What a trade costs

After this lesson you can compute the full cost of a round trip before taking it, and you'll know how frequency multiplies that cost.

Counting the tolls

Nearly every trade pays some of four tolls, and most beginners see only one of them.

The spread (course 2). You buy from the ask and sell to the bid, so a full round trip pays the full spread. A $40 stock with a 2-cent spread: 2 cents on $40 is 0.05% of the position, or 5 basis points, which is the unit costs are counted in (one basis point = 0.01%). It's worth learning to think in basis points, because it makes small numbers comparable.

Slippage. The price moves between your decision and your fill, and market orders of any size walk the order book (course 2 showed the mechanism). A limit entry in a liquid name pays almost nothing here. A market order thrown into a fast moment can pay a great deal, and there are real numbers on that below.

Commission and fees. Brokers charge per order or per share, or they advertise "zero," in which case course 10 has a note on how the broker gets paid anyway. For a European buying US stocks there's a toll that routinely dwarfs the commission: currency conversion, often 0.1–0.25% at retail tiers. You avoid most of it by converting once and keeping the dollars, then trading out of that balance. Convert around every trade and you pay it every trade.

Taxes. Entirely country-dependent. Some markets tax transactions themselves; the UK stamps 0.5% on purchases. Your own gains-versus-wealth regime is your accountant's department. What this course asks is that you know your regime before you count on any edge, since an edge smaller than the tax on it isn't worth running.

Real numbers, measured

This is why the course keeps asking for measured numbers rather than estimates. On our own live trading desk, in liquid US stocks, a careful round trip with limit entries and orderly exits measured out at about 6 basis points of position value. The same desk measured what lazy exits cost, and market orders dropped into the middle of the session leaked 35 to 47 basis points on the exit alone, in the same stocks and the same account. Six times the cost and more, and the only thing that differed was order discipline.

Then frequency multiplies whatever number you land on. At 6bp per round trip, three trades a week is roughly 150 round trips a year — about 9% of position value in annual cost, already a real hurdle for any edge to clear. The same year at 40bp costs 60%, which no edge survives. And a day-trader doing five round trips a day at even 10bp is paying around 125% of position value per year, before a single profit is counted. The per-trade number is small, and the per-year number is that small number times however often you choose to trade.

We didn't trust our own 6 basis points until we'd checked it one fill at a time. The desk's first day of live trading was reconciled afterwards, order by order, against what the system believed it had received, and the US fills matched to the cent. The same reconciliation turned up 63 European trades that had never happened at all, where the system had recorded fills that no exchange ever made. The cost number came through that audit intact. What changed afterwards was the trading plan, because seeing 6bp written next to a realistic count of trades per year moved us from taking the setups that appeared to taking the setups worth paying for.

Check yourself

1. A $30 stock with a 3-cent spread, no commission, no slippage. Round-trip cost in basis points? (3¢ / $30 = 0.1% = 10bp.)
2. Which toll do Europe-based traders most often miss? (Currency conversion. Convert once and stay in dollars, or the toll repeats every trade.)
3. Why does cost discipline matter more the more often you trade? (Costs are paid per round trip, so they scale linearly with frequency. Edge doesn't grow just because you trade more.)

The idea this lesson installs

Count the round trip's cost before you count any profit.

Next: Course 10 — "Order types: what happens when you press the button."