starting program to calculate the stock price on the nodes in the implied binomial tree, the tree of transition probabilities and the tree of Arrow-Debreu prices using Barle and Cakici's method.
starting program to calculate the stock price on the nodes in the implied tree, the transition probability tree and the Arrow-Debreu tree using Derman and Kani's method.
Auxiliary routine for rICfil: - if possible - generates for Scores Lambda~N(0,FI) (FI:: Fisher-Info) a Hampel-Krasker-IC psi to efficiency loss e, i.e. E psi Lambda' = unit(p) E psi=0 (1) E |psi|^2= (1+e) tr (FI^{-1}) (2) and psi= A Lambda w_b w_b=min(1,b/|A Lambda|) for di
Auxiliary routine for rICfil: - if possible - generates for Lambda=Lambda1+Lambda2, Lambda1~N(0,S1), Lambda2~N(0,S2) indep a Hampel-Krasker-IC psi to efficiency loss e, i.e. E psi Lambda' = EM, E psi=0 (1) E |psi|^2= (1+e) tr ((S1+S2)^{-1}) and psi= A (Lambda1 w_b + Lambda2) w_b=min(1,b/|
if does conditional branching, if the first element of the argument is true. If has to be followed by an endif. Else is an optional part of if sequences.
determines the implied volatilities assuming the Black Scholes model for a vector of European style options; uses either the method of bisections or the default Newton-Raphson method.
returns the optimal retention limit M, RBC capital, loading and expected rate of return from RBC, resulting either from minimization of the total premium or from maximization of the rate or return. The compound Poisson - truncated-Pareto model and FC1 or FC2 method of approximation of quantiles are
returns the optimal retention limit M, initial capital u and safety loading, resulting from minimization of premium c, under the assumption of some ruin probability level and dividend rate (fixed or flexible). The compound Poisson - truncated Pareto model and Beekman-Bowers or De Vylder method of a
interact estimates a model with interaction terms. It is using the marginal integration estimator with a local polynomial smoother. For details see Sperlich, Tjostheim, Yang (1997)
intertest1 is testing for interaction of x_1 and x_2 in an additive regression model. It is looking at the interation estimate and using wild bootstrap. For details see Sperlich, Tjostheim, Yang (1997)
intertest2 is testing for interaction of x_1 and x_2 in an additive regression model. It is looking at the estimate of the mixed derivative of the joint influence and using wild bootstrap. For details see Sperlich, Tjostheim, Yang (1997)
estimation of a bivariate joint influence function and its derivatives in a model with possible interaction. When loc.lin.smoother is chosen you get the function estimate and the first derivatives in the first and second direction, when loc.quadr.smoother is chosen you get the function and the mixe
Generation of menu system for impulse response analysis. A system of menus will appear on the screen that guides you through the stages of impulse response analysis for vector autoregressive models. This quantlet defines the following pop-up menus: irairmax, iracoverage, irabootci, irayscale, irai
isoreg computes the isotonic regression smoother via the Pool Adjacent Violators algorithm. Given a data set {(X_i,Y_i)} where X_i <= X_(i+1) i=1,...,n finds the values {mhat(X_i)} i=1,...,n, such that, minimizes (1/n) sum_i=1,...,n [Y_i - mhat(X_i)]^2 subject to mhat(X_i) <= mhat[X_(i+1)], i=
Auxiliary routine for rICfil: - if possible - it solves for Lambda~N(0,FI), where FI represents the Fisher-Info, following equations: A^{-1} =E [ Lambda Lambda' w_b ] (1) w_b=min(1,b/|A Lambda|) using a fixed-point-algorithm
Auxiliary routine for rICfil: It solves - if possible - for Lambda1~N(0,S1) and Lambda2~N(0,S2) following equations independently: A^{-1} =E [ Lambda1 Lambda1' w_b ] + E [ Lambda2 Lambda2' ] (1) w_b=min(1,b/|A Lambda1|) using a fixed-point-algorithm
main function for the Derman/Kani/Chriss method of implied trinomial trees (ITT). It computes the nodes of the ITT, the probability matrices, the Arrow-Debreu prices and the local volatility matrix.