The Silent Portfolio Killer: How Fees Compound Against You for Decades

Why you probably have a leaky portfolio

There are two ways to improve long-term investment returns. The first is finding better investments. The second is losing less of what you already earn.

Most investors obsess over the first. They research stocks, chase returns, rebalance constantly, and hunt for the next outperformer. Very few spend equivalent effort on the second — because fees feel small, invisible, and unglamorous. They are also the more reliable lever. You cannot underwrite next decade’s alpha. You can underwrite next decade’s cost.

A 1% annual fee on a $100,000 lump sum does not cost $1,000 in any year that matters. It costs the ending gap between 7% and 6%. Over 30 years that gap is about $187,000: $761,000 versus $574,000. That is not the sum of invoices. It is the wealth the fees prevented from compounding.

This is not an argument for being cheap. It is an argument for being systematic. Every structural friction you remove compounds for the rest of the investing life. The work is unglamorous. The arithmetic is not.

How the figures in this piece are calculated

Unless a section says otherwise:

• Gross return is 7% a year, nominal, compounded annually.
• AUM fees (MER, percentage custody) reduce the net return. A 0.69% MER on a 7% gross return is modelled as 6.31%.
• Flow costs (brokerage, FX conversion) reduce the dollars invested that year. The leaked dollars then miss out on 7% compounding.
• Tables are lump sum unless contributions are specified. Figures are rounded. Real returns, inflation, and tax on the growth itself are ignored so the fee effect is visible on its own.

Those assumptions are conservative in one sense and generous in another. Markets will not return 7% every year. Fees will not stay a fixed percentage of a smooth path. The point is comparison, not a forecast of your terminal wealth.

Do not add brokerage and FX percentages onto MER and call the total an AUM fee. They are different objects. MER hits the whole pile every year. Tickets and FX hit new money. The closing illustration later uses an AUM-equivalent friction rate only so two otherwise identical contribution machines can be compared in one line.

Here is the decision framework.

Part 1: ETF management fees — the fee you pay while you sleep

The mechanics

Every managed fund and ETF charges a management expense ratio (MER): an annual fee expressed as a percentage of assets, taken from the fund’s NAV in tiny daily increments. There is no invoice. The money never hits your return line.

MER is not the whole cost of owning an ETF. Bid–ask spreads, buy/sell spreads at creation/redemption, and tracking difference all sit on top. For a large, liquid broad-market fund those extras are usually small. For a thin thematic they are not. Use MER as the starting audit, not the complete bill.

The decision point

Compare two funds tracking a similar equity universe — one at 0.03% (a broad US market vehicle such as Vanguard’s US-listed VTS), the other at 0.69% (a typical thematic or active alternative). The gap is 0.66 percentage points.

On a $200,000 lump sum at 7% gross:

The 0.69% fund needed to beat the cheaper fund by at least 0.66% a year, after its own costs, just to break even. Most active managers do not clear that hurdle over long windows. S&P’s SPIVA Australia scorecard for year-end 2025 found 74% of Australian Equity General funds underperformed the S&P/ASX 200 in 2025 and 87% over 15 years. In Global Equity General, 70% underperformed in 2025 and more than 95% over 10- and 15-year windows.

Most thematic ETFs fail the same test against their parent index, not against cash.

The same arithmetic applies inside the cheap cohort. On that same $200,000 over 30 years, the 0.10% gap between VGS at 0.18% and BGBL at 0.08% is about $41,000. Liquidity, fund size, and distribution frequency can still justify VGS for some holders. The fee difference is no longer a rounding error once the position is large.

What to do differently

Audit every fund you hold. For core index exposure, the question is not “is this fund good?” It is “is this fund good enough to justify paying more than the liquid alternative?”

A common low-friction Australian core, funded in AUD on the ASX, looks like this:

• Global developed equities: BGBL (0.08%) or VGS (0.18%)
• US equities as a separate sleeve: IVV (0.04%)
• Australian equities: A200 (0.04%) or VAS (0.07%)

IVV versus VTS. Both track large-cap US equities. IVV is ASX-listed and bought with AUD, so there is no conversion spread on the way in. VTS is US-listed and cheaper on MER (0.03%), but buying it is usually an FX event unless you already hold USD. For an investor whose cash flow is in AUD, IVV is the cleaner default. VTS is a cost tool for people who have already decided to hold US-listed assets.

Treat anything above 0.30% MER in a core sleeve as needing a specific job. NDQ at 0.48% is not a substitute for IVV. It is a Nasdaq-100 concentration bet that costs twelve times IVV’s MER. On a $100,000 lump sum over 20 years at 7% gross, that fee gap alone is about $30,000 of terminal value — before you count the different risk.

Specialist factor exposures (value, quality, small-cap) can justify a higher fee when the product actually delivers the factor and you have a reason to hold it through the years it looks broken. Compare 0.35% against 0.55% on tracking difference and factor loading, not on the brochure. Premia in the academic literature are not a payout schedule.

A usable rule for thematics: if the fund has not beaten its parent index after fees over a full market cycle — not a three-year burst — it is a cost centre. Three-year rolling outperformance is mostly noise.

Part 2: Brokerage frequency — the tax on trading discipline

The mechanics

Every buy, sell, and rebalance can trigger brokerage. On older bank platforms that is often $10 to $20 a trade. On current retail platforms it is commonly $3 to $6.50 on the ASX, and $0 to a few dollars on US-listed stock. The headline looks small. Frequency makes it material.