Hi. Some people may recognize me from Yahoo Answers investing section where I have been using my brand of TA there for at least 4 years. Initially I was assaulted with every post I made as to the validity of TA (we get little respect {sigh}) but within a year I got grudging respect and was happy there...but now with the flawed redesign of the yahoo site it is not fun anymore. My philosophy is that there is NOT one magic bullet indicator. The all "lie" at one time or another. However, it is rare that they all lie at the same time. So the object of the game for me is to choose indicators that do not compete with each other (for example TRIX and MACD basically give the same information...pick one or the other) and then arrive at a consensus of the clues they provide. As a result of about 20 years of study I have arrived at 4-5 charts each giving a different perspective on a stock's position. My trading style is long term swing to long term as defined as a hold from 3 months to 3+ years. I do NOT day trade but do comment on such trades from time to time from a TA perspective (day traders with no TA skills are gamblers IMHO) So this is my introduction...the following posts will discuss various indicators and chart formats in detail...I like sharing and constructive comments...I have no secrets (unlike a post I see here that asks for comments on their indicators without explaining how they work...how can you constructively comment on that) I don't say my way is the only way....but it works for me
I will start this discussion with Bollinger bands. Books that I have read really do this overlay a disservice....You are lucky to read more than one page in most books on this topic. Most just say that Bollinger bands (BB's) are a sign of volativity... nothing more but to me there is a lot more you can do with this overlay...later I will combine this with its indicator BBWidth which gives this more power. I have chosen BAC as a ticker here as it demonstrates several points of BB's notice that major price changes occur when the BB's are tight...that is a given but how do you judge the direction of the break out?? that is the question. My general rule for this is to judge the position of the share price relative to the 20 day sma as the BB's tighten...above the 20day SMA there is a good chance the breakout will be positive...below the 20daySMA the price will have a severe correction. This is shown on Nov 11 and again on Dec 30.....right now the price is below the 20daysma so the prediction is a drop in price is coming (not a difficult thing to say right now Is this reliable by itself...not at all but as you will see later I will combine this with other clues (it does get better. I will point to prices rising above the upperBB...the pseudo math is that the upper BB represents 95% of possible results of deviations from the mean...so it is in rarefied atmosphere...same as when the share price is below the lowerBB. A share price rarely stays out there for more than 3-5 days before there is a pullback....a stock that is miles above or below the BB envelope will most certainly pullback and often signal the end of a run. You can see this on Sept 29 for lowerBB and again for upper BB Jan 4.....again as you will see combining this clue with others make it more reliable Finally...once a run is over....the share price trends to the 20daySMA...the chart until November is just random movement between resistance/support but Nov 28 is the end of the run...and in consolidation and you see the price trend toward the dotted green line. Again you see this on Jan 15....from there it oscillates above and below this 20daySMA til the BB's get tight again. this is a general overview of Bollinger bands...if you have comments or questions on this please do so....once we have dealt with them I will show how adding BBwidth makes predictions easier.
ok...ok...nothing earth shattering in that previous post so let's go one step further and discuss the use of the BBwidth indicator with Bollinger bands several problems happen if you rely exclusively on the BB's to give you all the information and not any head fakes...and head fakes often happen but there are many ways to identify them. The chart that I am building now for you is a buy/sell chart...it is NOT a sentiment chart...our objective here is to time purchases and sales. the BBwidth indicator helps us in this regard...and more so as we add other stuff to the chart later the BBWidth indicator in layman's language is related to the distance between the upper and lower BB's. One of the difficult thing to do at critical times is "are the BB's still converging/diverging or are they changing direction (IE from converging to diverging or the reverse)...Why is this important? because this is where "action points" occur...not subtle little changes but massive jumps. Here is our BAC chart with BBwidth added the two green lines at the bottom are the range of BBwith where the major buy/sell points occur In this case the BBwidth is between 5 - 7 at any time when the BBwidth is in this range a buy signal or a short (if negative) can be given. Now looking at Price alone you can see a couple of head-fakes at #1 and #2. How do I suspect them to be head-fakes?? because look at the response of the BBwidth at these times...lethargic or sloping down.... Now look at the BBwidth at #3 and #4...see how strong and positive that movement of the BBWidth is this time....that is a buy signal...(though as you will see we can improve on the reliability of such a signal as we add to the chart. NOW IF WE WERE A SHORT TERM SWING TRADER...Where is the sell signal?? (This works well in day trading too) Look at #5...see the rounding top of the BBwidth??? when the top has rounded and there is a definite negative slope....That is the Sell signal We can predict the place to sell using just Bollinger bands and BBwidth though it will get even more reliable as we add to this chart later. #6 line shows the BBwidth looking like it wants to turn a negtive slope and you might pull the trigger there but if you waited a day you would have seen the BBwidth flatten and not fall until it reached line #7...now it is in clear decline so you NOW sell with confidence. Just learning the interactions of these two TA features will improve your buying and selling...even in day-trading. Don't believe me....test this chart yourself, you can see the parameters I use for BB's and BBwidth on the chart...they are pretty traditional...I experimented with other parameters but they worked no better and sometimes worse. Comments? Questions?....if not we will move on in a a day or so. (I like questions/comments as it helps me know if I got my point across.
