© Andrew Coles       Terms & Conditions | Privacy

Modified VWAP Methodologies
Share on Facebook
Share on Twitter
Share via e-mail
Share on Delicious
Share on Digg
Share on Google Bookmarks
Share on LiveJournal
Share on Newsvine
Share on Reddit
Share on Stumble Upon

Essay two continued

Summary of blog post - types of curve continued

Third generation curves

Third generation curves are essentially Calibrated Curves. The rationale behind them and their application was covered in David Hawkins’ Chapter 9 of the book. He has not published on these curves elsewhere. Calibrated Curves can be either first generation curves (that is, they can be created by using market volume) or they can be second generation (that is, they can be plotted by using artificially inputted nominal volume). I’ll have more to say about these curves below.

Fourth generation curves

Fourth generation curves were developed by Andrew Coles and apply to any time series other than price provided it demonstrates the same fractal trend characteristics as a price trend. In my Chapter 16 of the book, I applied a fourth generation curve for reasons explained there to the On Balance Volume (OBV) indicator. Unfortunately there was no additional room in the book to illustrate these curves on other time series but I was grateful to Active Trader magazine for providing me with the opportunity to  illustrate them on momentum indicators such as the MACD, volatility (VIX), as well as familiar economic indicators such as the Baltic Dry Index.

It stands to reason that, like third generation curves, fourth generation curves can also be first or second generation curves. For example, if fourth generation curves are plotted on financial market indicators such as the MACD or OBV, they can be plotted using market volume or nominal volume. On the other hand, when fourth generation curves are plotted on economic indicators they will have to be created using nominally inputted volume and hence share the properties of second generation curves. Again I’ll have more to say about these curves below.

Fifth generation curves

Fifth generation curves were developed by Bob English, once of the Precision Report, and are discussed in English’s contribution in Chapter 17 of the book. Fifth generation curves consist of two curves whose plotting is related, namely MIDAS Average Curves and MIDAS Delta Curves.

Once a first or second generation curve is plotted, English takes the average of this curve to create what he calls the Average Curve. The Delta Curve is plotted equidistant from the first or second generation curve on its other side. I’ll return to these curves below.

It’s important to stress again that the term “generation” is not indicating the superiority of some curves over others. It’s merely a chronological designation. Each generation of curve has unique features pertinent either to differing market conditions or to more fundamental factors such as the absence of volume in certain markets or certain data types.

Summary of blog post - types of indicator

In the same blog post I (Andrew Coles) also discussed briefly indicators that had emerged during the MIDAS project. There was less of a need to do this in the post because, unlike the curves, the indicator rationale is easier to appreciate for newcomers. Nonetheless, because the creation of the indicators is intimately linked to the curve formulae, it was worth emphasizing that the properties of indicators varied depending on which generation of curve they consisted of. The following indicators can be created by all five curve generations:

Beyond the crucial point that each indicator is a function of the generation of curve that created it, I won’t be adding anything further on MIDAS indicators here.

Further comments on the five generations of MIDAS curve

Because detailed analysis occurs in the relevant chapters of the book, I (AC) will continue the orientation in this second essay with only the briefest of remarks. Moreover, because the rationale of each generation of curve is offset against first generation curves, there is little point in summarising what most people already know about first generation curves. Consequently I’ll start with second generation curves.

previous page next page