For most indicators (except the MACD), I use the standard settings.

For both Bollinger Bands and Keltner Channels, I use the 20 period lookback. It helps to keep the settings the same, but remember that shorter lookbacks lead to more signals (and more whipsaws) and longer periods lead to fewer signals, but more valid ones when they occur.

One more question, if you don’t mind:

What parameters do you normally use for the bollinger bands and Keltner Channels ?

Be sure to filter down the industries and view the individual chart components in Prophet to get the best out of the site, along with the historical trends. You are looking for money to be flowing into an industry group.

Thank you for the clarifications. I was only able to read the post you referenced and this is helpful to understand your method better.

The closer you can get to price action, the better. Indicators are helpful, but as long as we all realize that they are derivatives of price or volume.

I traded one minute charts for about a year solid before burning out doing that! I’m right there with ya! My rule now is never view a 1-minute chart again!

Regarding the volatility filter, yah I shouldn’t have just thrown that out there without an explination. Refer to market principle #4: The Market Alternates between Range Expansion and Range Contraction. I have not posted regarding this on my blog (but intend to) but Brett Steenbarger at TraderFeed recently posted regarding this issue and its statistical foundation.

A quick overview: When the market experiences price-bar overlap (swing high/low overlap), it is in a trading range. It is winding down to an equilibrium point. Oscillators will work best in this environment and moving averages (as support/resistance) will not. The market will (sometime) break to a trending period, where oscillators will no longer work, positive feedback will kick-in, and moving averages (for S/R) will work best.

What I mean is that oscillators and indicators are often sensitive to the condition of the market – be it range or trend. Also, this is (likely) why there is no magic indicator, because the market alternates states and changes ‘character’.

You can determine this condition by ‘eyeballing price’ or by using the DMI (Directional Moving Index) or the ADX (Average Directional Index). You are looking for a low ADX to identify range conditions and a high ADX to identify trend conditions.

You can also view the distance (squeeze/tightness) of the bollinger bands for volatility contraction and expect the price to breakout when the bands “squeeze” price, especially when they squeeze through Keltner Channels (if you overlay them both). This setup can allow you to play for a large target when price expands out of the “squeeze-play”.

Thanks for the info on sector rotation. It will take me some time to get to the books but the links, specially the one from Prophet is immediately useful for me. It has provided me with another quick method of getting some stocks for my watchlist.

Just wanted to point out that I use EMAs and Williams %R on index charts only to define long/short bias and market timing. I do not use any indicators on stock charts and use only price action. I do have keltenr and 20sma overlays but they are used only to kinda chart a map for the price and not to make any decisions for entries really.

I am planning to ADD hourly charts based trading to my daily routine and not switch from daily. I will still stick to daily and take entries when charts say so. And, as to getting overwhelmed, I have done my stint with trading the opening gap on 1 min charts, tracking ~ 10-15 charts with crazy scanners spewing out data at the same time. Hourly should be fine. Probably won’t even end up being day trades.

Can you also please explain the following from your comment: “Perhaps try to add a volatility filter to determine range/trend axis for increased validity of signals.”

Thanks,
Rahul
http://www.tradernirvana.com

I view candle charts and swing charts depending on my analysis. Swing charts are generally longer term and help me assess market structure and ’swings’ (think waves) in price. They are exclusively price based and I am seeking to determine swing highs or lows.

I have posted a few examples from my swing chart analysis (look at the chart construction, rather than the textual analysis): Momentum Precedes Price, Momentum Divergences, Anticipating a New Swing Low.

Basically, the color comes from a specified average true range function from the lowest or highest price 16 bars ago. Say, 16 bars ago, the price was $20 and the ATR is $1 so if the price now is 2 times the ATR (or $2), then the price would change color.

Actually, the mechanics of the color really aren’t that important – I don’t make decision on color, it just looks prettier and shows me data already there. I am analyzing bar charts to determine more data and look at longer time frames and identifying swing highs and lows in the structure of the price data.

For my decisions, I will only use candle charts.

If this doesn’t make much sense, or needs more clarity, please let me know.

Thank you for the comment!