Keltner Channels versus Bollinger Bands "There are two differences between Keltner Channels and Bollinger Bands. First, Keltner Channels are smoother than Bollinger Bands because the width of the Bollinger Bands is based on the standard deviation, which is more volatile than the Average True Range (ATR). Many consider this a plus because it creates a more constant width. This makes Keltner Channels well suited for trend following and trend identification. Second, Keltner Channels also use an exponential moving average, which is more sensitive than the simple moving average used in Bollinger Bands." http://stockcharts.com/school/doku....icators:keltner_channels#versus_bollinger_ban
I suppose from the quotation you posted, you are suggesting that Keltner Channels are superior to Bollinger bands??? You make no comments on them yourself. Well let us see....Both overlays are designed such that most share price values are supposed to lie with in their boundaries and if the share price drifts beyond the boundaries then they are in rarefied country. "Indicators based on channels, bands and envelopes are designed to encompass most price action. Therefore, moves above or below the channel lines warrant attention because they are relatively rare." well let us look at a graphic comparison the red envelope is Bollinger bands and Keltner bands are blue Notice that where it counts for a trigger tightness both overlays occur at about the same time. But look at data from Nov 11 to Nov 28...until the very end the price was basically climbing the upper Bollinger Band...very easy to interpret...with the Keltner Bands the price was above the upper bollinger band basically the whole distance...Are we in rarefied country in Keltner bands??? again it happens in January...at what point do we predict the end of the run??? I see no way....though with the BBs it is easier to see. I think that every time there is a strong breakout the price will shoot beyond the Keltner bands (upper or lower) with BB's the upper or lower bands are breached seriously ONLY with extreme surges, which to me are worth noting. Also there is no companion index equivalent to the BBwidth for Keltner...as you see so far With the BBwidth I can make a now rough estimate as to the place where to expect a run to happen (which a bbwidth falls to a certain level) AND I can predict the end of a run...the peaking and decline of the BBwidth. I see Keltner bands as being developed to make the constriction of the bands prior to the breakouts....nothing more. With BBs and BBwidth I get so much more information and as you will see later...I will integrate a BBwidth with 2 other indicators to make Bollinger bands and BBwidth even more powerful. I do confess not to use Keltner bands....it is not necessary to employ EVERY indicator and overlay...It is necessary to understand as best you can the indicators you choose and how they interact with others. But I am open to comments and education....can you show me other than having tighter bands (which I have already shown as a limitation...not an advantage for me) for Keltner bands can be used for prediction??? I will guarantee I can show you more yet on how to use BBwidth when I introduce other indicators.
Nah, I have no use for bands, channels or envelopes. But you said you wanted comments so I commented. Exactly. Which is why some technical traders design our own, rather than trying to make due with some cookbook indicators designed by people more involved in selling books than in trading. Old wisdom: if you want a job done right, do it yourself. New wisdom: if you want a technical indicator that does what you want and doesn't do what you don't want, design it yourself.
Fair enough....I did look briefly at Keltner Bands but I found they did not give me anything that BB's did not give. A big feature of Keltner bands are narrower and if they are outside those bands then the trend is strong.... Well fine...the BB's show that when they track either the upper or lower band...But BB's give more info since it is rare for the prices to lie above the upper BB or below the lower BB...this is very unnatural and when it happens you can usually anticipate a pullback depending on the distance from the band within 1-5days which is useful if you want to add to your position... How often have you seen pumped stocks with pumpers screaming that the stock is going to double or triple from where it currently lies which is well above the upper BB...you know that is a crock so if you did have the stock for some reason (maybe you play pumped stocks??) then you know it is time to exit your position asap. You don't really see that in Keltner bands as the price is always above the upper band in a strong uptrend. Also the lack of a companion index like the BBwidth for BB's you cannot do the things that I will show you later. I have created a couple of indicator charts myself, based on a consumer quadrant analysis and they are interesting to watch but they are a lot of work to update if you have a decent stable of watchlist + invested stocks . It is a hobby with me but conventional Overlays and Indicators are useful I have studied them for 20 years and arrived at a nice suite of them in 4 types of charts...currently I am showing the building of one chart and the components of it...it is a winner chart with some modified components but I will explain why I modified the parameters.
