Pillar 3a vs pillar 2 buy-ins: which to max first

This is the first piece of a short cluster on pillar 3a. The down-payment article introduced pillar 3a as the tax-advantaged vehicle that builds a property deposit and counts as hard equity. This piece widens the lens: for a household with discretionary capital after the fixed costs are covered, pillar 3a is not the only voluntary, tax-deductible option. The other is a voluntary buy-in into the occupational pension (pillar 2). They compete for the same franc, and they are not interchangeable.

Pillar 3a contributions are capped each year: CHF 7,258 if you are employed and in an occupational pension, or 20% of net earned income up to CHF 36,288 if you are self-employed without one (BVV3 art. 7). The contribution is deductible from taxable income at federal, cantonal, and communal level.

Pillar 2 voluntary buy-ins (Einkäufe)have no franc cap. You may buy in up to the gap between your current vested benefits and the maximum your plan's regulations allow at your age and salary, a figure shown on your annual pension certificate (Vorsorgeausweis). The buy-in is deductible at the same three levels, with no annual ceiling.

So the decision this piece answers: for a household with capital left over after mortgage saving, living costs, and an emergency fund, which pillar gets it first? The answer turns on five dimensions, and the most important of them is not the headline tax deduction.

On the face of it, the buy-in looks more powerful. Pillar 3a is capped and strictly annual: the allowance is use-it-or-lose-it, and an unfunded year cannot be recovered later in the established regime. (A separate 2025 mechanism allows retroactive 3a catch-up contributions under conditions; it is out of scope here and gets its own piece.) A buy-in, by contrast, carries no franc limit, so in a high-income year it can shelter far more.

The asymmetry is real. A salaried employee earning CHF 200,000 who makes a CHF 50,000 buy-in deducts the full CHF 50,000 from that year's taxable income, roughly seven times the pillar 3a cap. Both vehicles deduct at federal, cantonal, and communal level, so the per-franc tax saving is the same; the buy-in simply lets you deduct more francs at once.

But the deduction only sticks if the capital can stay where it is. The buy-in's advantage is conditional on a lock-up rule that pillar 3a does not share, and that rule is where most of the real decision lives.

A pillar 2 buy-in makes sense only if no 2nd-pillar capital will be withdrawn for at least three years.

This is the rule that reorders the whole comparison.

The concrete cases are unforgiving. A buy-in within three years of a property purchase financed by a WEF withdrawal is recaptured. A buy-in within three years of taking a retirement lump sum is recaptured. A buy-in within three years of a permanent departure from Switzerland is recaptured. In each case the tax saving you booked is reversed, and the planning that justified the buy-in unwinds.

Pillar 3a has no equivalent freeze. A 3a withdrawal for a property purchase does not unwind any prior 3a deduction; the deduction was clean in the year it was taken and stays clean.

The practical conclusion is sharp: a pillar 2 buy-in makes sense only if no 2nd-pillar capital will be withdrawn for at least three years. For a household with a property purchase on the near horizon, that condition usually fails, and the buy-in is the wrong tool regardless of how attractive its deduction looks in isolation.

The second structural difference is what the money earns while it sits.

The mandatory portion of pillar 2 must be credited at the BVG minimum interest rate, 1.25% for 2026, held by the Federal Council in August 2025. The recent path is informative: 1.0% through 2023, raised to 1.25% from 2024, and held for 2025 and 2026 (Federal Council / BSV). The supplementary portion(überobligatorisch) is credited at whatever the plan's regulations set, often a lower rate than the mandatory minimum under the common split-interest practice.

A pillar 3a account can be held the same low-yielding way, as a bank 3a, in which case the return roughly matches the pillar 2 mandatory rate and this section does not apply. The relevant comparison is an invested pillar 3a (VIAC, Finpension, Frankly), held in an equity-heavy strategy. Assume, purely as an assumption and not a forecast, a long-run nominal return of 5% for such a portfolio, with the volatility that implies. The contrast is then guaranteed-low against expected-higher-with-volatility.

Over a long horizon the gap compounds into real money. Contributing CHF 7,258 a year for 20 years:

• At the pillar 2 mandatory 1.25%, the balance grows to about CHF 164,000.
• At the assumed invested-3a 5%, to about CHF 240,000.

