 2 to be hit is about 16%, and for the 3 to be hit is around 7%.
Here's something all the readers of this thread can try doing. The default settings for the BB's are 20,2. Do something a little different. Set it for 28,3. The reason is a standard (Not a stereotype, just the rule.) cycle in any market is about 28 candles in length. The BB will measure and take into account that full cycle. When the 3 is hit, it, naturally is a far more extreme than a 2. When the candle releases its grip on the band, then be prepared to go the contrarian direction.
In order to prove that, plot a 28,2 and a 28,3 BB, simultaneously on your demo account, then notice how long the candle sticks to the -2 or +2 band. Then notice how quickly it will break its grip on the 3. This because the trend's extremity got hit, and it is time to go back the other way.
Here's some statistical facts about SD's. With "normal" SD's there is a 5% chance of any given 2 sigma to be touched and a 2% chance for a 3 sigma to be touched. Considering in use the BB's for trading, they become a moving SD. That being the case the calculations increase for the probability of the extremes to be hit. I did a more in-depth study of this and found the chances for the 2 to be hit is about 16%, and for the 3 to be hit is around 7%.
Here's something all the readers of this thread can try doing. The default settings for the BB's are 20,2. Do something a little different. Set it for 28,3. The reason is a standard (Not a stereotype, just the rule.) cycle in any market is about 28 candles in length. The BB will measure and take into account that full cycle. When the 3 is hit, it, naturally is a far more extreme than a 2. When the candle releases its grip on the band, then be prepared to go the contrarian direction.
In order to prove that, plot a 28,2 and a 28,3 BB, simultaneously on your demo account, then notice how long the candle sticks to the -2 or +2 band. Then notice how quickly it will break its grip on the 3. This because the trend's extremity got hit, and it is time to go back the other way.
Hi 4xpipcounter,
Though I have not tried BB(28),which I think is too WIDE, I posted BB(20)
here:
http://www.traderji.com/day-trading/48558-bollingerworm-3.html#post511623
Though I have not tried BB(28),which I think is too WIDE, I posted BB(20)
here:
http://www.traderji.com/day-trading/48558-bollingerworm-3.html#post511623
Last edited:
Dear murthymsr,
Yes ,all my posts here ,deal with CLOSE of candles, where as in new thread
they deal with HIGH & LOW of candles.
In fact ,it is observed that,narrow width will give to BREAK-OUT as you
have written.The same thing I mentioned in my posts also.To be frank
I won't give much emphasis to WIDTH of Bollinger Bands ,in fact ,Inclination
will give more information.(Inclination as seen by eye).Width over a period
is given below.
Source:NSEGUIDE
Yes ,all my posts here ,deal with CLOSE of candles, where as in new thread
they deal with HIGH & LOW of candles.
In fact ,it is observed that,narrow width will give to BREAK-OUT as you
have written.The same thing I mentioned in my posts also.To be frank
I won't give much emphasis to WIDTH of Bollinger Bands ,in fact ,Inclination
will give more information.(Inclination as seen by eye).Width over a period
is given below.
Source:NSEGUIDE
Last edited:
http://img218.imageshack.us/i/202l.png/
http://img267.imageshack.us/i/283.png/
Columbus, that's why I mentioned something to the effect of plotting the 20 and 28 on your demo account. 28 may seem to wide, but it takes into account the full cycle (Which sometimes can be up to 32 and even 34.). The 20 has a lot of action with the candles riding on the upper or lower bands, which, IMO, renders the bands useless.
Also, I prefer the 28,3 over the 28,2, because 3 sigma is more extreme and less likely to be hit, which renders a better trading opportunity.
Attached, are a couple of charts of the GBP/USD. One with the 28,3, and the other with the default, 20,2. The rule with the 28,3 is to wait for the break from the band, then enter going the opposite direction, and wait for at least the mean to be hit. Or, if the extremity is exceeded, then wait for the candle opposite the direction it is traveling in (E.g--Bear candle in an uptrend.), then enter the trade.
Also, notice how much more stable the 28 is. This is because of the measure of the trend, itself. The 20 oscillates all over because it does not account for the entire movement of the whole trend.
If there are too many candles accounted for, then it does not oscillate enough. In other words, it is too stable (e.g-plot the 100).
I believe in using the BB's, you want an indicator that will measure the movement of a trend once it has gotten to an extreme condition, so you can place a trade going the opposite direction in order to optimize the most profit from it.
BTW, nice charts in showing the normal distribution of the SD's. There is one slight difference. There is a greater percentage of sigma 1 and sigma 2 being hit, because this is a constant move SD. It is not arbitrary data, as an example, the height of all men in a classroom. So, the 68% ( 1 sigma), 95% (2 sigma), and 97% (3 sigma) will be much less. That is easy to test. Just plot 100 candles, then add up the times the 1'2, 2's, and 3's are hit.
bus;513159]Hi 4xpipcounter,
Though I have not tried BB(28),which I think is too WIDE, I posted BB(20)
here:
http://www.traderji.com/day-trading/48558-bollingerworm-3.html#post511623[/QUOTE]
http://img267.imageshack.us/i/283.png/
Columbus, that's why I mentioned something to the effect of plotting the 20 and 28 on your demo account. 28 may seem to wide, but it takes into account the full cycle (Which sometimes can be up to 32 and even 34.). The 20 has a lot of action with the candles riding on the upper or lower bands, which, IMO, renders the bands useless.
Also, I prefer the 28,3 over the 28,2, because 3 sigma is more extreme and less likely to be hit, which renders a better trading opportunity.
Attached, are a couple of charts of the GBP/USD. One with the 28,3, and the other with the default, 20,2. The rule with the 28,3 is to wait for the break from the band, then enter going the opposite direction, and wait for at least the mean to be hit. Or, if the extremity is exceeded, then wait for the candle opposite the direction it is traveling in (E.g--Bear candle in an uptrend.), then enter the trade.
Also, notice how much more stable the 28 is. This is because of the measure of the trend, itself. The 20 oscillates all over because it does not account for the entire movement of the whole trend.
If there are too many candles accounted for, then it does not oscillate enough. In other words, it is too stable (e.g-plot the 100).
I believe in using the BB's, you want an indicator that will measure the movement of a trend once it has gotten to an extreme condition, so you can place a trade going the opposite direction in order to optimize the most profit from it.
BTW, nice charts in showing the normal distribution of the SD's. There is one slight difference. There is a greater percentage of sigma 1 and sigma 2 being hit, because this is a constant move SD. It is not arbitrary data, as an example, the height of all men in a classroom. So, the 68% ( 1 sigma), 95% (2 sigma), and 97% (3 sigma) will be much less. That is easy to test. Just plot 100 candles, then add up the times the 1'2, 2's, and 3's are hit.
bus;513159]Hi 4xpipcounter,
Though I have not tried BB(28),which I think is too WIDE, I posted BB(20)
here:
http://www.traderji.com/day-trading/48558-bollingerworm-3.html#post511623[/QUOTE]
Last edited:
| Thread starter | Similar threads | Forum | Replies | Date |
|---|---|---|---|---|
| B | _bollingerbands overlay | AmiBroker | 0 |