\(EVI\) predicted all epidemic waves in Italy (Fig. 1) and New York (Fig. 2). Similar results were obtained for all countries (supplementary file S1, S2 and ). The overall \(Se\) and \(Sp\) may not be perfect but is, in all instances high and even close to unity. Further, \(Se\) and \(Sp\) refer to the ability of \(EVI\) to identify expected rises, as defined by Eq. 4, at individual time points. Perhaps the most important aspect in the use of \(EVI\) is its ability to predict future epidemic waves from the repetitive issuance of early warnings. A repetitive issuance of early warning as a predictor of new epidemic wave has been a consistent and remarkably stable finding across all countries. In a similar manner consecutive absence of early warnings indicates a flattening or drop in the future cases. The latter was also a consistent finding across all countries (supplementary file S1, S2 and S3). .
The Epidemic Volatility Index has been inspired by the use of volatility indices in the stock markets where a mainly negative association has been observed between volatility and stocks or stocks' future prices \cite{Fernandes_2014,Brenner_1989}. Most of  the stock price volatility indices are based on the standard deviation of historical price data sets and the moving standard deviation is the most common volatility index\cite{schwert2002stock}\(EVI\) is also based on the moving standard deviation and quantify the rate of change in this.