Three strategies for the withdrawal phase compared
Turning a portfolio you have built into a stream of income is a different discipline from building it in the first place. During the accumulation phase, time works for you; every weak market is a buying window. In the withdrawal phase the sign reverses. Now you are selling shares in order to live, and a weak market at the wrong moment can do lasting harm. The withdrawal strategy is the rule that fixes how much you take out each year and how that amount is allowed to change over the decades.
There is no single correct strategy. There is a trade-off between three quantities that cannot all be maximised at once: the stability of your income, the preservation of your purchasing power, and the likelihood that the capital outlives your time horizon. Each of the three classic strategies weights these quantities differently. This article describes them, places the tax consequences of a German withdrawal in context, and names the risk that binds all three together: the sequence of returns.
During the accumulation phase a market crash is an opportunity. In the withdrawal phase it is a bill the portfolio settles at once.
The three basic forms differ in one respect alone: how the annual withdrawal amount is determined, as a fixed euro figure, as an amount that grows with inflation, or as a fixed percentage of the current portfolio value. From this single setting all further properties follow, from inflation protection to the swing in income.
| Strategy | Withdrawal amount | Advantage | Risk |
|---|---|---|---|
| Constant | Fixed EUR amount per year | Predictable, secure income | Loss of purchasing power to inflation |
| Inflation-adjusted | Rises each year with the inflation rate | Purchasing power preserved | Portfolio more heavily strained |
| Percentage | Fixed % of the portfolio value | Portfolio always survives | Income fluctuates |
A fourth, rule-based variant (dynamic guardrails after Guyton-Klinger) combines elements of the inflation-adjusted and the percentage strategy. We treat it at the end, because it is best understood as a refinement of the classic forms. Do not read the table as a ranking. No row is superior; each swaps one worry for another.
With the constant nominal withdrawal you fix a set euro amount and take it out year after year unchanged, say EUR 30,000 per year, whether the portfolio rises or falls. It is the simplest rule imaginable, and its strength is predictability: you know today what amount will land in the account twenty years from now.
The quiet weakness lies in precisely that constancy. A nominally fixed amount loses real value, because prices keep rising while your withdrawal stands still. At an average inflation of around 2 percent per year, the purchasing power of a fixed amount halves over a span of roughly 35 years. The initial EUR 30,000 buys noticeably less at the end of a long retirement than at the start. For a short time horizon, or as a partial building block alongside an inflation-protected source (the gesetzliche Rente, Germany's statutory state pension, for example), the constant withdrawal is nonetheless a calm, foreseeable choice.
A constant nominal withdrawal is not to be confused with the constant real withdrawal of the 4 percent rule. The 4 percent rule keeps the amount constant in real terms, adjusting it each year for inflation. That is the second strategy.
The inflation-adjusted withdrawal is the strategy that stands behind the well-known 4 percent rule. In the first year you withdraw a set percentage of your starting capital, classically 4 percent, and thereafter raise that euro amount each year by the inflation rate. Your income thus stays constant in real terms: what you can buy stays the same over the years, and the euro amount grows only to offset the rise in prices.
The 4 percent rule has a concrete origin that also bounds its reach. William Bengen showed in 1994, in the Journal of Financial Planning, that an initial withdrawal of 4 percent, adjusted for inflation thereafter, would have survived every 30-year period in US history from 1926 onward. The Trinity Study (Cooley, Hubbard and Walz, Trinity University, 1998) confirmed this for US data from 1926 to 1995: an equity-heavy portfolio at 4 percent over 30 years had a historical success rate of around 95 percent, holding up almost always.
The 4 percent rule is a historical US backward look, not a law of nature and not a worldwide guarantee. It models neither taxes nor fund costs and assumes a time horizon of 30 years. Bengen himself held that for 50 years and more, around 3 percent was closer to “absolutely safe.”
From the percentage follows the familiar multiplier: 25 times annual spending equals 1 divided by 0.04. Lower the rate to 3.5 percent and the factor becomes around 29; at 3.0 percent, around 33. For long early-retirement horizons of 40 to 50 years, analyses by Kitces, Pfau and others suggest that the historically safe rate falls closer to 3.0 to 3.5 percent. We return to this in the section on choosing a strategy, because for a German early retirement three factors press independently on the rate.
The strength of this strategy is the standard of living it preserves. Its price is rigidity: because you keep withdrawing on an inflation-adjusted basis even in weak years, it strains the portfolio most heavily in exactly the phases where it is most vulnerable. That is the bridge to sequence-of-returns risk.
With the percentage withdrawal you take out a fixed percentage of the current portfolio value each year, say 4 percent of the level at the start of the year. If the portfolio stands at EUR 800,000, you withdraw EUR 32,000; if it has fallen to EUR 600,000, it is EUR 24,000.
This strategy solves the risk that plagues the first two: because you always withdraw a share and never a fixed amount, the portfolio can never, arithmetically, be fully exhausted. It shrinks in bad years, but it does not collapse. The price for this is a fluctuating income. In a year with a market drop of 25 percent your withdrawal also falls by a quarter, and that lands precisely when you are already nervous. Anyone who has to live on this withdrawal needs either a flexible lifestyle or a buffer that bridges the lean years.
