← All posts⏳ Hold or trade

Which ETFs to hold long term, and which need timing

VAS and SEMI are both called ETFs, but they behave nothing alike. One is built to hold for decades. The other can double in a year and then give it back. The data shows which is which, and why timing matters for one group only.

Joshua Stega, financial adviser6 min readUpdated 11 November 2026β–Ά Includes video
The short answer

Broad core ETFs are built to hold. Across 12 big core funds, five-year returns to 30 September 2026 ranged from +33% to +88%, and none lost money. Thematic and sector ETFs need a view on timing. Across 25 of them, results ranged from βˆ’14% to +352%, and last year's return was on average 24 percentage points away from their five-year trend. Hold the core; manage the satellites.

Two kinds of ETF, two different jobs

A core ETF tracks a whole market: the ASX 200, the S&P 500, the developed world, or a mix of all of them in one fund. Think VAS, A200, IVV, VGS or DHHF. Its returns come from owning economies over decades.

A thematic or sector ETF owns one slice: semiconductors, gold miners, Asian tech, small companies, clean energy. Think SEMI, GDX, ASIA or ATEC. Its returns come from a trend, and trends start and end.

To test the difference, we took 12 widely held core ETFs and 25 thematic, sector and regional ETFs, all with at least five years of history, and compared their returns using the TMX monthly report to 30 September 2026.

+33% to +88%range of core ETF returns over 5 years
βˆ’14% to +352%range of thematic ETF returns over 5 years
8Γ—more drift between 1-year and 5-year returns

Core funds cluster. Themes scatter.

Five years of returns

Best, average and worst: core vs thematic ETFs

Core (12 ETFs)Thematic and sector (25 ETFs)
Best5-year total return
+88%
+352%
Average5-year total return
+54%
+88%
Worst5-year total return
+33%
βˆ’14%
Cumulative total returns to 30 September 2026. Core: VAS, A200, IOZ, STW, VGS, IVV, VTS, VEU, DHHF, VDHG, VDGR, VGAD. Source: TMX Australia monthly funds report.

The thematic group earned more on average, but that average hides an enormous spread. The best, SEMI, returned +352% in five years. The worst, CLNE (clean energy), lost 14%. ATEC (Australian tech) and NDIA (India) were roughly flat after five years.

In the core group, the worst result was VDGR at +33%, and that fund is 30% bonds by design. Picking any one of them blindly would have worked out reasonably. Picking a theme blindly was a coin toss between a big win and a lost five years.

$10,000 invested five years ago

Core funds land close together; themes do not

$13,672
VAS
$15,537
DHHF
$18,779
IVV
$45,200
SEMI
$32,360
GDX
$9,450
ATEC
$8,580
CLNE
Value at 30 September 2026 with distributions reinvested, before tax and brokerage. Source: TMX Australia monthly funds report.

The timing test: last year vs the long-run trend

A simple way to see whether timing matters is to compare each fund's last 12 months with its average year over five years. We call the gap drift. Low drift means the fund just compounds. High drift means it moves in bursts, hot then cold.

  • Core ETFs: average drift of 3.1 points. DHHF returned 8.1% last year against 9.2% a year over five years. VDHG: 7.5% against 7.7%.
  • Thematic ETFs: average drift of 24.0 points, eight times as much.
Drift

Last 12 months minus the 5-year average year

β—€ Colder than its trendHotter than its trend β–Ά
IKOSouth Korea
+105.2
SEMIsemiconductors
+73.8
ASIAAsian tech
+37.8
DHHFall-in-one core
βˆ’1.1
VGSglobal core
βˆ’1.7
GDXgold miners
βˆ’16.0
NDIAIndia
βˆ’17.5
ATECAustralian tech
βˆ’31.9
Percentage points: 1-year total return minus 5-year annualised total return, to 30 September 2026.

IKO (South Korea) returned +125% last year against about 20% a year over five. ATEC fell 33% last year. GDX has been one of the best funds over five years, at about 26% a year, but returned only 10.5% in the latest 12 months as gold miners cooled. When you bought and sold these funds mattered far more than whether you owned them.

