© Andrew Coles       Terms & Conditions | Privacy

Modified VWAP Methodologies

New Research: Essay Six

Essay on MIDAS/AC Fourth Generation (Gen-4) Curves

© Andrew Coles


The first purpose of this essay is to draw together work published on my Gen-4 curves in Chapter 16 of the MIDAS book and in the June and July 2011 editions of Active Trader magazine ("MIDAS Curves Without VWAP: New Methods for Mapping the Hidden Changes in Trader and Investor Psychology”, parts 1 and 2). Although readers of the book will be aware of Chapter 16, far fewer people in touch with the MIDAS system will know about the broad application of Gen-4 curves in the Active Trader articles. This essay will draw out most of the important points from these three overlapping studies.

A second is to emphasize even further just how extensively Gen-4 curves can be applied to financial datasets. This versatile and extensive application can give rise to some remarkable market signals (and trading systems) that cannot be duplicated in other indicators or indeed in any other generation of MIDAS curve. MIDAS/AC Gen-4 curves thus represent another exciting development in the MIDAS project and provide extensive opportunities for new types of analysis.

A third purpose of this essay is to draw attention to important and recurrent behaviours in Gen-4 MIDAS curves on whatever dataset they are applied. I’ll illustrate this behaviour on numerous charts below. I first discussed one of these recurrent behaviours in Chapter 16 of the book in relation to the OBV indicator. There I called it the MIDAS/AC Dipper Setup and I also used this same phrase to describe the behaviour in the two Active Trader articles. In this Essay I’ll also be using the same phrase while explaining its meaning.

Gen-4 Curves and Paul Levine’s Philosophy of Price Movement

Part of my motivation in developing Gen-4 curves was the belief that Paul Levine’s philosophy of price movement could be extended much further than he had anticipated in his early lectures in the mid-1990s.

As discussed in the MIDAS book and other articles, Levine believed that the anchoring of MIDAS curves to trend reversals is such a fundamental methodology because the reversals represent major changes in market psychology. These psychological changes are consequently linked to ongoing changes in market psychology at every degree of price trend (hence the fractal basis of the MIDAS system). Moreover, because these ongoing changes are also trend reversals they are continuously supported or resisted as key price levels by the MIDAS curves. If (through eventual displacement) there is no support or resistance by the curves, these same levels (inflection points) simply become candidates for the launching of new anchored MIDAS curves.

While this philosophy is entirely reasonable, it is consistent with the possibility that there are other inflection points that are either hidden from the price trend or completely out of reach of it. At the time of writing the book, the obvious way of testing this assumption was to find an alternative dataset that was (a) similar to the price trend in also being trendlike and fractal, while (b) also being sufficiently different from it to possess the inflection points that were either hidden or out of reach on the price trend.

An obvious candidate for such a dataset was Joseph Granville’s On Balance Volume (OBV) indicator in so far as it easily satisfies requirements (a) and (b). In Chapter 16 I had great success in applying Gen-4 curves to OBV, but I also realised that there were many other datasets that satisfied (a) and (b) and even certain datasets – such as economic datatsets – that had nothing directly to do with market price trends while still having much broader forecasting utility in the financial markets. One of the aims of this essay is to look at some of these datasets.

The Makeup of Gen-4 Curves

Gen-4 curves can either contain market volume (in which case, they are also Gen-1 curves) or can be constructed from the Gen-2 volume methodology. Their crucial difference from earlier curves is that they’re obviously plotted on datasets other than the actual price trend. This is why Gen-4 curves are such a radical departure.

A Brief Survey of Current Gen-4 curves

As noted, Gen-4 curves began as applications to the OBV indicator. Later in the Active Trader article I began applying them to momentum indicators, volatility indicators, market sentiment indicators, as well as economic indicators. The following list highlights how extensive the application of Gen-4 curves has now become. Because Gen-4 curves are launched from normally hidden or out of reach areas on a price trend, they can - as emphasized earlier - create remarkable signals while also generating extremely interesting trading systems that are impossible to duplicate in other technical approaches, including standard MIDAS curves.

Here is a list of current curves:

  1. MIDAS/AC Gen-4 Volume Curves: Datasets with fractal trend characteristics other than price such as OBV or Accumulation/Distribution

  1. MIDAS/AC Gen-4 Momentum Curves: Unbound momentum oscillators such as Gerald Apple’s MACD (Moving Average Convergence Divergence)

  1. MIDAS/AC Gen-4 Momentum Curves: A limited number of bound momentum oscillators such as Welles Wilder’s Relative Strength Index (RSI)

  1. MIDAS/AC Gen-4 Economic Curves: Fractal trending economic time series such as the Baltic Dry Index

  1. MIDAS/AC Gen-4 Volatility Curves: Volatility indicators such as the VIX (subject to additional treatments not discussed here)

  1. MIDAS/AC Gen-4 Sentiment Curves: Market sentiment gauges such as the Put/Call Ratio (subject to additional treatments not discussed here)

  1. MIDAS/AC Gen-4 Rate Curves: Yield curve studies in the bond markets

  1. MIDAS/AC Gen-4 Spread Curves: Other financial gauges such as the TED Spread, the LIBOR/OIS Spread, futures contracts, etc

  1. MIDAS/AC Gen-4 RS Curves: Relative Strength (RS) Lines popularised by John Murphy and Martin Pring

A More Detailed Look at the Selected Curves

In the remainder of this essay I’ll focus briefly on the following:

I’m omitting Volatility and Sentiment Curves because the additional work done to the volatility and sentiment data prior to the application of Gen-4 curves for the moment remains proprietary.

MIDAS/AC Gen-4 Volume Curves (OBV, Accumulation/Distribution)

First, Figure 1 below illustrates the effectiveness of applying Gen-4 curves to On Balance Volume in the upper pane of the chart. Because its fractal trend characteristics are (crucially) distinct from price, the OBV indicator always has an extensive number of hidden inflection points (HIPs) and (more importantly) out of reach inflection points (RIPs). It is thus an excellent dataset for Gen-4 curves.

All the chart illustrations in Figure 1 have Granville’s OBV in the upper pane and futures prices in the lower pane. The charts are weekly and therefore the futures price trend is Primary (ie, 9 months to 2 years) and Secular (up to 10 years or more). The reader can compare the two curves launched from roughly the same time period and then compare how in the upper pane the Gen-4 curve captures the ongoing trend while the standard MIDAS curve in the lower pain fails to do so.

Figure 1: Gen-4 curves applied to Granville’s On Balance Volume in the upper pane and successfully capturing ongoing inflection points that are out of reach on the price chart

On the next page we’ll look at the application of Gen-4 curves to economic indicators, illustrating with the Baltic Dry Index, and then take a look at one or two other applications.  

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