Withdrawal Strategy
A withdrawal strategy is the set of rules for turning a portfolio into income: how much to take, which account to take it from, and what to change when markets fall. The rule for a bad year matters more than the starting rate, because a fixed rate sells the most units at the worst prices.
Accumulating is one problem and spending down is a different one. The second is harder, because a portfolio being drawn from cannot simply wait for a recovery — the withdrawals continue while it falls.
How it works
A withdrawal strategy has three parts, and most discussion only covers the first. How much to take. Which account to take it from. And what to change when the portfolio moves.
The simplest rule takes a fixed percentage of the starting balance, adjusted for inflation. It is easy to describe and it makes no reference at all to what the portfolio has done since.
Which means in a falling year it sells more units to raise the same cash, and those units are not there for the recovery. That is the mechanism behind every problem on this page.
Sequence is the risk
Two portfolios with identical average returns can end very differently, depending purely on the order the returns arrived in. Losses early, while the balance is largest and withdrawals are just starting, permanently reduce the base everything else compounds from.
On this site’s shared series, 95% of bars sit below a prior peak, the deepest drawdown was 3.76%
and the longest stretch under water ran 73 bars — and the series still finished up 3.61%. Drawing
income while below a high is the ordinary condition, not an emergency. The figures are in
research/series-measurements.json.
No fixed rate can see any of that. It is why the rule for bad years is the part that matters.
A worked example
Take 1,500,000 supporting 60,000 a year — a 4% rate.
The portfolio falls 25% to 1,125,000. Under a fixed rule, the withdrawal is still 60,000, which is now 5.33% of the balance.
Under a guardrail rule, spending is cut when the rate drifts too far above target — say back to 54,000, restoring the rate to 4.8% rather than 5.33%.
The 6,000 cut is what buys the portfolio time, and it is temporary. The whole point of the approach is that a small, reversible adjustment made early avoids a large, permanent one later.
Which account to draw from
A taxable account, a pre-tax account and a Roth-style account are taxed differently on withdrawal, so the same 60,000 of spending costs different amounts depending on where it comes from.
Drawing from the taxable account first is a common default, leaving sheltered balances to compound longer — but it is not universally right, because it can leave a large pre-tax balance forcing high required distributions later.
A common refinement is to fill lower tax bands from the pre-tax account each year rather than leaving it untouched. This is educational, not tax advice, and it depends heavily on jurisdiction.
The cash buffer
Holding one to three years of spending in cash means a bad year does not force a sale. The cost is that the cash earns less than the portfolio would have, every year, whether or not a fall arrives.
That is a real and permanent cost paid for an occasional benefit, which is a trade worth stating plainly rather than assuming. It is bought for the same reason any insurance is.
Costs
Raising cash means selling, and selling costs. On this site’s shared series a round trip measures about 2% of the median bar range of 0.493, and in a taxable account it is also a disposal.
Which argues for drawing annually or quarterly rather than monthly, and for holding the cash to spend from rather than selling every month.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 2 have a title about withdrawal
or drawdown strategy, at a median of 240,335 views — and 50% use beginner-shaped language. Safe
withdrawal appears in 1 at 283,990 and annuities in 2 at 430,119. The counts come from
site/rank_investing.py.
Two videos at a 240,335 median. Every retirement-income subject on this site shows the same shape — very high attention against almost no instructional supply — and this is the operational question underneath all of them.
The answer to the question on that chart is yes, and the size of the cut is the whole question. A temporary 10% reduction taken early is far cheaper than the permanent one a depleted portfolio forces later. The decision is easier if it was written down beforehand, because in the moment it feels like a defeat rather than the mechanism working.
When it fails
The failure is a fixed rule met by a bad first decade, and it fails invisibly for years. The withdrawals continue at the planned amount, the balance falls faster than any single year explains, and nothing signals a problem until the arithmetic no longer works at all. By then the correction required is large, permanent, and arrives at an age when returning to work is not a realistic option. The rule behaved exactly as specified the entire time.
The second failure is a plan with no flexibility in the date or the spending. Something has to be able to move.
A third is drawing monthly by selling monthly. It maximises transactions and taxable events.
A fourth is ignoring the withdrawal order across accounts. The same spending costs different amounts.
A fifth is confusing a nominal rate with a real one. Spending has to rise with prices.
And a sixth is treating a sustainable rate as a promise. It is arithmetic on an assumption about sequences nobody has seen yet.
Related
Annuities remove the failure mode entirely, at the cost of the capital. Asset allocation decides how violent the sequence risk is. And social security is the inflation-linked floor underneath all of it.
The plan I would want is not the one with the highest sustainable rate. It is the one with a written rule for what happens after a 30% fall, decided while nothing is falling. Everything else is arithmetic you can redo; that rule is the only part that has to survive being frightened.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.