Profitable ETF Trading Strategies: developing a new oscillator, the NDX()
In my last article I described two problems that I had with the otherwise excellent technical indicator known as Williams %R. I just like the scale which goes from zero to -100 which seems counterintuitive, and I don’t like that uses today’s price action when calculating the index value which hides information concerning breakout days.
I suggested that if I were going to design the indicator from scratch I would fix those two problems by making the scale read normally from 0 to 100 like a thermometer, and I would not use today’s price action in calculating the index value in order to highlight breakout’s above and below the trading range of the look back timeframe.
To support my own trading I have built just such an indicator in Microsoft Excel and I use it for analyzing large numbers of stocks and ETF’s as part of my daily trading practice. I have found the modified indicator which I call NDX or “index” does everything that Williams %R does while adding value by fixing the two problems I have identified.
In my spreadsheet reports to use the column heading of NDX(t), where “t” represents the look back timeframe. For example, NDX(10) would represent a 10 day look back period from yesterday to 11 days ago, with today being represented is the zero day.
If the asset traded in a range from 10 to 20 in the look back timeframe, and today closed at a price of 10, it would have an index value of zero. If the close today were 20, then the index value would be 100. If the close today were 21, the index value would be 110 which would indicate a breakout of 10% greater than the previous trading range.
You can see how this information would be very useful if you are scanning for breakout opportunities in a large population of stocks and ETF’s. The normal 0 to 100 scale is also more intuitive.
I use conditional formatting in Microsoft Excel to highlight index values greater than 90 in green, and less than 10 in red. I typically rank sets of symbols from highest to lowest or lowest to highest, in order to find those symbols at the most extreme condition compared to their peers. This is consistent with the use of oscillators and general and I have found it very useful to focus on these extremes for short-term trading targets.
When I use a year-long look back timeframe, I write the column header as NDX(52w), as you might expect.
I use Williams %R whenever I am using commercial or public stock screeners, but I use my own NDX indicator on my own spreadsheet reports.
These kinds of refinements come from deep understandings and appreciation for the construction, use and limitations of conventional technical analysis. The refinements don’t seem important unless you look very closely at the specific indicator. Attention to detail is rewarded in the practice of trading.