Assumptions
The safe withdrawal rate follows the logic of the 4% rule from the Trinity study: you withdraw a set percentage of the starting portfolio in year 1, then raise that dollar amount with inflation every year after, regardless of what the market does. Each cell is the highest first-year rate that still lasts the whole horizon in your chosen share of 10,000 simulated markets. Because the amount is indexed to inflation, it is the same spending power in today's dollars every year. Stocks and bonds are simulated separately from the nominal compound returns, volatilities, correlation and inflation you set, and the portfolio is rebalanced every year. The same simulated markets feed every row, so rows change smoothly.
“Independent” draws each year without memory. The variance ratio over 10 years, VR(10), measures memory: below 1 means a bad decade tends to be followed by recovery, above 1 means it tends to continue. “Historical” uses the ratios from 150 years of data across 16 countries (stocks 0.91, bonds 2.66), modelled as a simple one-year memory. That is a rough fit: it can't reproduce history's joint stock and bond behaviour, so it lands below the block-bootstrap result in the methodology notes. Bond persistence is what pulls bond-heavy rows down.
Acknowledgement: The default return, volatility and correlation assumptions are adopted from PWL Capital.