Volume based Technical Analysis

(terminology and basic principles)

Terminology

Before starting our discussion on how volume patterns drive index reversals, we need to define several terms.

Volume Moving Averages (VMA): Every trader is familiar with moving averages of securities prices, perhaps the most frequently used technical indicator. We simply apply the concept to volume, rather than to price, and plot Volume Moving Averages (VMA) that range in duration from as short as a few minutes to as long as several months. However, there is a slight twist to this: Volume activity typically follows certain predictable patterns throughout the trading day, with high levels prevalent immediately after the open, lower values around noon, and increased levels once more toward the close. We call this pattern the "time factor". Unfortunately, the time factor provides a rather distorted picture of the daily volume activity. It makes it difficult to differentiate those volume events, which are truly significant, from those that are simply part of the normal daily fluctuations. We have solved the time factor issue by normalizing volume data before charting it. Charting normalized volume allows a much clearer determination of whether or not volume levels are spiking above normal levels, an aspect that is at the core of our methodology.

We are particularly interested in the appearance of large peaks ("spikes" in the VMA - known as VMA spikes - and how an index reacts when they are generated. Sudden VMA surges are indicative of bursts of significant buying or selling activity. As such spikes occur, we determine whether the index is moving up or down at that time. If the direction is up, we call the associated volume surge a resistive VMA spike; if the index direction is down, we label the spike a supportive VMA spike. In the absence of distinct volume spikes, we still call any volume generated as the index is moving up resistive volume, as it moves down, supportive volume.

Basic principles

The most basic premise of volume analytics is that we can always anticipate an index will react to (significant) volume spikes - as a rule, resistive volume spikes will force a downward move in the index; supportive volume spikes will generate upward index momentum. This basic assertion must be qualified by two key questions:

• What determines the extent and characteristics of an anticipated move: Will it be short-lived or have "staying power" over the mid- to long-term? Will it be gradual, sudden, or volatile?
• What determines when an anticipated move will most likely occur: Will it happen immediately, promptly, or will there be a certain time lag (a "delayed volume reaction")?

Our research shows that the answers to these questions vary considerably, depending on (a) the general market context, and (b) the technical characteristics of the actual volume spike(s) being analyzed. Therefore, in order to get the most value from volume analytics, it must always be placed in the proper context:

Market context: Where in the larger market picture do supportive / resistive VMA spikes appear: During short-term pullbacks within a larger uptrend? As part of short-term upside corrections within a larger downtrend? At the presumed end of a weakening long-term trend? At the beginning of a new trend or somewhere in its middle? During distinct trend runs or in markets with choppy sideways trading action (i.e., in support / resistance corridors)?

Technical considerations: When analyzing a VMA spike, consider its magnitude, both vertically (the height of a thrust) and horizontally (its width or breadth). Comparatively larger and / or wider spikes obviously carry more weight. Caution must be exercised when analyzing volume spikes on a short time frame, as their potential impacts on mid- or long-term trends can easily be misjudged. A noteworthy spike appearing on a 5-minute chart could well affect an index in the short-term, but it may not necessarily have much of an impact on the prevailing long-term trend. A spike may look imposing and appear to be critical on 1-day chart, yet it may not loom as large on a 30-day chart or even seem significant at all on a 60-day chart.