EBTailIndex {ExtremeRisks} R Documentation

## Expectile Based Tail Index Estimation

### Description

Computes a point estimate of the tail index based on the Expectile Based (EB) estimator.

### Usage

EBTailIndex(data, tau, est=NULL)


### Arguments

 data A vector of (1 x n) observations. tau A real in (0,1) specifying the intermediate level τ_n. See Details\. est A real specifying the estimate of the expectile at the intermediate level tau.

### Details

For a dataset data of sample size n, the tail index γ of its (marginal) distribution is estimated using the EB estimator:

γ_n^E=(1+\frac{hat{bar{F}}_n(tilde{xi}_{tau_n})}{1-tau_n})^{-1},

where \hat{\bar{F}}_n is the empirical survival function of the observations, tilde{xi}_{tau_n} is an estimate of the τ_n-th expectile. The observations can be either independent or temporal dependent. See Padoan and Stupfler (2020) and Daouia et al. (2018) for details.

• The so-called intermediate level tau or tau_n is a sequence of positive reals such that τ_n -> 1 as n -> ∞. Practically, τ_n in (0,1) is the ratio between the empirical mean distance of the τ_n-th expectile from the smaller observations and the empirical mean distance of of the τ_n-th expectile from all the observations. An estimate of τ_n-th expectile is computed and used in turn to estimate γ.

• The value est, if provided, is meant to be an esitmate of the τ_n-th expectile which is used to estimate γ. On the contrary, if est=NULL, then the routine EBTailIndex estimate first the τ_n-th expectile expectile and then use it to estimate γ.

### Value

An estimate of the tain index γ.

### References

Padoan A.S. and Stupfler, G. (2020). Extreme expectile estimation for heavy-tailed time series. arXiv e-prints arXiv:2004.04078, https://arxiv.org/abs/2004.04078.

Daouia, A., Girard, S. and Stupfler, G. (2018). Estimation of tail risk based on extreme expectiles. Journal of the Royal Statistical Society: Series B, 80, 263-292.

### Examples

# Tail index estimation based on the Expectile based estimator obtained with data
# simulated from an AR(1) with 1-dimensional Student-t distributed innovations

tsDist <- "studentT"
tsType <- "AR"

# parameter setting
corr <- 0.8
df <- 3
par <- c(corr, df)

# Big- small-blocks setting
bigBlock <- 65
smallblock <- 15

# Intermediate level (or sample tail probability 1-tau)
tau <- 0.97

# sample size
ndata <- 2500

# Simulates a sample from an AR(1) model with Student-t innovations
data <- rtimeseries(ndata, tsDist, tsType, par)

# tail index estimation
gammaHat <- EBTailIndex(data, tau)
gammaHat


[Package ExtremeRisks version 0.0.4 Index]