optimalw: Optimiser to maximize Expected Utility

In sstoeckl/pensionfinanceLi: Optimal Pension Decisions in Liechtenstein

Description

Here we minimize negative expected utility (given by function 'util()'). Parameters can be fixed by setting "tight" boundaries

Usage

optimalw(
  initial_values,
  upper_bounds = NULL,
  lower_bounds = NULL,
  ret_age,
  cc,
  c2,
  nu2,
  nu3,
  ra,
  delta,
  alpha,
  beta,
  c_age,
  gender,
  gender_mortalityTable,
  w0,
  CF,
  li,
  lg,
  c1,
  s1,
  s2,
  s3,
  w2,
  rho2,
  rho3,
  ret,
  retr,
  psi,
  verbose = FALSE,
  warnings = FALSE,
  trace = 0,
  reltol = sqrt(.Machine$double.eps)
)

Arguments

initial_values	Starting values c(w3)
upper_bounds	Upper bounds for optimization
lower_bounds	Lower bounds for optimization
c2	Given Variable: second pillar savings as fraction of gross income (still missing: health, a-fonds-perdu payments)
nu2	Given Variable: fraction of second pillar savings that is converted to life-long pension
nu3	Given Variable: fraction of third pillar savings that is converted to life-long pension
alpha	Given Variable: parameter to choose fraction of wealth NOT consumed during retirement but kept for investment (and subsequent consumption)
c_age	Given variable: the investor's current age (assuming birthday is calculation-day)
gender	Given variable: gender, 0=male and 1=female
w0	Given variable: time c_age wealth that is not disposable, assumption: still available at retirement (no growth or decline), alternatively: expected wealth (that is not disposable) at retirement, stays the same over time
CF	Given Variables: income shocks, such as inheritance (not currently implemented)
li	Given variable: gross labor income at time 0 (in the last year before birthday)
lg	Given variable: labor growth rate (in real terms, constant)
c1	Given variable: first pillar savings as fraction of gross income
s1	Given variable: vector consisting of two components: c(number of contribution years at age=c_age,historical average yearly income until c_age)
s2	Given variable: savings in second pillar as of t=0
s3	Given variable: liquid wealth - invested in the third pillar (current assumption: no tax advantage for third pillar)
w2	Given variable: portfolio allocation in second pillar (assumed to be fixed and not influenced by the decision maker)
rho2	Given variable: conversion factor in second pillar for regular retirement age
rho3	Given variable: conversion factor in third pillar for regular retirement age
ret	Given variable: investment return scenarios (nominal)
retr	Given variable: investment return scenarios (real)
psi	Given variable: optional, spread to take a loan/leverage for third pillar savings
warnings	optional: should warnings be given? (default=TRUE)
c	Given Variable: fraction of income that is consumed while still working (current assumption: constant)

Value

Expected utility

Examples

## Not run: 
data(ret);data(retr)
.load_parameters(gend=0,type=1)
initial_values <- c(0.25, 0.25, 0.25)
outw <- optimalw(initial_values,upper_bounds,lower_bounds,
ret_age=ret_age,cc=cc,c2=c2,nu2=nu2,nu3=nu3,
         ra=ra,delta=delta,alpha=aalpha,beta=bbeta,c_age=c_age,
         gender=gender,gender_mortalityTable=gender_mortalityTable,
         w0=w0,CF=CF,li=li,lg=lg,c1=c1,s1=s1,s2=s2,s3=s3,w2=w2,
         rho2=rho2,rho3=rho3,ret=ret,retr=retr,psi=psi,trace=1,reltol=1e-4)
outw %>% tibble::rownames_to_column() %>% as_tibble %>% arrange (value) %>%
filter(convcode==0,kkt1==TRUE,kkt2==TRUE)
util(ret_age=ret_age,c=cc,c2=c2,nu2=nu2,nu3=nu3,tw3=outw2$xmin+c(0,-0.01,-0.01),
         ra=ra,delta=delta,alpha=aalpha,beta=bbeta,c_age=c_age,
         gender=gender,gender_mortalityTable=gender_mortalityTable,
         w0=w0,CF=CF,li=li,lg=lg,c1=c1,s1=s1,s2=s2,s3=s3,w2=w2,
         rho2=rho2,rho3=rho3,ret=ret,retr=retr,psi=psi)

## End(Not run)

sstoeckl/pensionfinanceLi documentation built on Dec. 2, 2020, 3:26 a.m.