Quant Memo
Foundational

Commissions, Exchange Fees And Transaction Taxes

The layered, unavoidable costs charged on every trade before spread or market impact even enter the picture — broker commissions, exchange and clearing fees, and government transaction taxes.

Before you think about spread or market impact, every trade already carries a floor of unavoidable cost: someone has to be paid for routing the order, the exchange has to be paid for matching it, and in some markets the government takes a cut just for the transaction happening. These costs are usually small per share, but they're paid on every single trade, so for high-turnover strategies they compound into a real drag that's easy to underestimate if a backtest only accounts for spread and slippage.

The three layers

Commissions go to the broker for routing and executing the order — anywhere from a fraction of a cent per share for institutional flow to a flat per-trade fee at a retail broker (or zero, where the broker instead earns from payment for order flow). Exchange and clearing fees are charged by the venue and clearinghouse for matching and settling the trade; many exchanges use a maker-taker model, paying a rebate to orders that add liquidity (resting limit orders) and charging a fee to orders that take it (marketable orders), so the same trade can cost differently depending on which side of it you're on. Transaction taxes, where they exist — the UK's stamp duty, France's financial transaction tax, historically similar levies elsewhere — are charged by the government on the trade value itself, independent of anything the broker or exchange does, and typically apply asymmetrically (only on buys, or only on certain instruments).

Worked example

A fund buys $1,000,000 of a UK-listed stock. Broker commission is 2 basis points: $200. The exchange charges a small per-share fee that nets to roughly $50 for this size. UK stamp duty reserve tax is 0.5% on the purchase (not the sale): $5,000. Total non-spread, non-impact cost: $5,250, or 52.5 basis points — more than 25 times the commission alone, and the stamp duty alone would have been invisible to a cost model that only tracked broker fees. On the sale, no stamp duty applies, so the round-trip cost is asymmetric: heavier on the buy leg than the sell leg, which matters for strategies that hold UK equities briefly and turn over frequently.

commission (\$200) exchange fee (\$50) stamp duty (\$5,000) Cost layers on a \$1m buy
The government transaction tax dwarfs commission and exchange fees combined on this UK purchase — a cost that only shows up on the buy leg, not the sell.

What this means in practice

A strategy's backtest that only models bid-ask spread and market impact is missing a cost that's fully known in advance and easy to model precisely — unlike spread or impact, commissions, exchange fees, and transaction taxes are published rates, so there's no excuse for leaving them out. For high-frequency or high-turnover strategies these costs alone can erase an edge that looks solid before fees. They also shape strategy design directly: a maker-taker fee schedule rewards resting limit orders over aggressive ones, and a jurisdiction with a transaction tax on buys only will favor strategies that turn over less or that structure trades to minimize taxed legs (for instance, using instruments like CFDs or futures that some jurisdictions exempt from the tax).

Commissions, exchange/clearing fees, and transaction taxes are known, published costs paid on every trade regardless of spread or market impact — they belong in a backtest as certainly as the trade itself, and for high-turnover strategies they can be the largest single cost layer.

When comparing execution costs across venues or instruments, check whether a transaction tax applies asymmetrically (buy-only, or specific to cash equities but not futures) — the same economic exposure can sometimes be achieved through an untaxed instrument, turning a real cost difference into a structuring decision rather than a fixed expense.

Related concepts

Practice in interviews

Further reading

  • Harris, Trading and Exchanges, ch. 15
ShareTwitterLinkedIn