Back to Portfolio Modeler

How ML Stats Calculates Portfolio Projections

The Portfolio Income Modeler estimates sustainability month by month using configured portfolio weights, annual price growth assumptions, annual dividend yields, and inflation-adjusted income targets.

What the model returns

Standard withdrawal mode

  1. Convert annual growth assumptions to monthly compounding rates and apply to each holding.
  2. Convert annual dividend yield to monthly income and add to cash balance.
  3. Withdraw income target from cash first, then sell holdings if needed based on sale strategy. Equal Dollars Across Holdings sells equal dollar amounts from each funded holding, using remaining holdings if one is exhausted.
  4. Record any unmet amount as monthly shortfall.
  5. If DRIP is enabled, reinvest remaining cash by current holding weights.
  6. If rebalance cadence triggers, rebalance to target weights using total holdings plus cash, even when DRIP is off. With monthly rebalancing, DRIP on and off therefore have the same result.

Allocation totals within 0.01 percentage points of 100% are accepted and normalized to invest the entire balance. Retained dividend cash earns no return. It appears separately in Final Holdings and as the end-of-period uninvested cash balance in Details; it is included in portfolio totals.

Three Bucket withdrawal mode

Three Bucket mode treats the portfolio as near-term, intermediate, and long-term reserves instead of one fully rebalanced portfolio.

  1. Bucket 1 is sized from monthly income, 12 months per year, and the Bucket 1 Years input.
  2. Bucket 2 is sized from monthly income, 12 months per year, and the Bucket 2 Years input.
  3. Remaining portfolio value is assigned to Bucket 3. Within each bucket, value is distributed across that bucket's holdings by their allocation weights.
  4. A bucket with no positive-weight holdings receives no allocation. Any remaining value is assigned to the highest available bucket. Income and bucket years determine reserve sizes, so the generic 60/40, 70/30 and 80/20 presets produce identical starting values in Three Bucket mode. The 3-Bucket and 50/30/20 presets use the same assets and weights.
  5. The user-facing bucket model uses broad asset class sleeves: near-term cash, intermediate bonds, and long-term stocks.
  6. Each month, the model applies growth, adds dividends, funds spending from cash, then withdraws from Bucket 1, Bucket 2, and Bucket 3 as a last resort.
  7. Sales inside the active bucket use the selected sale order. Proportional Within Bucket sells the same fraction of each holding in the active bucket.
  8. On the selected cadence, Bucket 1 is replenished from Bucket 2 first. Bucket 2, and any remaining Bucket 1 need, can be replenished from Bucket 3 only when Bucket 3 had positive growth that month.
  9. Normal full-portfolio rebalancing is not applied in Three Bucket mode because it would undo the bucket structure.

DRIP reinvests surplus dividends after spending by current holding weights. Uninvested cash is included in Bucket 1 ending value and the portfolio total; the separate cash column is a component of those totals.

Max monthly income mode

When enabled, the model runs repeated simulations to search for the highest monthly income that keeps ending value near a target within a configured tolerance band. The search uses the selected withdrawal mode.

Monte Carlo analysis

Monte Carlo analysis reuses the same monthly engine and selected withdrawal mode while sampling inflation, portfolio growth, and dividend yield assumptions across repeated scenarios.

Important assumptions

Important note

Results are informational and educational. They are not guarantees of future investment performance and are not recommendations to buy, sell, hold, allocate, or rebalance.