Transparent costs

Commissions, exchange fees, taker fees, clearing: the costs you can read off a fee schedule. Easy to count, and easy to underestimate at high turnover.

The idea

What counts as a transparent (explicit) trading cost?

An explicit cost is any charge that is billed, line-itemed and knowable in advance. It appears on a contract, a fee schedule or an invoice, so you can compute it exactly before the trade ever happens. That is the dividing line from implicit costs, which are never invoiced (the spread you cross, slippage, the impact of your own order) and must be estimated. Explicit costs are arithmetic; implicit costs are statistics.

The test is simple. If you can look it up in a published schedule and multiply, it is explicit. If you only learn it by comparing the price you wanted to the price you got, it is implicit. The explicit families are commissions, exchange/ECN fees (including the maker-taker schedule), clearing and settlement, regulatory fees, transaction taxes and market-data fees. Some are per-share, some per-trade, some per-notional, some a fixed monthly bill; the units differ, which is exactly why people mis-total them.

The honest framing decides how much this page matters to you. For a low-frequency investor the explicit stack is the whole cost conversation. For an HFT it is the small, easy half, but a half that, mis-counted, still flips a thin per-trade edge, because in high-frequency trading the gross edge and the cost stack are the same order of magnitude. You cannot afford to forget a line.

Commissions: what your broker charges

A commission is the fee your broker or futures commission merchant charges to route and execute an order. Historically per-share or per-contract, it is now often zero for retail equities, paid for instead by payment-for-order-flow. For professional and HFT flow it is a tiny negotiated per-share or per-notional rate, frequently with monthly minimums and tiered breakpoints by volume.

The units differ by asset class, and that is the trap. Equities: retail US commissions are widely zero as of 2026; professional DMA rates are fractions of a cent per share, tiered by monthly volume. Futures: charged per contract per side by the FCM, plus the exchange's own per-contract fee, small in dollars but scaling with size, not notional. Crypto: the commission and the exchange fee collapse into one published trading fee (a maker/taker basis-point rate) on a centralised venue; on a DEX there is no commission but a protocol fee plus gas.

The catch worth internalising: zero commission is not zero cost. It usually routes your order for a worse implicit price through internalisation or payment-for-order-flow, trading a visible explicit fee for a hidden implicit one. That is the clearest possible illustration of why this page and implicit costs must be read together: the commission line alone tells you nothing about the all-in cost.

Exchange / ECN fees and the maker-taker schedule

Venues charge an access fee to take liquidity and usually pay a rebate to post it: the maker-taker model. The same trade costs a few mils (thousandths of a cent) per share if you take, but earns a rebate if you make. This single schedule drives venue choice and the entire passive-versus-aggressive decision.

The intuition: the exchange is a two-sided marketplace. It subsidises the side that supplies liquidity (makers) by charging the side that consumes it (takers), keeping a small margin for itself. Hence a negative explicit cost is possible: you can be paid to trade.

This schedule is the explicit-cost engine behind two whole strategies: rebate capture, where you trade for the rebate, and the maker/taker choice inside every market-making book. The crypto analogue is identical in spirit: published maker/taker basis-point tiers that fall with 30-day volume, with many venues offering negative maker fees at the top tiers, exactly mirroring lit equities. Always re-check the venue's current schedule; these change frequently.

Clearing, settlement and regulatory fees

Beyond the venue, a central counterparty clears the trade and a depository settles it (each charging small per-trade or per-notional fees) and regulators levy their own. In US equities the seller pays the SEC Section 31 fee and members pay the FINRA Trading Activity Fee (TAF); other jurisdictions add transaction taxes. These are the most-forgotten lines, and they are never zero.

Clearing and settlement: the clearing house and central securities depository charge per-trade and per-notional fees; FCMs pass exchange clearing fees through on futures. Often netted and small per trade, but real at HFT volumes. Regulatory fees (US equities, 2026): the SEC Section 31 fee is charged on the sell side as a rate per dollar of notional, reset periodically by the SEC, so look up the current rate, never hard-code it. The FINRA TAF is a tiny per-share fee on covered sales. Both are pennies-level but mandatory and asymmetric, weighted to the sell side.

Transaction taxes are the wildcard: several jurisdictions levy a financial-transaction tax or stamp duty (UK stamp duty reserve tax on certain share purchases, FTTs in some EU states) a per-notional cost that can dwarf every other explicit line where it applies. Always check the instrument's jurisdiction before you total. The honest point is that these lines are most-forgotten precisely because they are small per trade and buried in the back-office statement, yet they are pure, unavoidable explicit cost. Total them.

Market-data fees: the fixed cost everyone under-counts

To trade well you must see the book, and exchanges sell that view. Market-data fees (for the consolidated tape, for each venue's proprietary depth-of-book feed, plus per-user, per-device and non-display licences) are a substantial fixed monthly cost that does not scale with trades. Amortised across your flow, they are a real per-trade cost the trade ticket never shows.

The intuition: data is the venue's other product. Top-of-book may be cheap or bundled; the deep, low-latency proprietary feeds an HFT actually needs (full depth, order-by-order) are licensed separately and priced for professionals, with extra fees for redistribution and non-display (algorithmic) use. Because the bill is fixed, it sets a minimum viable volume: below some trade count the data bill alone makes the strategy uneconomic. That is a quiet but real barrier to entry.

This is where the explicit-cost pages meet the data pages: the engineering of the feed lives there, the fee for it lives here. The crypto contrast is sharp: venue data is largely free over public APIs, which is one structural reason the barrier to entry is lower there. The fixed-cost wall that protects incumbent equity HFTs is mostly absent.

Worked example

A single round trip (buy then sell) of 10,000 shares of a \$50 US equity (notional \$500,000 each way) as a professional DMA participant, illustrative and as of 2026. Re-check every rate against the current schedule before you rely on it.

Take the order as 5,000 shares lifted aggressively (taker) and 15,000 shares posted passively (maker) across the round trip. Commission is \$0.0002/share on all 20,000 shares; the take fee is \$0.0030/share, the maker rebate \$0.0025/share.

The net per-trade explicit cost is −\$8.40 before data: the rebates more than paid for the entire explicit stack. That is the whole point: explicit cost can be negative, which is why venue and maker-taker choice is a profit lever, not just a line to minimise.

As basis points of the \$500,000 notional, the per-trip explicit cost (excluding the fixed data bill) is well under 0.2 bps: tiny next to the spread and impact you will pay on the same trade, modelled in implicit costs. That is the recurring lesson of these cost pages: the easy, knowable half is small; the hard, estimated half is where the strategy lives or dies. Numbers are synthetic and rounded; rebate tiers, the Section 31 rate and the TAF are reset periodically, so always read the current published schedule.