Ready to go further
Next: count what each trade really costs
The break-even calculator shows how much each trade must make before costs stop eating it.
What is included
- What it is, in plain words
- The formula or mechanism, with exact arithmetic
- How it hits grid, martingale and DCA bots
- A check you can run before you switch a bot on
Who it is for
- People who run or plan to run a trading bot
- Anyone reading a backtest or a bot sales page
- Readers who want the numbers, not the pitch
- Not advice, and not a signal service
Key facts
- High frequency trading (HFT) means placing very large numbers of orders at extremely high speed, aiming to earn a tiny edge on each trade. It is a contest of speed, cost and scale more than of prediction.
- Delay matters because prices keep changing. If the best price changes once every 5 ms (made-up), an order that takes 100 ms to arrive meets a price that has changed 20 times since the bot looked (calculated by us on 29 Oct 2026, exact).
- Light in optical fibre travels about 200 km in a millisecond, so distance to the exchange is a real cost. A delay of 1 ms is roughly 200 km of fibre one way.
- A fee that looks tiny is large against a tiny edge. A fee of 0.004 a share takes 40% of an edge of 0.01 a share, and 167 trades of 100 shares are needed to make $100 instead of 100 trades.
- A retail bot on a home line or a cloud server cannot win this contest, and the sensible response is a slower strategy where speed is not the edge.
What high frequency trading is
There is no single strict definition, but the pattern is consistent. The firm holds positions for a very short time, often well under a second, trades in very large numbers, and makes a very small amount on each trade. Common ideas include quoting both sides of a market and earning the gap between them, or acting on a price difference between two venues before it closes. The returns come from volume, so the firm needs low costs, fast connections, and close access to the exchange's computers.
Nothing in this is about guessing where a price will be in an hour. It is about being first to a small, short-lived opportunity, which is why delay is the main input. Most retail trading bots work on different time scales and have different edges, and they should be judged on those.
Delay, in numbers
Take a made-up market where the best price changes once every 5 ms. A bot that sees a price and sends an order is acting on information that gets older every millisecond. The table shows how much older, for several delays, and how far light in fibre can travel in that time.
| Delay from price to order | Typical setup (made-up) | Price changes missed | Fibre distance, one way (km) |
|---|---|---|---|
| 0.1 ms | Co-located firm | 0.02 | 20 |
| 1 ms | Nearby server | 0.20 | 200 |
| 10 ms | Cloud server, same region | 2 | 2,000 |
| 100 ms | Home line plus a broker API | 20 | 20,000 |
| 500 ms | Slow API or busy server | 100 | 100,000 |
At 0.1 ms the bot is ahead of nearly every change. At 100 ms it is 20 changes behind. Orders that arrive after the price has moved either miss the trade or fill at a worse price. That is the same effect as slippage, and it grows with the delay. The fibre column is why firms pay to place their servers inside or next to the exchange's data centre: a distance of 2,000 km costs 10 ms each way, and nothing in the code can recover it.
Edge and fees
Suppose the edge on each trade is 0.01 a share, on 100 shares, so $1.00 a trade. To make $100 in a day takes 100 winning trades. If a fee of 0.004 a share applies, the net edge is $0.60 a trade, and 167 trades are needed. The edge and the fee are made up, but the lesson is general: when the edge is a fraction of a cent, the cost schedule is part of the strategy, and a firm with lower fees has a different business from one with higher fees. See spread and algorithmic trading for how costs eat into frequent trading.
Net edge per trade = (edge per share - fee per share) x shares
Trades needed = daily goal / net edge per trade
What this means for a retail bot
- Do not build a bot that needs to be first. If the strategy only works when the order arrives within a few milliseconds, a home or cloud setup is the wrong place to run it.
- Measure your real delay. Time from the price to the confirmed fill, on your own setup, over many trades. A backtest that assumes instant fills is too kind, as covered in backtesting.
- Prefer edges that last longer than your delay. A rule that works on prices held for minutes or hours is not hurt by 100 ms. A rule that works on prices held for a few milliseconds is.
- Count every cost. Spread, fees and slippage are paid on every trade, and frequent trading pays them most. See the break-even calculator.
- Be wary of sales claims. A product that promises high frequency results on a home connection is promising something the arithmetic above does not support. Ask for audited fills.
How to check a fast bot
Log the time the bot saw the price, the time the order was sent and the time it filled, and compare the fill price with the price the bot saw. Run it in paper trading first, then with small size. Then ask how much of the edge remains after the average delay and the fee, and use the expected value calculator on the result.
What this page does not tell you
It does not describe any firm, market or rule, and it does not say how much any high frequency trader makes. The price changes, edges and fees are made up, the only physical figure is the speed of light in fibre, and this is education, not advice.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What is high frequency trading?
Trading in very large numbers, at very high speed, to earn a tiny edge on each trade. It depends on low delay, low costs and scale.
Why does speed matter so much in HFT?
Prices change constantly, and an order that arrives late meets an older price. When the edge is tiny, being first is most of the edge.
Can a retail trader do high frequency trading?
Not in the way professional firms do. A home or cloud setup has a delay of tens or hundreds of milliseconds, which is too slow for edges that last a few milliseconds.
Is HFT the same as algorithmic trading?
No. Algorithmic trading means using rules run by a program, at any speed. HFT is a fast, high volume form of it.
What should a retail bot do instead?
Use an edge that lasts longer than its delay, count all costs, and test with realistic fills. Speed is not the retail advantage.
Are the numbers on this page real?
Example numbers are made up to show how the maths works, and the page says so. A figure only counts as exact when the page says it was calculated on a stated date. Nothing you enter in a calculator leaves your browser.
Can I use this to decide whether to run a bot?
It can show whether a bot's numbers hold together and how much an account can lose. It cannot say what the market will do, and it is not advice. Use it to ask better questions, then read the terms of your own broker or exchange.
Is this financial advice?
No. It is general education. The prices, edges and fees are made up and it does not consider your situation.
Know the Risk Before You Run It
Free calculators and guides, added every week. Leave your details and we will email you when new ones go up.
Trading carries a risk of loss