MACD - why I hate traditional usage and How I use it in a chart When I came to developing I wanted a momentum oscillator and MACD was an obvious choice...I could have chosen TRIX but when you are looking for a consensus among indicators...never choose more than one indicator of the same type. Trix gives basically the same signs that MACD does though MACD in the traditional sense employs a signal line. MACD is simply the subtraction of the 12-day EMA from the 26-day EMA with a 9 day EMa signal line....you can play with the parameters but I found the default parameters are fine I very much dislike the signal line of the MACD...buy/sell signals are usually a result of the signal line crossing/recrossing the MACD line. here is our friend BAC again Here I show a comparison between traditional MACD and MACD with no signal line. If you religiously follow traditional MACD you get an awful lot of buy/sell signals...most of which are pretty well useless IMHO. Especially on the sell side, I fine that the signal lags to the point that if you executed the sale when the signal line crosses above the MACD you have lost serious profit. By itself I much prefer to trend the basic MACD values...linking peak to peak and valley to valley. Note that between Sept to end of Nov the peaks of the MACD are progressively lower....this is not a range that I consider BUY/SELL signals to be valid.... breaking that dotted Blue line is a Buy Signal...though not a perfect one but as you well see we can improve on that....the sell signal comes with the peaking of that MACD on Dec 1 The higher low on Dec 29 is another buy....but note that the peak did not rise above the last peak...a certain sell sign and when you see the MACD fall below the extended trend lines at the present time....this is another clue that the future of the BAC price performance is not looking good. See how we are accumulating one clue after another...some times the clue might be BULLISH other times they may be Bearish....at the end of the analysis you cancel out bear/bull and what is left over is the consensus. OK...so you can see that by itself I perfer to trend a basic MACD rather than use it as described in most if not all books. BUT How can we make MACD more usefull??? That happens when we combine it with BBwidth as in this chart there are only 2 clear buy/sell signals...see how the MACD leads but wait for the BBwidth confirmation ...as you can see sometimes the MACD starts to rise but the BBwidth is falling as shown on Oct 9....sometimes the BBwidth is rising but the MACD falls...even though the price rises...that shows a headfake. Also note that you judge the end of a swing run by the simultaneous reversal of the BBwidth and MACD....if only one turns but not the other....the run may continue (another indicator will help that decision (as well as other charts too)) So by looking at the interaction of BBWidth/Price/Bollinger bands/MACD we get a much better Buy/Sell Signal....I can even take this chart one set more by adding another indicator. this will help with the buy/sell aspect but also help in deciding if the end of the swing run is worth waiting to see if it will run again NOTE: the circle at the far right....the price is falling, the MACD is falling, the BBWidth is rising....this is an indication of a correction in price... let us finish this with looking at a case where the price plummets instead of rises (shorting) Now I do not short stocks...but I know many do so...if I did decide to short this is what I would look for the MACD is falling but you wait til the BBWith rises on Sept 17....most shorters would probably cover the short on Sept 23 or so....but the BBwidth is still rising and the MACD is falling....so you wait....on Oct 7 the BBwidth is flatlined but MACD is falling...the short is still on.....on Oct 11 there is a hint of a downturn in the MACD and BBwidth so you could cover the short there except the MACD is still falling...the best place is a few days later on Oct 16 when MACD RISES and BBWidth falls....I call it a MACD/BBwidth PINCH This is not necessarily the best place to cover but it was a pretty good return and turned out to be a good cover. Anyway....this is how I use MACD and how I apply it to my BUY/SELL chart.....still have another indicator to add (tomorrow probably) Check out this chart on your own...you will be surprised how well it works. Questions?? (Please show a sample chart or at least the ticker of the problem if you have one.)
I don't backtest this chart any more....this chart I have used for at least 5 years. It is reliable for me....test it yourself ....you may not find everything as clear as I show it....but I doubt you will find a false buy or sell.....for a swing trade...long term holds require more charts