Does the 4% rule work for a large portfolio?

Will my wealth last? is one of the toughest questions around your wealth, and the 4% rule is the most quoted answer to it.

It says you can withdraw 4% of your portfolio in year one, raise that amount with inflation each year after, and be reasonably confident the money outlasts you.

It's useful research and a poor plan.

The short answer. No, not on its own. The rule assumes a 30 year retirement, one fixed portfolio, no taxes, no fees, and a retiree who never adjusts their spending. For a large portfolio the tax assumption alone breaks it, because 4% from an IRA and 4% from a Roth are completely different amounts of money.

What the original research assumed.

The rule comes from studies of historical market data, most influentially William Bengen's 1994 paper in the Journal of Financial Planning, and like all such work it had to fix a set of assumptions to produce one number.

• A 30 year retirement. Not 35, not 40.
• One fixed portfolio, broadly a simple split of large company stocks and government bonds.
• A fixed real withdrawal. The same inflation-adjusted amount every year, whatever markets did.
• One country's market history, over one span of decades.
• No taxes. None at all.
• No fees.
• A retiree who never changed their mind, never spent more in a good year, never cut back in a bad one.

Each is a reasonable simplification for research. Several are unrecognisable as a description of an actual household.

The assumption that matters most.

The fixed withdrawal is the crux.

The rule works by finding a rate that survives the worst historical stretch. Because it assumes you'll keep withdrawing the same real amount straight through a severe decline, it has to be cautious enough to survive that. The caution is the price of the rigidity.

No real retiree behaves that way. Anyone watching a portfolio fall by a third while drawing from it will consider spending less, and most will. That flexibility is worth a great deal, and the rule gives you no credit for it because the model doesn't contain it.

So the number is simultaneously too cautious for a household with real flexibility and too risky for one whose spending is almost entirely committed. It can't be both.

Why taxes break it for you specifically.

For a household with substantial assets, the missing tax assumption isn't a rounding error. It's the whole problem.

Withdrawing 4% means something completely different depending on where the money comes from. Four percent from a tax deferred account is ordinary income. The same amount from a taxable account is largely a return of your own basis, meaning the money you originally put in, plus some gain. Publication 551 is where the IRS sets out how basis is determined, and required withdrawals eventually override any percentage you had chosen anyway. From a Roth, a qualified withdrawal is tax free.

Three households with identical portfolios and identical 4% withdrawals can face wildly different tax bills, and therefore wildly different amounts actually available to spend. A rule expressed as a percentage of assets can't capture that, because the answer depends on which accounts you hold rather than on the total.

This is why we think in terms of which account funds which year rather than what percentage comes out. See which account you should draw from first.

What the research does get right.

It identified something real: the order in which returns arrive matters enormously once you're withdrawing.

Two portfolios with identical average returns produce very different outcomes if one has its poor years early. A decline in the first few years means selling shares to live on while they're cheap, and that damage doesn't reverse when the average recovers. This is the single largest risk in retirement income planning, and the rule deserves credit for making it visible. What it can't tell you is which market news should change anything, which is a separate discipline.

Where it goes wrong is the response. It manages that risk by permanently under-spending. There are better tools.

What to use instead.

Not a different percentage. A different structure.

Separate must-spend from choose-to-spend. What continues regardless, and what you'd trim and still recognize your life. This one distinction does more work than any withdrawal rate, because it tells you exactly how much flexibility you have. See how much do you actually spend.

Fund the must-spend from steady sources. If the necessities are covered by things that don't fluctuate much, a decline stops threatening your life and becomes a question about the discretionary layer.

Hold a cash reserve. A year of must-spend outside the market means no bad week ever forces a sale.

Decide the adjustment rule in advance. Rather than a fixed withdrawal, agree now what would make you spend less, by how much, and for how long. A decision made calmly beforehand is a far better decision than one made during a decline.

Plan the tax sequence, not the percentage. Which account, which year, and what it does to your bracket. That's the lever the rule can't see.

Run the survivor scenario on its own. A plan that works for two can strain badly for one, and that has nothing to do with withdrawal rates.

So is the number useless.

No. It's a useful first sanity check. If you're contemplating 9% a year, the research tells you something immediately. If you're drawing 2%, it tells you that you probably have more room than you're using, which is its own kind of failure to plan.

Treat it as a rough boundary on the conversation rather than the end of it. The real answer depends on your account mix, your tax picture, how much of your spending is genuinely fixed, and what you'd do if the first five years disappointed.

No two households produce the same answer, which is why a single percentage was never going to be one. If you'd like yours worked out properly, let's talk.

Investing involves risk, including the possible loss of principal. Past performance is not a guarantee of future results, and nothing here is a projection or a recommendation for your circumstances.