zhao-hailei/prime-harvesting

A collection of python modules and jupyter notebooks for retirement research

This is a series of retirement financial simulations and investigations, initially inspired by Michael McClung's book Living Off Your Money.

QUICKSTART

$ virtualenv . $ bin/pip install --requirement requirements.txt $ bin/jupyter-notebook

JUPYTER NOTEBOOKS

• Prime Harvesting. The one that started the project. A comparison of Prime Harvesting and Annual Rebalancing. Blog post.
• EM vs. VPW. A comparison of portfolio values and incomes for EM and VPW (from bogleheads).
• Inverted Withdrawals. A look at Inverted Withdrawals and blog post.
• Liability Matching Portfolio. Haven't done anything with this one yet...
• Sleep Well At Night. A first look at bond levels with Prime Harvesting. Blog post.
• Monthly Harvesting. The difference between checking annually and checking monthly. Blog post.

DATA SOURCES

CSV imports are easier for the code to deal with, so I've frequently munged original data source into CSV files.

• 1871_returns.csv comes from Simba's backtesting spreadsheet on bogleheads
• 2004-period-life-table.csv comes from https://www.kitces.com/joint-life-expectancy-and-mortality-calculator/
• PortfolioChart.com 'simulated indexes'. The blog post introducing them is https://portfoliocharts.com/stock-index-calculator/

I've also included local copies of some of the Excel sheets the CSV files were derived from.

• Simba's backtesting spreadsheet
• PortfolioChart.com 'simulated indexes' raw Excel file

FUTURE IDEAS

• Probability of failure and SIZE of failure might be different. You might choose the plan with the higher probability of failure because the size of failure is smaller. See: Estrada's new paper
• Try to tie together Prime Harvesting and valuations better.
• CEW (and by extension WER & HREFF) punishing declining withdrawal rates. However, the evidence from actual retirees (Bernicke & others) shows that's exactly what retirees do. Have a metric that follows this?
• Mortality-weighted shortfall calculations from Gardner's paper. Except...when you do Monte Carlo with stochastic mortality, you get this already.
• Look at how bond percentages change with monthly Prime Harvesting.
• Look at how valuation/expected returns in PMT does long term smoothing
• Try putting Pye's higher rate with Siegel & Waring's average lifespan Does it tilt TOO much towards early income?
• Map early income tilts from PMT shapes against Bernicke & Blanchett's research on actual retiree spending.
• Walton's paper on inverted withdrawals
• Treat SS as a bond, argument against: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=200572&newpost=3072525#p3072379
• "Alpha, Beta, and now Gamma" includes parameters for asset classes that allows use of a Truncated Levy Flight distribution to create Monte Carlo analysis. That sounds fun?
• redo Blanchett's Revisiting the Optimal Distribution Glide Path with a variable withdrawal strategy instead of constant withdrawals