| Library: | finance |
| See also: | archest annarchtest bigarch jump lochomest volatility |
| Quantlet: | lochomtest | |
| Description: | computes the most recent interval of homogeneity of the volatilities of financial time series data and the local mean relative to this interval. |
| Usage: | {xest, hom} = lochomtest(powtr, lambda, mu, m0) | |
| Input: | ||
| powtr | T x 1 vector, powtr = abs(S)^(0.5) where S is a time series of financial returns | |
| lambda | scalar, a smoothing parameter; 0.5 <= lambda <= 4 | |
| mu | scalar, a smoothing parameter, 0.5 <= mu <= 4 | |
| m0 | scalar, the grid step; m0 >= 5 | |
| Output: | ||
| xest | T x 1 vector, the estimated volatility | |
| hom | T x 1 vector, the estimated interval lengths of homogeneity at each grid point | |
library("finance")
x = read("dmus58")
z = x[2:rows(x)] - x[1:rows(x)-1]
powtr = sqrt(abs(z))
lambda = 2
mu = 3
m0 = 18
{xest, hom} = lochomtest(powtr, lambda, mu, m0)
xest
hom
Contents of xest [1,] 0.00055794 Contents of hom [1,] 523