hedgedRetRCov: Realized variance - covariance matrix of hedged option...

In piotrek-orlowski/impliedCF: Implied Characteristic Functions: power-divergence option portfolios. Magic.

Description

The function calculates the approximate matrix of realized variance-covariance between hedged GLT portfolios. The hedging frequency is determined by the frequency of stock futures data. If there is missing data in the time series of GLT values, these days are ommitted from hedging. The approximate nature of the result comes from the fact that we use the stochastic integral representation of the hedged GLT portfolio return and use Ito's isometry to build the approximation.

Usage

hedgedRetRCov(igft.list, igft.days, hf.dframe, weekly.grid,
  trade.end.grid = NULL, u.seed, fixed.maturities = c(1/12, 6/12),
  psd.ensure = TRUE, trade.int = 5/252)

Arguments

igft.list	A list of buy and sell values of the igft (the difference comes from the maturity decreasing over the trading interval + vol. changes) as output by constMatiGFT
igft.days	A vector containing the days corresponding to igft.list
hf.dframe	A data frame containing high frequency futures prices. It has 2+#TTM columns, day, time and F_ttm(1)...F_ttm(T), T = length(fixed.maturities).
weekly.grid	The approximately weekly grid of days where we empirically observe option prices (or their transformed values)
trade.end.grid	(optional) A grid of days on which trades starting at weekly.grid are supposed to end. Use if weekly.grid has some 'gaps'.
u.seed	Vector of base frequencies at which the portfolio values have been calculated. The vector's length is N
fixed.maturities	The maturities of the option portfolios we are mimicking
psd.ensure	If set to TRUE (default), the estimated variance covariance matrix is projected onto the space of positive semi-definite matrices as in [paper].
trade.int	The trading interval (annualized in business year) over which we hold the option position.

Value

Array of size (NxT)x(NxT)xS, where N is the number of base frequencies, T is the number of maturities, S is the length of weekly.grid.

Note

TODO: handle missing data (NA observations in the igft); add paper citation in @param psd.ensure

piotrek-orlowski/impliedCF documentation built on May 7, 2019, 8:18 a.m.