Skip to contents

Changelog

Source: NEWS.md

R4GoodPersonalFinances (development version)

R4GoodPersonalFinances 1.1.0

This version improve the performance of portfolio optimization algorithms, ability to parallelize the Monte Carlo simulations, and simulation caching. It brings also new useful functions as render_scenario_snapshot(), plot_future_saving_rates(), and get_default_gompertz_parameters(). With render_scenario_snapshot() function, it is possible to get a table with cash flow and net worth statements for a given year in a simulated scenario. The result of new features in this version can be seen in the blog post: Can Juliet manage without Romeo? How much she should spend, save, and invest to achieve that?.

Minor changes

  • Changed default portfolio optimization algorithm to NLOPT_LD_SLSQP (Sequential Least-Squares Quadratic Programming). The optimization is now more robust and faster.
  • Added ability to pass opts list to simulate_scenario() function with optimization parameters nloptr optimization function and thus changing algorithm or its options. The list of possible options and their default values can be obtain with nloptr::nloptr.print.options().
  • Added ability to pass a single seed value or vector of seeds simulate_scenario() function to reproduce the same results for the same scenario, also for Monte Carlo samples.
  • Added auto_parallel argument to simulate_scenario() function to automatically parallelize the Monte Carlo simulations.
  • Changed behavior of use_cache argument in simulate_scenarios() function to caching also all data of all Monte Carlo samples for a scenario (not only each sample separately).
  • Changed the reordering behavior in the plot_scenarios() which does not reorder scenarios anymore.
  • Changed optimal allocation for the last year of a simulation. It is now overwritten with the optimal allocation from the previous year, to avoid the edge case when net-worth is going to zero.
  • Added render_scenario_snapshot() function to render a table with cash flow and net worth statements for a given year (first by default) in a simulated scenario.
  • Added plot_future_saving_rates() for plotting the future saving rates.
  • Added print methods for Household and HouseholdMember objects.
  • Added ability to plot expected allocations based on the aggregated results from Monte Carlo samples with plot_expected_allocation().
  • Added age() function as convenient wrapper for getting age of a household member when defining triggers for expected income or spending streams.
  • Added get_default_gompertz_parameters() function to get default parameters for the Gompertz model for a given country, sex, and age.
  • Added, changed, or fixed multiple minor details in plots.
  • Updated default capital market assumptions.

R4GoodPersonalFinances 1.0.0

CRAN release: 2025-06-04

This major release introduces probably first in the world (open-sourced) implementation of a new multilevel life-cycle modeling of a household finances. It connects life-cycle models with single-period net-worth Markowitz optimization models and allows for high degree of personalization. The implementation is based on a novel theoretical framework, described by Thomas M. Idzorek and Paul D. Kaplan in “Lifetime Financial Advice: A Personalized Optimal Multilevel Approach” (2024).

Major changes

  • Added Household R6 class that aggregates information about a household and its members and HouseholdMember class that aggregates information about a single member of a household.
  • Added functions responsible for setting, resetting and getting information about disk cache for caching the results of the simulations: set_cache(), reset_cache() and get_cache().
  • Added calc_life_expectancy() function to calculate the life expectancy of a person based on the Gompertz model.
  • Added plot_life_expectancy() function to visualize the probability of dying and life expectancy of every member of a Household object and similar plot_survival() function for visualizing the probability of survival of each of the members of a Household and the probability of at least one member of a Household to survive.
  • Added create_portfolio_template() function to create an exemplary Portoflio object - a template for default portfolio with two asset classes: GlobalStocksIndexFund and InflationProtectedBonds and with some reasonable default values that should be updated and customised further.
  • Added calc_portfolio_parameters() function to calculate the parameters of a Portfolio object like current value, weights of assets, expected return and standard deviation of the whole portfolio.
  • Added calc_optimal_asset_allocation() function to calculate optimal allocation for a particular Household and Portfolio objects.
  • Added plot_optimal_portfolio() function to visualize the current and optimal allocation to assets for a particular Portfolio object.
  • Added simulate_scenario() function to simulate a single scenario of cash flows for a particular Household and Portfolio objects using expected returns of the assets in portfolio or random returns in multiple Monte Carlo samples.
  • Added simulate_scenarios() function to simulate multiple scenarios of cash flows for a particular Household and Portfolio objects. The scenarios parameters can triggered by different events like different years of retirement.
  • Added plot_scenarios() function to visualize the scenarios metrics that help with choosing the optimal parameters for the best scenario, e.g. the optimal year for retirement.
  • Added plot_expected_allocation() function to visualize the expected allocation of assets for a particular scenario over the household life cycle.
  • Added plot_expected_capital() function to visualize the expected financial, human, and total capital, as well as liabilities for a particular scenario over the household life cycle.
  • Added plot_future_income() and plot_future_spending() functions to visualize the future income and spending for a particular scenario over the household life cycle. The plot_future_spending() function allows to plot the relation of discretionary and non-discretionary spending, structure of non-discretionary spending and possible levels of discretionary spending over time based on Monte Carlo simulations.

R4GoodPersonalFinances 0.2.0

CRAN release: 2025-03-13

This release provides functions for calculating probability of retirement ruin and functions for calibrating the underlying Gompertz model with the use of the mortality rates.

Major changes

R4GoodPersonalFinances 0.1.1

Bugfixes

  • Fixed apps errors while executing locally.

R4GoodPersonalFinances 0.1.0

Initial release of the package includes functions designed to support the optimal allocation of your investment portfolio with the use of the ‘Merton Share’ formula.