Hello, I've been researching a system to indicate in advance which days are / will be directional (volatile) and which sideways, and would like some input. The system uses Volatility as defined by the Width of the Bollinger Bands (UpperBB - LowerBB) divided by the BB moving average. If you apply this on intraday data (15m, 5m) on the major indeces (I use DJIA) you will notice oscillations where the market expands (directional) and contracts (sideways). When the above Volatility oscillator is rising, the market is usually directional. When it is falling, or below a certain level, the market is sideways. I'm looking to improve on this indicator, possibly using a Moving Average on it or in combination with the ADX for example or some other indicator. Any thoughts on this? What tools are other people using to get an idea of whether the market will be directional or sideways? This could also be used for "time of day" trading, ie. when is it best to enter the market (expecting a move) and when is it best to sit on the sidelines and wait, if we apply it on 5m data. Any thoughts on that? Thanks in advance, Loukas
If you want to orient to making money, then use the Connors Hayward stuff. The better thing to do is turn it into an absolute indicator from the relatavistic one so you can work with a neutral bias. you can catch the BO's in anticipationquite nicely. That Bollinger stuff is not too swift. you make money off the right side of the price formations and Bollinger is stuck what mistakenly looks like left side stuff.
I should specify that I do have a profitable system to make money, using oscillators (MACD, Stoch, RSI, Price ROC etc) and other indications (formations, candlesticks, support/resistance / retracement) to get an initial idea of where the market is headed, and then a breakout system to enter the market, along with trailing stop / stop loss to protect my capital and proper position sizing. BUT, I'm looking to improve the % winning of my trades, by avoiding (or scaling down) in sideways days and focusing on pyramiding in directional days. That's why the original post. Thanks.
Connors Hayward is a leading indicator. When you use it, you get an advantage ina manner that allows you to anticipate the market. Your set up is comprehensive and I am sure that you bracket entry on BO's and that you deal with failures to BO after a BO begins by using a reversal strategy. Form your comments, I felt I gleaned that you may be surprised on occasion from using primarily what you mentioned in the first post. I simply offered a thing to bridge over the situation for you. since it was a biased indicator I suggested how to remove the bias as well since you seem to have a neutral bias. Those who do not appreciate the possibility of C-H being a leading indicator, can give it a try and come to see that as the compression poops out, they are sitting on the edge of a BO coming up. Their are bunches of indicators that expire periodically. The C-H made absolute is one of them. ennds of lateral moves are very impotant simply because the remaining two trend alternatives both make money. By using bracket entries you give up predicting and thus assure that your entry will make money. A failure to BO just makes more money by using the indicator set you have to trigger your reversal.
Use the front end of the C-H to establish that a lateral is coming up. You may want to tune to another fractal to slalom the lateral.
This is a really tough question. On a "quiet" day I think your method is on target or approximately. But you have to take into account shocks: news surprise, M&A bid, and a plethora of earnings and numbers .... what this all means is that on a given day, whether the trading range is directional or sideway is contingent on the shocks, as evident in the last two session. (Sideway followed by aftermarket shock). Keep us posted on the result.
As for me I can tell looking at my model for example see http://www.elitetrader.com/vb/showthread.php?threadid=19770 You can see a dreadfull straight green bearish line from 9209 to 9069 that we began to attack since the low has crossed the segment so this is indicative of big vol for the session that was coming.
is this described in http://www.amazon.com/exec/obidos/tg/detail/-/1557389004 do you have code available maybe?
plenty more on goole "Volatility for me is complex and rich as a method or as an adjunct to a person's algorithm.. I have posted on it before and maybe those posts are available. A good read that would be much more rewarding to anyone here than googling would be "Investment Secrets of a Hedge Fund Manager" by Laurence A. Connors & Blake E. Hayward. The results appear on pages 126 to 151 (An appendix related to Chapter Six (beginning page 113)) and they are an outgrowth of volatility work On page 117 you will find the VIX indicator which Haley Tosis can use to succeed with his ES. You will see that by using a leading indicator from The NAZ 100 it is possible to anticipate moves in ES (the work is with the S&P futures rather than ES but I suggest that there is no statistically significant difference.) I commented that there are results because as of late there is a continuous accusation here that no one posts results. The vast resources of futures and commodities are replete with results for almost all the classes of methods posted here. So I suggested they are there for volatility specifically then I took the trouble to qualify them."