That is roughly CHF 76,000 more from the same nominal contributions, before any withdrawal-tax considerations, purely from the return differential. The 5% is a stated assumption; a lower assumption narrows the gap and a higher one widens it, but at any plausible equity return the invested 3a outpaces the guaranteed 1.25%.

Three smaller dimensions round out the comparison.

Insurance.Pillar 2 is not only savings: it includes mandatory disability (Invalidenrente) and survivor (Hinterlassenenrente) cover. Under the BVG minimum, the surviving spouse's pension is 60%of the insured retirement or disability pension (subject to marriage-duration and age conditions, or dependent children), each child's pension is 20% (to age 18, or 25 in education or training), and the disability pension scales with the degree of disability (BVG art. 23–26; Swiss Life). The combined first and second pillars target roughly 60% of prior salary in payment. Pillar 3a contains no insurance, only savings. For a young sole breadwinner, the pillar 2 cover is real value that 3a savings cannot replicate, though the franc amount is plan-specific and any single figure is illustrative rather than exact.

Inheritance.Pillar 2 follows a strict statutory cascade (BVG art. 20a): spouse or registered partner and orphans first, with cohabiting partners eligible only if the fund's own regulations provide for it, typically on proof of a shared household or joint children. Pillar 3a (BVV3 art. 2) follows its own ordered cascade but with an explicit, standardised path to name a cohabiting partner (BVG art. 20a; BVV3 art. 2). For unmarried couples, pillar 3a is materially more reliable as a way to direct the capital to a partner.

Exit on departure. On a permanent move to an EU or EFTA state, only the supplementary portion of pillar 2 can be cashed out; the mandatory portion must stay in a Swiss vested-benefits foundation until retirement age, because the person remains insured for the old-age, death, and disability risks in the destination state. On a move to a non-EU/EFTA country, the full pillar 2 can generally be withdrawn in cash. Pillar 3a can be withdrawn in full in either case (FZG, SR 831.42; Auffangeinrichtung BVG). Withdrawal or source tax applies, and the destination country may tax the lump sum on receipt; the cross-border tax detail is the subject of a forthcoming piece in this cluster.

Continue the Zug couple from the mortgage cluster. Household income CHF 200,000, a property purchase planned in five to seven years, and after fixed costs, mortgage saving, and an emergency fund, assume CHF 25,000 a year of discretionary capital to allocate.

The first allocation is uncontested. CHF 14,516 maxes both pillar 3a accounts (CHF 7,258 each, both employed), which is exactly the deposit-building case from the down-payment piece. That leaves about CHF 10,484 a year as the marginal decision.

For this couple, at this margin, the buy-in is the wrong move:

• A pillar 2 buy-in within three years of the planned WEF property withdrawal triggers the art. 79b recapture. With a purchase five to seven years out, an early buy-in risks the freeze, and a late one is pointless.
• The natural alternative over a five-to-seven-year horizon is a taxable equity investment. Switzerland does not levy capital-gains tax on private individuals for most securities, so the post-tax return on a diversified portfolio is essentially the gross return. At any reasonable equity assumption that beats the 1.25% pillar 2 mandatory rate over the holding period.
• So the marginal capital goes to a taxable brokerage account (or a higher-yielding savings vehicle if the couple is risk-averse), and any buy-in is deferred until after the property purchase plus the three-year quarantine.

Five heuristics generalise the decision.

1. Property purchase within three years? Skip the buy-in and max pillar 3a. The buy-in deduction would be recaptured at the WEF withdrawal anyway.
2. Long retirement horizon, no near-term property plans? The buy-in becomes viable. Its uncapped deduction is a powerful single-year shield in a high-income year, and the three-year freeze does not bind when no withdrawal is imminent.
3. Unmarried partner? Pillar 3a wins decisively on inheritance flexibility (BVV3 art. 2 versus BVG art. 20a).
4. Equity-invested 3a over a 20-year-plus horizon? The compound differential against the 1.25% mandatory rate is large enough to tilt 3a-first even outside the property case.
5. Self-employed without pillar 2? No buy-in option exists, and the large 3a (CHF 36,288 cap) dominates by default. The question does not arise.

The default for most salaried Swiss residents under 60 with a property goal in the next five to ten years: max pillar 3a first. The pillar 2 buy-in is the more specialised tool, for households with very high income, established occupational cover, no near-term property plans, and a comfortable retirement horizon.