Most NZ investors own six ETFs. Two would do the job better.

Start with the thing that trips people up first: a "Global" or "Total World" fund is not a neutral, evenly-spread basket. It's roughly 60-65% United States by weight, because that's what the US actually is as a share of global market capitalisation. Add a separate US 500 fund on top of that "for extra US exposure" and you haven't diversified anything - you've just overweighted the US further while telling yourself you've spread your risk.

This is the overlap problem, and it's the single most common thing wrong with the six-ETF portfolios I come across. Global fund plus US fund plus NZ fund plus an Australian fund plus a property fund plus a bond fund looks thorough. Half of it is the same exposure counted twice.

What's actually in a NZ-domiciled ETF, tax-wise

Every fund listed by Kernel or Smart is structured as a Portfolio Investment Entity. That matters more than the ticker does. PIE tax is settled inside the fund at your Prescribed Investor Rate, capped at 28% no matter your income, and none of it touches the Foreign Investment Fund rules - not at $10,000 invested, not at $500,000. A NZ-domiciled global equity fund and a directly-held US ETF bought through Hatch or Interactive Brokers can hold near-identical underlying companies and sit under completely different tax regimes. One never touches FIF. The other does, once your cost basis crosses $50,000.

That's not a reason to avoid direct overseas ETFs - there are legitimate reasons to hold them, and Issue 3 covered when that makes sense. It's a reason to know which bucket a given holding sits in before you assume your portfolio is simpler or more complicated than it is.

The fee gap most comparisons miss

Kernel's cheapest, most diversified option - Total World, 0.12% p.a., which just holds Vanguard's own Total World ETF wrapped in a PIE - is not available on Kernel's free Core plan. You need the Plus tier, $50 a year, to unlock it. Sign up on Core because Kernel looks cheap and you're actually paying 0.25% to 0.50% on funds that are individually less diversified than the one you can't access yet.

Smart runs the wider shelf - over 40 funds across NZ, Australian, US, international, thematic, property and bond categories, more depth in any single slice of the market than Kernel offers. If you specifically want exposure to something narrow - global infrastructure, dividend aristocrats, a hedged bond fund - Smart is more likely to have built it. For a plain global equity core holding, the two are increasingly similar on cost; check the current TER on the specific fund you're buying rather than trusting a headline "cheapest provider" claim, because both have moved their pricing around this year.

A working starting point, not a template

For someone building a long-term portfolio from scratch, wanting genuine global diversification without duplicating exposure: one global equity fund as the core - Total World or an equivalent all-world index, not a US-only fund dressed up as "global." Add a New Zealand fund alongside it if you want a deliberate home tilt (there are reasonable arguments for this - imputation credits being one, covered in Issue 12), sized as a conscious percentage rather than a leftover habit from before you understood FIF. Everything past those two funds should be able to answer a specific question: what does this actually add that the first two don't already cover.

A bond or cash fund earns its place once capital preservation matters to you, not before - that's a risk-tolerance and time-horizon decision, not a diversification one, and it's a different conversation than the one this issue is having.

Where this goes wrong

Buying funds because a comparison site ranked them, not because you checked what's underneath. Two funds that both hold Apple, Microsoft and Nvidia as top-five positions aren't two different investments, whatever their names suggest. Chasing a 0.05% fee difference between two nearly-identical global funds while ignoring that you're also paying for a third fund with 80% overlap with the first two.

None of this means fewer funds is automatically better. It means each fund in the portfolio should be there for a reason you could actually state out loud, not because it showed up on a "best ETFs NZ 2026" list.