How much do I need to retire?
The 4% rule and its limits.

What the 4% rule is

The rule is a planning heuristic for how much you can safely withdraw from a retirement portfolio without running out of money.

The mechanics: take 4% of the portfolio's value on the day you retire. That is your year-one spending number. Every subsequent year, raise that dollar amount by the previous year's inflation. You are NOT withdrawing 4% each year. You are withdrawing a fixed (inflation-adjusted) dollar amount based on the starting balance.

On a $1,000,000 portfolio, year-one withdrawal is $40,000. If inflation runs 3% the next year, year-two withdrawal is $41,200. And so on.

The Trinity Study origins

William Bengen (1994) published the original research. The Trinity Study (Cooley, Hubbard, Walz 1998) extended it with a fuller set of historical simulations. Both used rolling 30-year windows of US stock and bond returns going back to 1926.

What they found, for a 50/50 stock/bond portfolio withdrawing 4% with annual inflation adjustments over 30 years:

• Roughly 95% of historical 30-year windows ended with money still in the account.
• Roughly 5% of windows ran out. The failures clustered around retirements starting in 1929, 1937, 1966, and 1973.
• Higher stock allocations (75/25) had higher terminal wealth in most cases but similar survival rates at 4%.

What the 4% rule was NOT: a guarantee, a target, or advice for a 60-year retirement. It was a probabilistic finding about a specific historical dataset.

Why modern researchers argue 3.3% today

Wade Pfau, Michael Kitces, and Morningstar's annual "State of Retirement Income" report have all argued that current market conditions make 4% more aggressive than it sounds. William Bernstein (The Ages of the Investor, 2012) adds a different critique: the 4% rule assumes a level of equity risk tolerance most actual retirees do not have. His "If you've won the game, stop playing" framing argues for de-risking the allocation itself once you have enough, rather than trusting that the historical 4% number holds through the next drawdown.

Two reasons:

• High equity valuations. Starting P/E ratios have historically predicted long-run returns. When CAPE is elevated, forward-looking 30-year equity returns tend to be below the historical average.
• Bond yield environment. The Trinity study's bond sleeve benefited from a 40-year bond bull market (1981 to 2020). A retiree starting in today's yield environment cannot count on the same tailwind.

Morningstar's recent "State of Retirement Income" has put the safe starting withdrawal rate around 3.7 to 4.0% depending on assumptions; Pfau has argued 3.3% for a 30-year retirement beginning in a high-valuation regime.

The path-dependence problem

The 4% rule works on the average historical path. If you happen to retire into one of the bad 5% of paths, 4% fails. This is the sequence-of-returns problem. See sequence of returns. You cannot know in advance whether you are on a good or bad path. That is why the rule needs guardrails.

How Bitcoin changes the calculus

Adding a Bitcoin sleeve to a retirement portfolio has two opposing effects on the safe withdrawal rate.

• Short-term, the volatility hurts you. BTC's 50 to 80% drawdowns raise sequence-of-returns risk in the first retirement years. If you retire into a BTC bear market and have to sell to eat, the math is brutal.
• Long-term, the return tail helps. BTC's historical return, even after huge drawdowns, has been far above stocks. A small permanent allocation raises the median terminal wealth and typically raises the sustainable withdrawal rate in Monte Carlo simulations.

The net: a 5 to 10% BTC sleeve tends to increase the SWR in simulation IF the retiree has a non-BTC cash buffer large enough to avoid forced selling in the bad first years. Without the buffer, the volatility wins.

The SCHD alternative (dividend strategy)

A completely different approach: do not withdraw principal at all. Live off the dividends and let the portfolio compound.

SCHD (Schwab US Dividend Equity ETF) currently yields somewhere in the 3.5 to 4% range. On a $2,000,000 portfolio, that is $70,000 to $80,000 of annual income without ever selling a share. No sequence risk in the traditional sense, because you are not drawing on principal.

The tradeoffs:

• Lower total return than a total-market portfolio (VTI) over long horizons in most historical windows.
• Dividend tax drag if held in a taxable account. Works much better in Roth.
• Concentration risk: you are heavily tilted toward large-cap US dividend payers.
• Dividend cuts happen in deep recessions and cut your income at the worst time.

It is not strictly better or worse than a total-return 4% rule approach. For retirees who cannot psychologically tolerate selling principal, it is a legitimate alternative.

Guardrails: the Guyton-Klinger rules

Guyton and Klinger (2006) showed that a retiree willing to adjust spending dynamically can safely start at a higher rate than the static 4%. The rules, simplified:

• If the current withdrawal rate rises above an upper guardrail (e.g. 20% above the starting rate), cut spending 10%.
• If it falls below a lower guardrail (e.g. 20% below), increase spending 10%.
• Skip the inflation adjustment in any year the portfolio lost money.

With guardrails, the safe starting rate rises to roughly 5 to 5.5% on a 60/40 portfolio. The tradeoff is that you have to actually execute the cuts when markets are down, which is psychologically hard. Set the rule in writing now, automate it, do not renegotiate with yourself in a bear market.

Related