Module 4 of 14 · Foundations

Costs and the speed limit

There are only two ways to improve a trading strategy: make more before costs, or pay less in costs. The second is the only one you control with certainty, and it is the one most automated systems get wrong.

Every trade pays a spread cost (crossing half the bid–ask spread) plus commission. In dollars those numbers are meaningless across markets — $5 is nothing on a full-size crude contract and a lot on a micro. Divide the dollar cost by the annual dollar risk of one contract and you get the risk-adjusted cost per trade, measured in Sharpe-ratio units. Now every market is on the same scale.

Annual cost is then simple: cost per trade × turnover (how many times a year you trade your average position), plus two trades for every roll. Faster strategies have higher turnover, so they pay more for every unit of risk.

Finally set a budget. A realistic single-market Sharpe ratio before costs is modest — a few tenths — so a sensible rule is to spend no more than about a third of it on costs. Budget ÷ cost per trade gives the maximum turnover a market can afford: its speed limit. Markets whose cost per trade is so high that even slow trading breaks the budget should not be traded at all.

🎛 Cost budget calculator

1
20
0.3
33%
4

Cost per trade

$1.60

spread $1.00 + commission

Risk-adjusted cost

0.00071

Sharpe units / trade

Annual cost

0.020

budget 0.099 SR

Max affordable turnover

130

the speed limit

Within budget: costs use 7% of the expected edge.

Cost per trade = half the bid–ask spread (in ticks × tick value) + commission. Dividing by one contract's annual risk turns dollars into Sharpe-ratio units, which makes cheap and expensive markets directly comparable regardless of contract size. Turnover counts how many times a year you trade your average position; rolling adds two trades per roll. Spreads and commissions here are illustrative.

Why speed is a cost decision first

Turn the turnover slider up to a few hundred — the turnover of a rule that holds positions for a day or two. Even the cheapest liquid contracts blow through the budget. An intraday system on minute bars has turnover in the thousands: it needs a pre-cost edge many times larger than anything a directional futures strategy can realistically deliver.

The speed limit runs the other way too: it tells you which markets each rule may trade. A medium-speed rule might suit equity index micros but be unaffordable in a thin agricultural contract. Choose rules per market, not one rule for all.

Things to try

  • • Compare MCL and NG: the dollar cost per trade differs by about seven times, yet the risk-adjusted cost is almost identical. Dollar costs mislead; risk units don't.
  • • Widen the half-spread to 3 ticks to mimic trading in the overnight session or right at the open. Watch the speed limit collapse.
  • • Set rolls to 12 (a monthly contract): rolling alone uses a visible slice of the budget before you have placed a single signal trade.

Original interactive lesson. Labs run on a seeded synthetic market in your browser, built to illustrate mechanics, not to predict any real market. Contract figures are illustrative; check exchange specifications. Educational analysis, not investment advice.

How are we doing?