Summary Essay: Essay Seven (continued)
© Andrew Coles - no unauthorized reproduction
MIDAS Standard Deviation Bands
The MIDAS Standard Deviation Bands are the focus of my Chapter 15 of the MIDAS book.
As stated there, the Bands have a mixed provenance, most likely first emerging in 2003 on a TradeStation forum. A few years later, versions of the Bands emerged in 2007 in Ninja Trader and the anchoring technique was used for the first time. This technique was applied to the Bands in 2007 by the trading platform Investor/RT. In 2009 Bob English also anchored the Bands in a TradeStation version of the indicator. My own contribution was to code the Bands for the first time in the Metastock coding language while also applying the MIDAS formula to them for the first time. As stated in several essays on this site, the MIDAS formula differs from the original VWAP formula, resulting in a curve that is not coextensive.
Although in Chapter 15 I illustrated the use of the Bands in sideways markets as well as in uptrends and downtrends, I expressed in December 2013 a great deal of caution in their application and suggested limiting their use to uncommon patterns known as Broadening Formations (tops and bottoms) and sharp, angular price moves springing out of low volatility conditions or as part of larger zigzag patterns. The relevant discussion can be found in Essay Three here or in a blog post entitled, “Introducing a New Indicator to the MIDAS Canon” (entry: December 16, 2013.)
The reason for my caution is that the Bands’ formula creates a fanning problem. In other words, the curves move out from price (a) too quickly, while (b) doing so much too excessively. In Chapter 15 it was slightly easier to control this problem for two reasons: first, because I (and Bob English) used at least three curves either side of the central standard MIDAS curves; and second, because the degree to which the curves fanned out from the central curve was preset within the formula itself – this allowed the tiniest increments to be used and thus offset to some extent the fanning problem. The major drawback with (b) is that it established absolutely no relationship with the price trend since the increments were random.
The alternative to using preset increments within the formula is to rearrange the coding of the indicator so that users could again “fit” the curves to the pullbacks. The major problem here is that even when the Deviation Bands are fitted to the smallest pullbacks, the fanning problem is often unworkable. The only way of defeating this problem – and even then the success rate is still modest – is to fit the Bands to the tiniest pullbacks that occur within two to three price bars of the launch point. (“Fitting” has to be reintroduced because as I first showed in the methodology of the Displacement Channel, this is the only meaningful way of establishing an ongoing relationship between price and the curves.)
Figure 3 below is a gallery of charts used in Chapter 15 of the book. The charts are intraday index futures and range across timeframes from 60m to 1m. The charts confirm that the MIDAS Standard Deviation Bands still have a role to play when the analyst appreciates the type of patterns they work best on and the vital importance of “fitting” them to the tightest possible pullbacks.
Figure 3: gallery of intraday futures charts with the MIDAS Standard Deviation Bands
MIDAS/AC Normal Deviation Bands
This indicator is the subject of Essay Three here and the blog post, “Introducing a New Indicator to the MIDAS Canon” (entry: December 16, 2013). At present, written work on this new indicator is not being published outside this website.
Work on this indicator was completed in late 2013 and was inspired by the fanning problem in the MIDAS Standard Deviation Bands described above. A wide range of changes and additions were made to the formula of the Standard Deviation Bands, resulting in an indicator that is now very different to the Deviation Bands.
The new indicator has two bands either side of the central MIDAS curve, with the inner bands based on a different formula to the outer bands. Both bands were thought to be necessary because they each catch different aspects of the trend: the wider bands capture the outermost swing highs and lows, while the inner bands capture the more modest internal price moves.
The indicator is also based on a “fitting” procedure, which means that it retains the crucial relationship with the ongoing price movement that was created at the very outset of the trend by means of the “fitting” procedure itself. In this indicator the “fitting” procedure involves not one but three inputs from the user. This ensures the widest possible freedom and fine-tuning during the procedure.
Unlike the MIDAS Standard Deviation Bands, which are best suited for Broadening Formations and sharp angular price moves, the lack of rapid or excess fanning means that the new indicator can be fit to normal trending conditions. Moreover because it avoids the faning problem, it doesn’t displace from the trend as quickly as a Standard MIDAS curve does, which is why it can stay with the trend for three to four times the amount of time than the MIDAS/AC Displacement Channel. Finally, because the new indicator isn’t designed for sideways markets, there is no overlap between the new indicator and the MIDAS/AC Displacement Channel – both have vital roles to play.
I will not say more about this indicator here because of the detailed treatment in my essay covering it.
Figure 4 below is a gallery of charts highlighting the new indicator on daily charts.
Figure 4: the new MIDAS/AC Normal Deviation Bands on daily charts
Figure 5 below is a gallery of charts illustrating the new indicator on intraday charts.
Figure 5: the new MIDAS/AC Normal Deviation Bands on intraday charts
The essay continues overleaf with the MIDAS/AC Quadrating Price Levels and Bob English’s detrended oscillator.
|Andrew Coles Blog|
|David Hawkins Blog|
|Essay One page 1|
|Essay One page 2|
|Essay one page 3|
|Essay one page 4|
|Essay one page 5|
|Essay Two page 1|
|Essay Two page 2|
|Essay Two page 3|
|Essay Two page 4|
|Essay Two page 5|