The hold list: funds built to compound

ETFExposureFee1 year5 years (a year)
A200ASX 2000.04%+2.9%+6.9%
VASASX 3000.07%+2.2%+6.5%
IVVS&P 5000.04%+10.1%+13.4%
VGSDeveloped world0.18%+10.1%+11.8%
VEUWorld outside US0.07%+14.0%+9.2%
DHHFAll-in-one, 100% growth0.19%+8.1%+9.2%
VDHGAll-in-one, 90% growth0.27%+7.5%+7.7%

Total returns to 30 September 2026. Source: TMX Australia monthly funds report.

Low fees, thousands of companies, and returns that track the economy. Australian shares have had a quiet year, but that is a normal part of holding a market, not a reason to sell it.

The watch list: funds where timing matters

ETFThemeFee1 year5 years (a year)
SEMISemiconductors0.45%+109.0%+35.2%
ASIAAsian tech0.67%+55.4%+17.6%
HACKCybersecurity0.67%+25.8%+15.2%
GDXGold miners0.53%+10.5%+26.5%
SMLLAustralian small companies0.39%+4.3%+4.7%
ATECAustralian tech0.48%βˆ’33.0%βˆ’1.1%
CLNEClean energy0.65%+15.2%βˆ’3.0%

Total returns to 30 September 2026. Source: TMX Australia monthly funds report.

The cost of a theme

Average fee: core vs thematic

Core ETFs12 funds
0.12% a year
Thematic and sector ETFs25 funds
0.55% a year
Simple average of management fees for the funds in this study. Source: TMX Australia monthly funds report, September 2026.

Themes also cost more, about 0.55% a year on average against 0.12% for the core. That is another reason not to hold a fading theme for years out of habit.

What "trade" really means here

This is not day trading. For satellite holdings, it means having a rule for when you get in and when you get out, and reviewing it regularly. ETF Adviser's approach is to follow the trend: markets tend to move through quiet bases and then trends, so the aim is to buy a theme as a new uptrend begins and step aside when it breaks down.

  • Enter on strength, not on hope. A fund breaking out of a long base is a different proposition from one that has already doubled.
  • Have an exit before you buy. A stop-loss or a trend rule stops a fading theme turning into a long-term loss.
  • Expect false starts. Some breakouts fail. Small positions keep those mistakes affordable.
  • Keep it in proportion. Satellites work best as small positions around a large, steady core.
The one rule: the broader the fund, the less timing matters. If an ETF owns a whole market, hold it. If it owns one theme, know your exit before you buy.

Putting it together

This is the logic behind core and satellite investing. The core, often most of the portfolio, sits in one or a few broad ETFs and is left alone. The satellites are smaller positions in themes, each with an entry and exit plan. Read more in our guides to building your core and building your satellites, or browse tech, gold and Asia ETFs.

Compare a core fund with two themes β†’

Common questions

Can you buy and hold thematic ETFs?

You can, but the results vary widely. Over five years to September 2026 thematic ETFs ranged from a 14% loss to a 352% gain. Holding one through a full cycle can mean giving back years of gains, which is why many investors manage them actively.

Which ETFs are best for long-term holding in Australia?

Broad, low-cost funds that own whole markets, such as ASX 200 or ASX 300 trackers, global share ETFs and diversified all-in-one funds. Their returns have been far more consistent than sector or theme funds.

Is it bad to trade ETFs?

Frequent trading of core funds usually adds costs and tax without adding return. For narrow thematic funds, having clear entry and exit rules can make a large difference, because their returns come in bursts.

How much of a portfolio should be in thematic ETFs?

There is no single answer, but most core and satellite investors keep themes as a minority of the portfolio and size each position so that a failed theme cannot do serious damage.

Joshua Stega

Joshua Stega is a financial adviser with 20 years' experience and the founder of ETF Adviser. He built ReviewETF to make every Australian ETF easy to find and compare.

General information only. It does not take into account your objectives, financial situation or needs. Consider the PDS and TMD before investing. Past performance is not a reliable indicator of future performance. Data as at 30 September 2026.

Keep in touch with the markets

Get Joshua's weekly ETF research, trade alerts and the model portfolio. Free for 30 days.

Start your free trial

Keep reading