- Get link
- X
- Other Apps
An ETF's expense ratio is only half the story: a tighter bid-ask spread can outweigh a lower annual fee over a short holding period. For two funds already suitable for your goal, estimate both costs before treating the cheapest annual fee as the cheapest choice.
Separate a recurring charge from a trading friction
An expense ratio is an annual percentage paid from fund assets, not a one-time entry fee. Read the fund's fee table using the SEC's fund-fee guide. The bid-ask spread is different: FINRA explains ETF trading costs and why commission-free does not mean cost-free.
Put both costs on the same timeline
Hypothetical $10,000 investment, constant value and equivalent exposure: Fund A charges 0.05% annually with a 0.20% spread; Fund B charges 0.15% with a 0.04% spread. Assume one purchase at the ask and one sale at the bid, with the same midpoint and spread. The approximate round-trip drag is one full spread—not two.
| Holding period | Fund A | Fund B |
|---|---|---|
| 3 months | $21.25 | $7.75 |
| 1 year | $25 | $19 |
| 3 years | $35 | $49 |
A costs approximately $20 + $5 × years; B costs $4 + $15 × years. A's $16 extra trading drag is offset by $10 annual savings after 1.6 years, about 19 months. This is an illustrative calculation, not a forecast. Vanguard's cost-of-ownership discussion also emphasizes holding time.
Stress-test the crossover before using it
The comparison above is an original worked example, not a quotation for two real funds. A useful sensitivity test is to change the uncertain spread while keeping the annual fee difference at 0.10 percentage points. Approximate crossover years = extra round-trip spread percentage ÷ annual expense-ratio savings percentage. Use percentage points consistently: 0.16 ÷ 0.10 = 1.6 years.
| Extra spread for the lower-fee fund | Extra drag on $10,000 | Crossover at 0.10-point annual savings |
|---|---|---|
| 0.08 percentage points | $8 | 0.8 years (about 10 months) |
| 0.16 percentage points | $16 | 1.6 years (about 19 months) |
| 0.32 percentage points | $32 | 3.2 years (about 38 months) |
This shows why a quoted spread should not be treated as a permanent property. Doubling the extra trading drag doubles the waiting time in this simplified model. A larger investment scales both dollar costs and savings proportionally, so it does not change the crossover here; a fixed brokerage charge would break that proportionality. If there is no positive annual fee saving, this formula provides no future crossover.
A repeatable comparison worksheet
- Confirm that exposure, benchmark, currency and risks are sufficiently comparable.
- Record each prospectus expense ratio, any temporary fee-waiver expiry, and your broker commission schedule.
- Record the date and time of the live bid and ask. Calculate spread percentage as (ask − bid) ÷ midpoint × 100. Historical median spreads are context, not a guaranteed execution price.
- Estimate the number of purchases and sales, then rerun the costs for your intended holding period and a wider-spread scenario.
Repeated trading can add spread costs each time, while the original table assumes only one purchase and sale. Do not add two full spreads for a single round trip under its unchanged-midpoint assumption: entry and exit each contribute approximately half a spread. Different execution prices, market movement and changing spreads require a more detailed model.
Primary reference: the SEC investor bulletin on ETFs explains trading spreads and the documents investors should examine. Calculations and sensitivity cases here are illustrative editorial analysis, added October 4, 2026.
Reject a false like-for-like comparison
First compare holdings, benchmark and risk. A bond ETF and a stock ETF are not substitutes because their fees cross. For comparable funds, read current prospectuses, inspect live quotes and estimate your own holding period.
Actual spreads can change before you sell; balances fluctuate, so annual dollar fees will too. This screen excludes commissions, taxes, premiums or discounts to asset value, and tracking differences. It does not measure investment returns or justify trading an unsuitable fund. Do not deduct the expense ratio again from performance already reported net of fund expenses.
U.S. investor education, not investment advice; checked September 25, 2026.
Comments
Post a Comment