The tax on a withdrawal is often overestimated in Germany, because an obvious rule of thumb does the arithmetic wrong. It is not the amount withdrawn that is taxed, but solely the capital gain contained in the shares that are sold (§ 20 Abs. 4 EStG, the German Income Tax Act). The invested capital flowing back to you stays tax-free. Someone who withdraws EUR 40,000, half of which is originally contributed capital, does not pay tax on EUR 40,000 but only on the gain portion.
Kernaussagen
Four building blocks determine the actual burden, and their order matters:
Order of the tax building blocks
Raw gain portion of the shares sold, then × 0.70 (Teilfreistellung), then − Sparerpauschbetrag, then × 26.375 % Abgeltungsteuer.
A deterministic example makes the difference visible. It is illustrative arithmetic, not a forecast and not a measured quantity. Suppose a single investor holds EUR 1,000,000 in an equity ETF, withdraws 4 percent (EUR 40,000), and the shares sold consist half of capital gain (an assumption, not a natural value):
| Step | Amount |
|---|---|
| Withdrawal (4 % of EUR 1,000,000) | EUR 40,000 |
| Contained gain (assumption 50 %) | EUR 20,000 |
| After Teilfreistellung (× 0.70) | EUR 14,000 |
| After Sparerpauschbetrag (− EUR 1,000) | EUR 13,000 taxable |
| Abgeltungsteuer (× 26.375 %) | EUR 3,428.75 |
The tax of EUR 3,428.75 amounts to 8.57 percent of the withdrawal, not 26.375 percent. The gross withdrawal rate of 4 percent thus becomes a net rate of around 3.66 percent, a tax deduction of about 0.34 percentage points. The naive calculation “26.375 percent of EUR 40,000 is EUR 10,550” overstates the burden in this case by roughly threefold, because it taxes the returning capital as well and ignores both the Teilfreistellung and the allowance.
The overstatement is not constant. In the early years the gain portion of the oldest shares is small, so the tax deduction is slight. With FIFO it rises over the decades, because the oldest lots sold first carry the highest embedded gain. The 3.66 percent named here applies to the 50 percent assumption chosen here, not as a fixed quantity.
The tables above show the principle for a single year. How a strategy behaves over an entire retirement depends on your capital, your desired withdrawal, the time horizon and the assumed return. The calculator below assembles these quantities into a multi-year withdrawal plan. The underlying return is an assumption, not a return forecast.
Treat the result as a planning anchor, not a promise. A small change in the return assumption shifts the final value considerably over three or four decades, and no single path captures what the markets will actually do.
Before you choose a strategy, it is worth looking at the risk common to all three: sequence-of-returns risk. In the withdrawal phase, weak returns in the first years do lasting, disproportionate harm. Anyone who sells shares early in a downturn removes them from the portfolio for good, so they no longer take part in the later recovery. Two retirees with an identical average return can end up very differently, depending on whether the bad years came at the start or at the end. In the accumulation phase this order plays no role; in the withdrawal phase it is central. The risk can be cushioned by a liquidity buffer of roughly two to three years of spending, by flexible spending, and by a cautious initial rate.
This is exactly where the dynamic guardrails after Guyton-Klinger come in, the rule-based fourth variant. You start with a somewhat higher initial withdrawal and define guardrails of around plus/minus 20 percent around the original withdrawal rate. If the running rate climbs above the upper guardrail after a weak year, you cut the withdrawal by about 10 percent; after a loss year you skip the inflation adjustment. In this way the plan responds to sequence risk instead of ignoring it. The price is an income that fluctuates within limits, the gain is a markedly higher likelihood that the capital carries you a long way.
For a German early retirement the sensible anchor shifts downward relative to the American 4 percent rule, toward 3.0 to 3.5 percent (around 29 to 33 times annual spending, roughly rounded to about 30 times). Three independent factors stack up:
A proven lever against this cost item is so-called Barista-FIRE: anyone who takes up genuine part-time work subject to social insurance (above a Minijob, Germany's marginal mini-employment) becomes compulsorily insured under § 226 SGB V and then pays statutory health insurance contributions only on the wage. Capital and rental income carry no additional contribution in that case. The full charging of the entire income hits voluntarily insured members alone.
From this no universally valid answer can be derived, but a sense of direction can. Anyone who places a stable, predictable income above all else and has a shorter horizon or an inflation-protected second source is well served by the constant or the inflation-adjusted withdrawal. Anyone who does not want to exhaust the portfolio at any price and can live with swings in income tends toward the percentage withdrawal. Anyone who wants to balance both finds in the Guyton-Klinger guardrails the most considered compromise. In every case the same holds: a fixed withdrawal rate, whatever it may be, can fail through an unfavourable sequence of returns. The lower initial rate and the willingness to cut flexibly in weak years are the most effective protection.
The example calculations and model assumptions in this article are illustrative and not a forecast. Return assumptions are not promises, and the historical success rate of the 4 percent rule is a US backward look, not a guarantee. Tax figures reflect the 2026 position and do not replace individual tax or investment advice.
Run the numbers on your withdrawal plan
In Investboard you can see how different withdrawal strategies affect your capital over the decades, with taxes and inflation taken into account.
View your withdrawal plan in Investboard →