I have waited since my last post on Bollinger bands to describe how I used them until I had a real world example to share. BRCM is the stock, timeframe is 15 minute bars off of the open. Previous close was 41.19, Today's open 37.87, a gap down worth $3.32 Value of the lower 15 minute Bollinger Band 39.01 (2 standard deviations/20 period ma) Opening price was beyond 3 SD's. The 2 inherent principles of volatility are: ***Persistancy ***Reversion to the mean How do I capitalize on this potential? I Overlay the 15 minute BB's on my 1 minute chart - I can see that price is well below the band. The two above principles should begin to manifest. I look for my trade setups off of a 40 tick chart, basing entries and exits off of a 10 tick chart in conjunction with L2. I entered at 0944 at 38.24 keying ISLD for 1500 shares, filled for 800 canceling the rest. In addition to the above, price action was strong and constructive on the 40 tick chart. L2 was thick with support on the bids at single and 2 penny spread, each level with thousands of shares, (supportive) if it stalled, I knew I could bail easily. For the next minute, the buyers had control. Steady advancing price, the last price bounced above and back into the inside market, being chased by the inside market, then locked and crossed, followed by my exit at 38.46 at 0947 into oscillator weakness. This was a classic demand driven advance. I successfully used volatility when stretched to it limits to capture a profitable trade. I did not buy the bottom, I did not buy the top. I bought the middle, the momentum. With the lack of range, volume and volatility, tick charts have been my playground. You can see the action within a 2-3-4-5 minute bar, by dropping down a time frame. Ticks can allow you to see inside the 1 minute bar. This and a few other themes are working in this market. The aforementioned represents my opinions and experience. http://www.elitetrader.com/vb/showthread.php?threadid=1718 -Hoyler
Thanks for your detailed example showing exactly how you used BBs to trade BRCM. I find it very interesting and helpful. In your linked message, you explain that Persistency is the continuance of an effect after the cause which first gave rise to it is removed. Would you say more about how this applies in your example? That is, in your example, what is the effect that continued? Plus, what was the cause which first gave rise to it, and what indicates this cause was removed? I understand the concept of regression toward the mean, and I have some understanding of the definition of persistency, but I don't understand its operation in the example you gave or to stock trading in general. Any further explanation will be appreciated. Fast
Hi Fast, I will try and answer your questions. Allow me to go back to the basic's. These are concepts that would be immeasurably easier to understand with repeated visual examples. I am writing from off the top of my head here, these concepts are easier to trade with than explain. Look at a few charts where price has peirced the band. Also note the relative width of the bands. Narrow BB's highlight low volatility & portend a potential move, aligning yourself for the coming move is what works here. Wide BB's indicate a volitile market place, & here is where fading the band works. It would be foolish to say these occurence are fruitfull each and every time, like anything else the can and do fail. But, conceptualy they are correct, offering you and edge. Bollinger bands are a measure of volatility around an average price. Usually 20 periods and 2 standard deviations (SD) Standard deviation, i.e., volatility has some peculiar traits that you as a trader can harness for your benefit. These traits are: ***Reversion to the mean or average price; In this case the 20 period ma. moves outside of the bands are anomalous, and should correct themselves back inside the band toward the mean or average price. ***Persistancy; The continuance of an effect after the cause which first gave rise to it is removed. Once moving in a direction, tends to continue in that direction. This especially true when price is outside the band and attempts to make it's way back towards it average/mean price. ***Autocorrelation; A high degree of. Forecast's of volatility can be highly accurate, more accurate and predictable than price. This particular feature of volatility is the birthmark of the two above principle's in action. All right Fast, how to take this from a laboratory curiosity and put it in to a real world practical application? When looking at a bell shaped distribution curve, you will see that the "majority" of the distribution falls in the middle of the bell. This would be akin to 1 standard deviation. The farther away from the middle of the bell, the less action or distribution is taking place. Concerning volatility when price moves beyond 2 SD (away from the middle of the bell/or average price or mean) We can expect the above inherent principles to begin to work. If 95% of all action is contained by 2 SD's, those times when prices is greater than 2 SD's "should" be ripe for exploitation. Recall the following; 1 standard deviation encompasses 67% of all occurrences 2 standard deviation's encompasses 95% of all occurrences 3 standard deviation's encompasses 97% of all occurrences How this applies to my example. Moves outside the BB's are less reliable the smaller the time frame in use. A 15 minute time frame has solid credentials for allowing me to harness the attributes in question, it is less failure prone than say a 5 minute time frame IMO. When BRCM traded below the 15 minute BB, I knew it had the makings of a high probability direction based trade. You ask what would indicates the cause in no longer in affect. In a perfect world a return to the 20 period ma would say that things are "back to normal" in this case it did not work that way. I sold out when signs of weakness showed. This kind of opportunity works because the rubberband has simply been stretched to far and it now seeking to return to it's original shape. BRCM was in the midst of a sell off with no real news. While CIEN when they reported earnings is very different -- that rubber band has been broken for the time being. If someone else can post simpler examples of explanation I would be obliged. FYI - the Sept 1996 issue of TASC has an outstanding article on Standard Error Bands - Recognize low volatility situation and trend direction. The aforementioned represents my opinions and experience, I hope this helps. -Hoyler
Hoyler, Your response took a lot of time and effort on your part, which I recognize and appreciate. Part of the trouble I was having was that I couldn't see that the concept of persistency added much over the the concept of reversion to the mean. That is, it seemed to me that you could have made the trade just knowing that the price was outside 2SDs and was likely to revert back to the mean. I can see that persistency could come into play when, after gapping down and continuing down for a short time, the price first turned up. After the price turned up, I think the principle of persistency would predict that the price will continue up even if the cause for its turning up is removed. If true, that argues that you could have entered a long position right after price turned up. Strengthening this argument is your statement that persistence in movement toward the mean is especially true when "price is outside the band and attempts to make it's way back towards it average/mean price." Based on what you said, it looks like you don't just depend on these principles; rather, you also check the trading environment to see whether you can get into and out of positions. I think that's smart. Gives you another edge. I like all the edges I can get. Thanks for the reference. I meant to ask whether you could point me to some place where I could read about this to keep you from having to give me a long explanation, but I forgot. Again, thanks for putting in so much time and effort. Your explanation helped a lot and your example of how you used BBs has added to my trading knowledge. The more specific examples I study, the more understanding I have. Fast
First I would like to say "good post." Now I will add a correction to part (only because it is important), and My trading "lives and dies" [lives ] by statistics; so just in case someone is looking to do an evaluation the correct SD's are: 1 Standard Deviation (68.26%) 2 Standard Deviations (95.44%) 3 Standard Deviations (99.73%!)
Fast, You got the idea right ^5 Below are a couple links for bb's and SD. http://www.charthelp.com/ http://trends-online.com/ind15691.htm#Bollinger Bands http://www.bollingerbands.com http://www.traderspress.com RAY, Thnxx you, it has been a good long while since I've read the concepts from a book. Maxito, Glad to hear that! Let us know how it works out for you. Lancer, You read my mind <g> -Hoyler