Withdrawal Strategy

In Sweden, different account types are taxed differently on withdrawal. ISK withdrawals are tax-free (you already pay annual schablonskatt), depå withdrawals trigger 30% capital gains tax on the profit portion, and pension withdrawals are taxed as income.

The order you draw from these accounts affects your total lifetime tax bill by tens or even hundreds of thousands of kronor. Drawing pension income up to the grundavdrag (basic deduction) is nearly tax-free, while ISK above 300 000 kr costs schablonskatt whether you withdraw or not. The optimal sequence depends on your specific balances, gain fractions, and pension timing.

ISK First (naïve): Drain ISK completely before touching depå. This is what most people do by default. Simple, but often leaves tax savings on the table.

Community Order: Draw ISK above the 300 000 kr tax-free threshold first, then depå, then dip into the ISK free zone last. Preserves the schablon-free base as long as possible. This is the advice you find on forums and in Rika tillsammans.

Tax Smoothed: Compares the marginal cost of ISK (ongoing schablon drag above 300k) vs depå (capital gains tax on withdrawal) using a breakeven calculation. When depå has low gain fraction, it is cheaper to draw from depå first. When gain fraction is high, ISK is cheaper. This strategy adapts dynamically each year.

For a typical retiree with 3 Mkr in ISK and 1 Mkr in depå (50% gains), the difference between the best and worst strategy can be 100 000–300 000 kr in lifetime tax savings. The gap widens with larger balances and longer retirement horizons.

The biggest single lever is usually grundavdrag harvesting: drawing pension income up to the basic deduction each year at ~0% tax, rather than deferring it into larger, higher-taxed lump sums later. The second lever is depå gain fraction timing: selling depå assets while the unrealised gain is low, before they appreciate further.

ISK charges annual schablonskatt on the account balance regardless of whether you make withdrawals. Every year you delay drawing down your ISK, you pay that drag on assets you are not even spending. Depåkonto only triggers tax when you actually sell and realise a gain.

As a starting rule: if both accounts hold similar assets, draw ISK first to stop the ongoing schablon cost. But the answer depends on your depå gain fraction. If most of your depå value is unrealised profit, the 30% capital gains tax on withdrawal may cost more than the ISK schablon you would pay by waiting. The tax-smoothed strategy in this tool calculates the breakeven point for your specific situation each year.

A common approach for FIRE retirees is to draw ISK down to the 300 000 kr tax-free threshold, then shift to depå, then return to ISK once pension income begins. See the FIRE calculator to model withdrawal phases alongside your pension bridge.

ISK schablonskatt only applies to account value above 300 000 kr. Below that threshold, the account is entirely free of the annual tax drag. This creates a useful zone: if you can draw your ISK down to near 300 000 kr and keep it there, the remaining balance grows completely tax-free.

For withdrawal planning, this means the last 300 000 kr in your ISK is your most efficient money. Many advisors suggest preserving this floor as long as possible, drawing from depå or pension first, then spending from the ISK free zone only when necessary.

Timing the depletion of your ISK around when pension income starts can also help. If you can empty your ISK during a low-income year before full pension kicks in, you avoid paying schablon on the balance in future higher-income years. The ISK vs pension comparison explains the schablon mechanics in more detail.

The 4% rule comes from the Trinity Study (US, 1998): withdraw 4% of your starting portfolio in year one, then adjust for inflation each year. The study found this survived 95%+ of 30-year periods in US market history.

In Sweden, the rule needs adjustment. ISK drag of roughly 1.07% per year reduces your effective return, which means a 4% withdrawal rate consumes more real portfolio growth than the US figure implies. Swedish FIRE planners typically use 3% to 3.5% as a more conservative baseline when most savings sit in ISK.

The bigger adjustment for Swedes is the pension bridge. If allmän pension and tjänstepension together cover a significant portion of expenses once active, your private portfolio only needs to bridge 10–20 years rather than 30–40. That allows a higher withdrawal rate during the bridge period. The FIRE calculator models this bridge effect directly.

Sequence matters. The highest-tax mistake is realising large depå gains in the same year you receive full pension income. At that point your income is already high, and capital gains stack on top, potentially pushing you over the state tax threshold at 643 000 kr annual income where an extra 20% state tax applies.

A lower-tax approach: realise depå gains in the years before pension income starts in full. In those years your total income is lower, grundavdrag reduces your taxable base, and capital gains at 30% may be the only tax you pay. Spreading large gains across several years keeps each year below the state tax cliff.

The ISK schablon of 1.07% is fixed regardless of your income level, so there is no bracket effect. In a high-income year ISK is proportionally cheaper than depå realisation. In a low-income year depå realisation may be cheaper. This is exactly what the tax-smoothed strategy calculates. For the full picture of how your marginal tax changes with income, see the salary calculator.

The success rate answers: if you had retired with this plan at the start of each year from 1970 onward, how many of those cohorts would still have money left at your target end age? A 100% rate means every historical starting year survived. This is exactly the method behind the original 4% rule (the Trinity Study, 1998).

56 years of Swedish data is both useful and limited. It contains the 1970s stagflation, the 1990s banking crisis, the dot-com crash, the 2008 financial crisis, and the 2022 rate shock, so it covers real stress. But it is one country's history, and 56 data points is fewer than the US datasets most FIRE research uses (typically 1926–present). A 90% success rate here is informative, not a guarantee.