Skip to content
Financial Context Wiki

Portfolio Turnover and Trading Cost: From Disclosed Rate to Net Return

For educational purposes only; not investment advice.

Direct answer

Section titled “Direct answer”

Portfolio turnover measures how much of a portfolio is replaced during a period. The SEC Form N-1A calculation for a registered fund generally divides the lesser of eligible purchases or sales by average monthly portfolio value, excluding specified short-term securities:

Turnover rate = min(eligible purchases, eligible sales) ÷ average portfolio value

Turnover is not a direct expense ratio. It is a trading-activity indicator. The economic cost depends on the buy and sell notional, bid-ask spreads, commissions and fees, market impact, execution delay, taxes, cash drag, and any benefit obtained from the trades.

One-sided turnover versus two-sided cost

Section titled “One-sided turnover versus two-sided cost”

The lesser-side formula helps prevent net subscriptions or withdrawals from being mistaken for portfolio replacement. If a fund with $100m average assets buys $80m and sells $90m of eligible securities, turnover is 80%. It does not mean total traded notional was $80m; the two sides total $170m.

Costs can be organized as:

Net strategy return ≈ gross paper return - explicit fees - spread - market impact - delay/opportunity cost - taxes

For a trade, implementation shortfall compares the actual portfolio result with a hypothetical execution at a chosen decision or arrival price. It can capture commissions, spread, price movement while executing, and the opportunity cost of shares not completed. The benchmark timestamp and treatment of subsequent market moves must be documented.

Fund expense ratios generally do not include every trading cost. Brokerage commissions can be reflected in transaction accounting, while spread and market impact are embedded in execution prices. A zero-commission account therefore can still incur meaningful turnover costs.

Converting turnover into annual drag

Section titled “Converting turnover into annual drag”

A portfolio has $100,000 average value, eligible purchases of $220,000, and eligible sales of $200,000:

Turnover = $200,000 ÷ $100,000 = 200%

Total traded notional is $420,000. If measured average implementation shortfall is 0.10% per executed dollar across both sides:

Execution cost = $420,000 × 0.10% = $420

Portfolio drag = $420 ÷ $100,000 = 0.42%

An equivalent replacement-capital method might apply a 0.20% round-trip cost to the $200,000 lesser side, producing $400 or 0.40%. Do not add both estimates; they are alternative approximations using different definitions.

If a backtest reports 8.0% gross return, subtracting 0.42% gives about 7.58% before taxes and other costs. Taxes cannot be estimated from turnover alone: gains, losses, holding periods, account type, tax lots, jurisdiction, and investor circumstances matter. A sale at a loss can have a different effect from realizing a short-term gain.

Measurement and decision checklist

Section titled “Measurement and decision checklist”
  • Reconcile broker executions or fund reports into purchases, sales, commissions, fees, and average portfolio value using one consistent period.
  • Separate subscriptions, withdrawals, distributions, maturities, derivatives, short-term instruments, in-kind ETF activity, and corporate actions according to the chosen definition.
  • Measure execution against the quote or midpoint at a documented decision or arrival time, not an arbitrary later price.
  • Segment costs by liquidity, order size, participation rate, volatility, session, venue, and market-cap category.
  • Include incomplete orders and time spent in cash; analyzing only completed fills creates selection bias.
  • Compare after-cost, after-tax results with a suitable low-turnover benchmark and with the strategy’s forecast error.
  • Set a trade hurdle: the expected improvement from replacing a holding should exceed implementation cost, tax cost, and uncertainty by a meaningful margin.
  • Use contributions and tolerance bands to rebalance when they reduce unnecessary selling, but exit when the underlying thesis or risk limit truly fails.

Low turnover is not automatically efficient: retaining a deteriorating asset can be more costly than selling it. High turnover is not automatically harmful if a repeatable edge exceeds all implementation costs. The relevant evidence is realized net performance under realistic capacity, not a universal “good” turnover threshold.

Common misconceptions

Section titled “Common misconceptions”
  • “A 100% turnover rate means total trades equal portfolio value.” Under the lesser-side convention, purchases plus sales can be about twice that value or differ because of flows.
  • “Turnover multiplied by one-way cost always gives total cost.” Confirm whether the cost assumption is one-way, round-trip, or per total notional.
  • “Zero commissions mean zero trading cost.” Spread, impact, delay, fees, and taxes can remain.
  • “Expense ratio includes all turnover cost.” Important execution costs can be embedded in trade prices rather than the reported ratio.
  • “Lower turnover is always better.” Strategy horizon, information decay, risk changes, and realized net value added matter.
  • “A backtest’s gross return is investable.” Realistic fills, capacity, unavailable liquidity, taxes, and cash timing can materially reduce it.
Section titled “Related topics”

Authoritative sources

Section titled “Authoritative sources”