A strategy for an affluent passive income lifestyle, with a $20M head-start
How a hands-on strategy with guard rails performs across bull markets, crashes, stagflation, and everything in between.
April 2026
Here's the question: you have $20 million. You want to withdraw roughly $500,000 a year after taxes, adjusted for inflation. You want to preserve the portfolio's real purchasing power over 30 years. And you'd rather not panic-sell at the bottom of a crash.
This article presents a simulation of one approach to that problem — a three-bucket strategy with explicit rules for when to buy, when to trim, and when to get defensive. The simulation below is fully interactive: you can run it against actual market history (1996–2025), a base-case forward projection (2026–2055), or an adverse scenario that opens with a severe crash. You can also toggle between California and Washington tax residency to see how state taxes change the picture.
The model is simplified. It uses annual time steps, assumes a fixed gain ratio for share sales, and doesn't model individual lot-level tax accounting. But it captures the core dynamics: income generation, tax drag, rebalancing discipline, and the compounding cost of selling into a down market.
The Three Buckets
The portfolio is split into three components at inception, each with a distinct job:
Bucket 1 — Income (40%)
SCHD + NOBL. High-quality dividend stocks. Generates the bulk of annual cash flow through qualified dividends, which are taxed favorably.
Bucket 2 — Growth (30%)
VOO + VTI. Broad U.S. equity exposure. This is the engine for long-term appreciation. Volatile year-to-year, but the rebalancing rules exploit that volatility.
Bucket 3 — Reserves (30%)
VTES + SGOV. A mix of 50% California munis, 33% Treasuries, and 17% TIPS. Low volatility. Provides bond income and acts as dry powder for buying crashes.
How Income Works
Each year, the portfolio generates income from three sources: dividends from Buckets 1 and 2, bond interest from Bucket 3, and — only if needed — share sales. The simulation always tries to meet the income target from natural yield first. If dividends and bond income fall short after taxes, it sells shares from Bucket 2 first (growth), then Bucket 3 (reserves), then Bucket 1 (income) as a last resort. Sales assume a 40% embedded gain ratio, so every dollar sold triggers long-term capital gains tax on $0.40 of that dollar.
The income target starts at $500,000 per year and ratchets up with inflation. There's an absolute floor of $200,000 — in the worst-case scenario, you'll never drop below that in nominal terms.
The Guard Rails
Above $24M real value: The portfolio has grown enough to coast. The strategy signals a switch to passive — no more active rebalancing needed.
The Rebalancing Rules
This is where the strategy earns its keep. Instead of rebalancing on a calendar, it rebalances on signal — based on what the market actually did last year:
- Crash >30% (B2 down hard): Move up to 8% of total portfolio from reserves into 70% growth + 30% income. This is the big opportunistic buy.
- Correction >15%: Move up to 5% from reserves into growth. A smaller but still meaningful dip-buy.
- First recovery after a crash: If last year was the first positive year after a >15% drop, deploy an additional 3% from reserves into growth. Follow-through conviction.
- Boom >20%: Trim 5% of growth into reserves. Lock in some gains while the market is euphoric.
- Good year >10% with low reserves: Smaller trim of 3% to refill the reserves bucket if it has drifted below 28%.
Any surplus income (post-tax income exceeding the withdrawal target) gets routed back into the portfolio. Where it goes depends on the current state: if the portfolio is in drawdown, surplus goes entirely to reserves. If reserves are low, it's split 65/35 reserves/income. If reserves are flush, it's 75/25 growth/income. Otherwise, it's distributed across all three buckets.
Tax Modeling
Taxes are modeled for two jurisdictions. California applies a 37.1% combined rate on qualified dividends and long-term capital gains, 54.1% on ordinary income, 0% on California municipal bond income, and 40.8% on Treasury interest. Washington (no state income tax, but with the capital gains tax) comes in at 23.8% on qualified dividends, 30.8% on long-term capital gains, and 40.8% on ordinary income. The difference is significant — run the simulation in both modes to see it compound.
The Historical Data
The backtest uses actual annual returns from 1996 through 2025 for U.S. dividend stocks (B1 proxy), broad market (B2 proxy), and bonds (Treasury, muni, and TIPS indices). It includes the dot-com crash, 9/11, the 2008 financial crisis, the COVID shock, the 2022 rate shock, and every bull and bear in between. Bond yields in the backtest are derived from the 10-year Treasury rate for each year.
The forward projections are scenario-based. The base case models a typical cycle: moderate growth punctuated by corrections roughly every 5–7 years and two major bear markets over 30 years. The adverse case opens with a severe crash (−38% in year one), followed by a grinding bottom and stagflation, before eventually recovering. It's designed to stress-test sequencing risk — the danger of bad returns early in the withdrawal period.
The Simulation
Toggle between scenarios and tax states. Hover over charts for year-by-year detail. The Actions tab shows every rebalancing decision the strategy made and why.