Blog
How Taxes and Fees Compound Against Active Traders
On a single trade each of these looks ignorable. Across several hundred trades, with tax on every gain, a strategy that wins on paper settles somewhere below zero in the account.
Gross numbers run through every backtest. They run through most journal reviews and very nearly all trading conversation besides. The account uses net ones. Between the two sits a collection of small costs, each of them individually ignorable, which together settle the question of whether a strategy pays.
The costs, one trade at a time
On any household spreadsheet, the line that surprises people is never the mortgage. It is the small recurring things nobody thinks to cancel. A trading account behaves in exactly the same way.
Nobody reads a costs article for pleasure, so this will be quick, and then you can go and look at your own statement, which is the genuinely uncomfortable part.
The spread comes first. You buy at the ask and sell at the bid, so the gap is paid twice on every round trip. On a thin name, several percent. A liquid large cap charges a few hundredths of one, which is why the same method costs different amounts in different instruments.
Slippage comes next. It widens in fast markets. It widens again with larger orders. All of it is the gap between the quote on the screen and the fill that actually came back.
Commissions and fees. Largely zero on US stocks now. They are not zero on options, nor in a good many international markets, and the regulatory and exchange fees keep appearing on statements regardless of what the headline rate says.
Tax last. In most jurisdictions a gain on a position held for less than a year is taxed as income, at a higher rate than long-term gains, and the active trader pays that rate on every winning trade, every year, with nothing deferred to a more convenient moment.
What multiplication does to them
Start with a genuine gross edge. 0.3R per trade, risking 1% of the account on each. A respectable figure by any standard. That works out at roughly 0.3% of the account per trade before costs.
Now put costs against it. Round-trip costs of 0.1% of position value on a liquid stock, with positions averaging about ten times the amount risked, which is what a stop roughly 10% away tends to produce in practice. The cost works out at about 1% of the position. That is 0.1R per trade, so the edge is now 0.2R.
Trade it two hundred times in a year and the gross return sits in the region of 60% of the account, of which costs have already taken twenty percentage points. Tax what remains at a short-term rate. The net figure lands somewhere near half the gross one. Push the count to five hundred trades a year, or move into less liquid names where a round trip costs 0.35%, and the same strategy loses money, steadily, for as long as you keep running it.
| Trades a year | Costs at 0.1% of position | Costs at 0.35% |
|---|---|---|
| 20 | 2% of account | 7% |
| 100 | 10% | 35% |
| 500 | 50% | 175% |
That last cell is correct. A strategy showing 40% a year on paper, at five hundred trades, is a losing strategy once the costs are modelled honestly, and nothing in the equity curve will say so.
Two things follow from the arithmetic. Costs scale with activity while tax scales with gains, which means the more you trade and the better you do, the larger the share of your gross result that ends up belonging to somebody else. The second is that an edge has to be judged after both have been taken out, and most edges that look thin before costs turn out to be absent afterwards.
An illustration
Suppose a short-term mid-cap method looks profitable in the journal. The journal is kept at mid-prices. Rebuild that year using the bid and the ask actually quoted at each fill, add back the tax already paid, and a healthy positive falling to roughly nothing would surprise nobody. Now take the same rules, restrict them to the most liquid names on the list, and run them with a holding period that halves the trade count. That version may well come out ahead on both counts. The trader in the first version worked twice as hard. All the difference went to market makers and the tax authority.
Keeping costs from eating the edge
- Model costs at the high end of what you observe, because if the edge disappears when you double your cost estimate it was never robust to begin with.
- Trade the most liquid instruments available. The spread is the one cost you get to choose.
- Count trades. Where two methods show similar gross expectancy, the one with fewer trades nets more, and it is generally the one executed better as well, because fewer decisions leaves fewer chances to decide badly.
- Each quarter, set the account’s net line beside the journal’s gross line. That gap is your activity cost. It ought to be a number you can say out loud.
- Hold what the analysis allows. A position carried past the short-term tax threshold keeps meaningfully more of its gains.
The trading costs guide works through each cost in detail, and the stock profit calculator computes what a trade nets after fees, so that the journal can begin from the right number.
Frequently asked questions
How much do trading costs reduce returns?
It scales with activity, which is why two traders with the same strategy can end the year in very different places. At a round-trip cost of 0.1% of position value, 100 trades a year costs roughly 10% of the account and 500 trades roughly 50%. At 0.35%, which is ordinary enough in less liquid names, 500 trades a year costs more than the account is worth, a level of expense that overwhelms any gross edge.
Are short-term trading profits taxed differently?
In most jurisdictions, gains on positions held under a year are taxed as income, at a higher rate than long-term gains, and an active trader pays that rate on every winning trade every year with no deferral at all. Net of tax, a gross return is often somewhere near half the figure sitting in the journal, which is a gap worth knowing about before you draw conclusions from it.