VariationOfInformation: Variation of Information

In sa-ccr/Trading: CCR, Advanced Correlation & Beta Estimates, Betting Strategies

Variation of Information

Description

Calculates the variation of information of the track records of two assets/strategies based on the sample entropy.

Usage

VariationOfInformation(x, y, m = 2, r = 0.2, normalized = TRUE)

Arguments

x	a vector containing the track record of the underlying asset/strategy, this will be normalized during the algorithm
y	a vector containing the track record of the underlying asset/strategy, this will be normalized during the algorithm
m	an integer value defining the embedding dimension , default value is 2
r	a double value defining the tolerance, default value is 0.2
normalized	a boolean value so as to bound the return value between 0 and 1, default value is TRUE

Value

A value containing the variation of information

Author(s)

Tasos Grivas <[e-mail]>

References

Lopez de Prado, Marcos, Codependence (Presentation Slides) (January 2, 2020). Available at SSRN: https://ssrn.com/abstract=3512994

Examples

x = PerformanceAnalytics::edhec[,c("Short Selling")]
y = PerformanceAnalytics::edhec[,c("Convertible Arbitrage")]
variation_of_information = VariationOfInformation(x, y, m=2, r=0.2, normalized = TRUE)

sa-ccr/Trading documentation built on Feb. 23, 2024, 9:26 p.m.