1993issue C061-4

A volatility-direction filter for dual moving-average crossovers

Keep a dual moving-average crossover as the timing clock, then gate it with a relative volatility index built from a lookback standard deviation of closes rather than daily net price change. Run the same gate on the relative strength index to see when a second reading is only restating the first.

A moving-average crossover remains the timing clock. The second oscillator is a confirming filter that accepts, delays, or exits that clock rather than originating the trade.
The relative volatility index applies the relative-strength calculation to a 10-day standard deviation of closing prices, so it reports directional volatility rather than daily net price change.
Crossover buys are taken only when that index is above 50, and crossover sells only when it is below 50, with delayed entries and exits at 60 and 40.
The same confirmation, delayed-entry, and exit rules were also applied to the relative strength index so a price-change filter could be compared with the volatility-direction filter.

Entries in this reading3 entries

Current trackRelative Volatility Index7 readings Moving-average crossover126 readings Relative Strength Index261 readings

Keep the timing clock separate from the confirming filter

The pairing is motivated by the observation that many confirming indicators reuse price-change information already present in the timing model, so a second reading should be based on a different measurement of market strength.

In this archive workflow, the timing signal is a moving-average crossover: a 10-bar simple moving average crossing a 20-bar simple moving average. The confirming filter is the relative volatility index. That second reading accepts, delays, or exits the crossover. It does not originate the trade by itself.

Directional volatility is not daily net price change

The relative volatility index applies the relative strength index calculation after substituting a 10-day standard deviation of closing prices for daily net price change. It is scaled from zero to 100, with readings above 50 treated as upside-directed volatility and readings below 50 as downside-directed volatility.

That midline summary is directional volatility: whether recent close-to-close variability has been concentrated on the upside or the downside.

Confirmation, delayed entry, and exit

Under the test rules, a moving-average crossover buy is taken only when the relative volatility index is above 50, and a crossover sell is taken only when it is below 50.

If a crossover buy is ignored because the relative volatility index is below 50, a long entry is still allowed if that index later rises to 60 or above. If a crossover sell is ignored because the index is above 50, a short entry is still allowed if it later falls to 40 or below.

A long position is exited if the relative volatility index falls to 40 or below, and a short position is exited if it rises to 60 or above.

The same gate on the relative strength index

The same confirmation, delayed-entry, and exit rules were also applied to the relative strength index so a price-change filter could be compared with the volatility-direction filter. In this pairing the relative strength index is a same-rules control filter, not a new timing model.

The historical sample covers six markets, each observed for 650 hourly bars.

Editorial interpretation: agreement between the two gates suggests the second reading may only be restating price-change information already present in the moving-average clock. Disagreement suggests the standard-deviation substitution has produced a different measurement of market strength. The archive records the shared rule set. It does not rank the two oscillators.

10/20 SMA crossover profit with RSI versus RVI confirmation

Across six hourly markets the unfiltered 10/20 moving-average crossover loses money in every contract. The same six-rule gate applied with RSI still leaves a net loss, while the RVI gate turns four of the six markets profitable and flips the combined result from a $12,762.50 loss to an $8,175 gain. Dollar figures are one-contract, no-commission totals read from Dorsey's Figure 1 results table.1 hour

Each market covers 650 hourly bars. Buys are taken only when the oscillator is above 50 and sells only when it is below 50, with delayed entries at 60/40 and exits at the opposite 40/60 band. No commission adjustment.

Educational research material, not investment advice. Historical source context does not establish present-day performance.

2 of 7 in the Relative Volatility Index track

19951-4 pp.Next on Relative Volatility IndexConstructing the relative volatility index and inertiaThe relative volatility index keeps the two-stream ratio of a relative-strength oscillator but substitutes a lookback standard deviation of recent closes for one-day net change.

All readings on this track · 7 readings

By fieldIndicators and filters874 readings By year1993106 readings

All 7 readings tagged Relative Volatility Index

Also on Relative Volatility Index5 readings

2005Genetic optimization of relative-volatility crossover rules 1995Constructing a high-low relative volatility trend filter 1990Constructing a two-line relative volatility index 1995Constructing the relative volatility index and inertia 2017Building a signed relative-volatility oscillator