Evaluating Volume-price confirmation versus On-balance volume and MACD

A second vote, not a first signal

This piece reviews an archive evaluation design for Volume-price analysis, On-balance volume, and MACD. The archive facts describe a historical workflow. Editorial comments are labelled as such and are not archive claims.

Editorial interpretation: evaluate a volume overlay as a second vote, not a first signal. Lock trend intensity and price direction first, then ask whether participation agrees.

How Volume-price analysis is formed

Volume-price analysis is formed by subtracting a Simple moving average from a Volume-weighted moving average of the same lookback, reading a positive gap as Volume-price confirmation and a negative gap as Volume-price contradiction.

The confirmation reading multiplies that long-horizon gap by a short-horizon volume-weighted-to-simple average ratio and then by a short-to-long average-volume factor that overweights rising participation and underweights falling participation.

In the two-day arithmetic example, closes of 10 then 12 with volumes of 100,000 then 300,000 produce a Simple moving average of 11 and a Volume-weighted moving average of 11.5.

Four participation states

Four pairings of price change and volume change are treated as distinct states: rising price with rising volume, falling price with falling volume, rising price with falling volume, and falling price with rising volume.

Confirmation cues

Described confirmation cues include a reading above zero, a rising or falling slope, a smoothed volume-weighted average of the reading as a momentum trigger, and a V-shaped reclaim after the reading drops through the lower band of an envelope around itself.

On-balance volume asks a different question

On-balance volume is described as a running total that adds volume on up closes and subtracts volume on down closes, with the stated aim of exposing hidden coils in otherwise quiet markets.

Volume-price analysis is distinguished from On-balance volume as a fixed-window snapshot of how volume weights a price trend, intended to judge the health of an already trending market rather than to accumulate history in a trendless one.

The evaluation design

The evaluation design took long entries only when a seven-period ADX classified the market as trending, a 12-26-9 MACD supplied direction, and either an On-balance volume or Volume-price confirmation crossover of its own smoothed trigger agreed.

The comparison was run on the S&P 500 exchange-traded fund from February 1993 through 2006 using unoptimized standard settings: a 28-period On-balance volume series with an eight-period smooth, versus a 7/28 Volume-price confirmation setup.

A scaling extension

One described extension is to scale another series, such as a trailing average of lows or a momentum reading, by the confirmation value so that agreement and contradiction move the scaled level in opposite directions.