Weighted Portfolio Expense Ratio: Why Coverage Matters

0.0422% with 90% coverage can be a more honest portfolio expense ratio than a smaller number labeled complete. A weighted expense ratio averages annual fund expenses according to position value. When a fund's fee is unavailable, the average needs to say which holdings it covers. This synthetic example, computed September 14, 2026 using the PortLens fee methodology, shows exactly why.

How do you calculate a weighted expense ratio?

Multiply each position's current value by its annual expense ratio, add those amounts, and divide by the value included in the calculation. A large position has more influence than a small one. Counting each fund once gives every label the same vote, even when the money behind those labels is very different.

Written as a formula:

Weighted expense ratio = sum of (position value × annual expense ratio) ÷ included position value.

Keep the units consistent. If a fee is written as 0.03%, its decimal fraction is 0.0003. A spreadsheet cell formatted as a percentage normally stores that fraction. Typing 0.03 into a plain decimal calculation instead would mean 3%, a different input before the averaging has even started.

The denominator is the value underneath the line. It might represent all your holdings, only your funds, or only the positions with known costs. Each scope answers a different question. A percentage without its denominator is like a receipt with the quantity column torn off: the final number survives, but the calculation cannot be checked.

In its July 23, 2025 investor bulletin, rechecked September 14, 2026, the SEC explains that a fund's expense ratio expresses annual operating expenses relative to average net assets. The bulletin separates those expenses from other charges. Combining fund ratios therefore measures one defined layer of cost. It doesn't turn that layer into a complete bill for investing.

What happens when one fund's fee is missing?

An unavailable fee must stay unknown; putting zero in its cell makes the average look lower without supplying any new evidence. PortLens excludes that fund from the covered average and identifies the excluded holding alongside the headline's coverage.

Here is the full illustration, computed September 14, 2026. It describes no real investor or security. Imagine a first fund, a second added later for a different exposure, an ordinary company share, and finally another fund whose expense ratio the data source cannot supply. The names and rates are synthetic arithmetic inputs, not current prospectus figures or a suggested allocation.

The downloadable input CSV deliberately leaves the unknown ratio blank. Zero and an empty cell must remain different when the file is imported into a spreadsheet.

Using the same calculation as the scan, Fund A contributes $18 and Fund B contributes $20 to a static annual-cost illustration. The ordinary share contributes zero annual fund expense. The known contribution is therefore $38, over $90,000 of covered value:

$38 ÷ $90,000 = 0.042222…%, with 90% coverage.

That $38 assumes the displayed values and rates stay fixed for a year. It is neither a statement of expenses actually paid nor a forecast of next year's charge. The final fund's expense remains outside it. The empty cell is an unanswered question; it cannot serve as a discount.

Why do different denominators produce different answers?

The denominator changes the population being averaged, so two correct calculations can disagree because they describe different holdings. A third calculation can be wrong because it silently treats missing data as a known zero.

These are outputs from the same synthetic inputs above, evaluated September 14, 2026 with the PortLens calculation and stated scope. The funds-only case excludes the ordinary share as well as the unknown fund. The zero-filled case deliberately changes the unknown input to demonstrate the error; it is not how PortLens handles missing ratios.

The first two rows share the same known $38 contribution. Their difference comes entirely from whether the ordinary share belongs in the average. Neither number alone says the underlying fund prices changed.

A simple unweighted average of the two known fund rates would be 0.065%. That gives the smaller fund the same influence as the larger fund, despite their unequal position values. It answers a question about two listed rates, not the weighted cost carried by the positions in this example.

Multiplying the covered average by the entire portfolio value creates another error. The covered rate belongs to $90,000 here. Applying it to $100,000 assumes the missing position carries the same rate as the covered group. That assumption may be useful in a separately labeled model, but it is not an observed fee.

Coverage is the second dial next to the percentage. Turning one without looking at the other can make a changed data set look like a changed cost. A like-for-like comparison keeps both the included positions and the fee definitions visible.

Do individual stocks count as having zero expenses?

An ordinary company share has no annual fund expense ratio, which is why PortLens includes a confirmed non-fund holding at zero for this particular measure. That doesn't say every cost of owning or trading the share is zero.

An ETF whose ratio is missing is different. So is a fund that a price provider happens to label as a stock. PortLens checks its SEC fund registers as well as the supplied asset type. Those registers help distinguish ordinary company shares from fund structures that cannot safely be assumed to carry no annual fund fee.

If the required register cannot be read, the product reports the cost as unknown. It doesn't classify everything as an ordinary share and announce a fee-free portfolio. A ruler with the labels rubbed off cannot certify that every object is zero inches long.

Coverage for expenses is also separate from coverage for holdings. A scan may resolve the companies inside a fund while lacking that fund's expense ratio. Conversely, a published fee can exist even when constituent information is incomplete. The look-through explanation describes the holdings side; the fee card describes the expense side. Those two coverage figures are not interchangeable.

How can I check the fund-cost measure in PortLens?

The free scan shows a weighted expense-ratio headline with its coverage and excluded funds when the required data is available. An account is not required for that headline; per-fund cost detail is a separate Pro feature. Missing information remains visible on the free result.

You can scan your own holdings by entering supported ticker symbols and share counts. The scan derives position values from those inputs and current available prices. Dollar values in this article explain the arithmetic; they are not instructions to type allocation percentages into the share-count field.

The expense data comes from supported SEC prospectus disclosures. When a net ratio after reported waivers is available, the product uses it; otherwise it uses the available gross ratio. Some fund structures and non-US funds fall outside that source's coverage. A supported ticker does not guarantee a supported expense-ratio record.

At the result, read the headline together with the excluded holdings. A lower percentage after adding a position can come from a different denominator. A lower percentage after a data refresh can come from a newly available ratio. Neither observation alone explains what changed.

A verified account saves the entered holdings for later use. It does not preserve today's prices and fee data as a frozen historical report. The next calculation can reflect refreshed inputs. The CSV above is retained precisely so this educational example remains checkable after live data moves on.

Which source details make two fee snapshots comparable?

Two snapshots need the same position scope, matching share classes and clearly identified fee definitions before their averages can be compared. Date labels help identify those inputs; changing a page's date cannot make an old fee current.

The SEC bulletin explains that different classes of a mutual fund can carry different expenses even when they invest in the same underlying securities. A fund name alone is therefore insufficient. A reproducible calculation identifies the actual class or ticker and preserves the rate used for it.

For a comparison you can audit, retain the position values and their valuation date, the expense ratios with their source dates, whether each rate is gross or net, and the treatment of every unavailable value. A waiver can change which published rate applies. A price move can change the weights even while all the fund rates stay constant.

This article's illustration avoids both moving inputs by publishing fixed synthetic values and rates. It demonstrates the calculation's treatment of missing information. It makes no claim about a current fund, typical investor costs, or whether a security is worth owning. A complete numerical average needs complete inputs; an incomplete one earns its usefulness by stating exactly what is missing.

Key takeaways

• A weighted expense ratio uses position values, so unequal holdings do not receive equal weight.
• Unknown fund expenses are excluded from PortLens's covered average and shown through coverage and excluded holdings.
• Ordinary company shares contribute zero annual fund expense; that is a narrower statement than having no investing costs.
• Covered-portfolio and covered-funds-only averages use different denominators and can both be valid when labeled.
• A static dollar-cost illustration is not an actual charge or a prediction, and fee coverage is separate from holdings coverage.

This article is for information and education only and is not investment advice. Analytics referenced are computed as described in our methodology; see our disclosures.