Relative Volatility Index was elaborated by Donald Dorsey in 1993 and introduced in the "Technical Analysis of Stocks and Commodities" magazine.
Relative Volatility Index is not a separate indicator; it is used to confirm signals and improve trading systems.
This index measures forces of volatility movement. It does not duplicate signals of other oscillators, but just confirms them. RVI and Relative Strength Index are very much alike, but they additionally demonstrate highest and lowest points of price of standard deviation in a certain range.
RVI is calculated in a way similar to calculation of RSI, but here 10-days' standard deviation of close price is taken instead of change of price.
RVIorig = 100 * (EMA[W14] of U) / ( EMA[W14] of S)
RVI = (RVIorig of highs + RVIorig of lows) / 2
EMA (w14) – is an exponential moving average for 14-days' period;
U = S, on condition that current price is above the price of previous period;
U = 0, on condition that current price is below the price of previous period;
S = Stddev (10 days) — 10-days' standard deviation;
RVIorig of low – Relative Volatility Index for minimums;
RVIorig of highs – Relative Volatility Index for maximums;
If the source does not provide for a low or high volatility of figures, 10-days' Stddev figure is changed for 14-days' period.
Buy – when index is above 50;
Sell - when index is below 50;
If the first buy signal is missed, it will be necessary to make a purchase when index reaches value above 60;
If the first signal for sale is missed, it will be necessary to make a purchase when index reaches value below 40;
When the index of relative volatility gets below 40 or grows higher than 60, in such a case it is better to close position.
Comparing Relative Volatility Index with other indicators, one can say that RVI focuses on measuring other figures of market dynamics. A strong point of RVI is that it accounts for diversification levels omitted by RSI, because RVI was devised as a confirming indicator. More often, RVI is applied in combination with a moving average.
You might be also interested in: