|Discussion of Howard's Moving Average|
Howard's Discussion on His Moving Average
18:31 Ensign Howard: On 9 of our studies that involve some kind of average, the study property form for the study will have a Formula selection for chosing the average formula to use in the study.
18:32 Ensign Howard: For example, the Kelter channel is an average true range offset from a center line which is an average. So there are 2 averages in this study, one to average the True Range for the band offset from the center line and one to calculate the center line.
18:34 Ensign Howard: Now, in working with average, which we had as Simple, Exponential, Smoothed, Weighted, and Volume Weighted, I realized that there was the possibility for another way to calculate averages, that would incorporate a bit of an adjustment based on the acceleration or first derivative of the slope of the average line.
18:36 Ensign Howard: All of these other averaged just take a set of input points and crunch an average. My clever average does the same, but also considers the rate of change of the data points
18:36 Ensign Howard: Let's do a trivial example
18:37 Ensign Howard: If we do a 3 number simple average of 3, 5, 10 the simple average is 6 because the sum is 18 and 18 / 3 elements is a 6. But, there is acceleration in the step from 5 to 10 which is a bigger change then the step from 3 to 5. So if we to 'forecast' the next number, the forecast would suggest we should have a number higher than the simple average 6, because we are accelerating. That is kind of what I am factoring in, is some consideration of the acceleration. You can say we connect the data points to make a curve, and the first derivative of the curve is its slope. If the curve is velosity, then the slope of the curve is acceleration.
8:41 Ensign Howard: Anyway, that is all I want to say about Howard's Average. I do not want my genius to be robbed and put in the charting applications I compete with. So I have to talk in general terms and keep the specifics to myself.
18:42 Ensign Howard: However, the math for the calculation is not burdensome. The amount of calculation would fall in the neighborhood of the Weighted average.
18:43 Ensign Howard: So if the speed of calculation of all the averages were on a scale from fastest to slowest, I would rank order them in this order.
Smoothed and Exponential, Simple, Howards, Weighted, Volume Weighted, and then the DYO for Hull is much farther away because is it a DYO
18:45 @Buffy2: good to know thanks
18:45 Ensign Howard: The new Howard's MA formula has been retrofitted into Moving Averages, Bollinger Bands, Keltner channel, Uni-Channel, Stochatic, Stochastic Momentum, Triple Average, TRIX Oscillator, and MACD Oscillator. I think that is a complete list of studies that utilize averages in their base formulas.
18:47 Ensign Howard: The nine studies listed have Howard's MA as a selection choice
8:47 Ensign Howard: Others like RSI do not use averages other than we plot a 2nd line which is an average of the RSI, and this 2nd line is typically Exponential, but some studies also offer a check box for Simple as the formula for the 2nd line. But as shown the Average comparison illustration, you will discover that Howard's average is smoother and yet tracks the data a bit better in my opinion. When the market moves up or down quickly, Howard's average does a much better job of staying with the move because we factor in the acceleration effect. You can see that in the comparison chart.
18:49 Ensign Howard: let me post it again.
18:52 Ensign Howard: buffy, in this post you will recognize your B-Swing template, but the Keltner bands have the Howard's average checked and actually end up being a bit smoother in their flow, which I consider a good thing
18:53 Ensign Howard: Just one more option to consider and investigate possible usefulness
18:54 @Buffy2: :-) tks
18:54 Ensign Howard: Anyway, BLinford was asking in a PM what the Howard's MA was about, so I thought it best to give this discussion in public for the benefit of several readers. But, please permit me to be mum about the exact formula (that so far as I know) I have invented.
18:55 @Buffy2: Appreciated Howard -- Thanks for asking BLinford
18:56 BLinford: sure buffy
18:56 Ensign Howard: If a lack of specific details bother a user, well I suggest they just continue with the other standard average formulas that are available in Ensign
Last updated 07/17/2006