Automated Trading: What It Is and How to Automate a Trade Safely
Automated trading means software places or manages trades for you from rules you set in advance, such as a price to buy at or a loss to stop at. The rules can be simple orders, a bot, a copy of another wallet, or an AI agent. Automation removes hesitation, not risk: a bad rule runs just as fast as a good one.
What automated trading is
Automated trading is the habit of letting software carry out a decision you already made. You choose the token, the price, the size, and the exit. The software waits, then sends the order when the rule is true. You are not in the click at that moment. That is the whole idea, and it is also the whole risk. A rule you would refuse by hand will still fire if you left it armed.
The Wikipedia page on automated trading systems is a background definition of this family of tools. Use it as a map of the idea, not as a claim that any system makes money. This page stays on the shapes people actually automate, what a fill costs, and how to arm one rule without handing the account to a slogan.
People mix this up with algorithmic trading. Algorithmic trading is the narrower case: a precise rule written as code, sometimes a statistical model. Automated trading includes that, and it also includes a plain limit order you typed yourself. If the software can send the order while you are away, it is automated. If a person still has to press the button every time, it is not.
Automation removes hesitation. It does not remove a bad price, a thin market, or a fee. The CFTC customer advisory SI Won't Turn Trading Bots into Money Machines says SI technology can't predict the future or sudden market changes, and that promises of unreasonably high or guaranteed returns are a red flag. No AI can predict markets. A bot that runs is not evidence that the rule is sound. Read does AI trading work before you treat a screenshot as a plan.
The 5 ways people automate trades
1. Orders you arm in advance. A limit buys only at your price. A take-profit sells into strength. A stop-loss sells into weakness. A DCA order adds on a schedule or at steps you set. You already decided. The venue just waits. This is the safest shape because the price and the size are visible before anything fires. On Moby, You can keep multiple resting orders. Limit, take-profit, stop-loss and DCA fills are 0.75%.
2. A bot that follows a fixed rule. The rule might be a moving-average cross, a percent move, or a clock. A crypto trading bot is this shape when it is not an agent: software, a condition, an order. An AI trading bot is the same shape with a model in the loop. The model does not change the duty. You still own the rule, the size, and the loss. Signals are a common input. AI trading signals explains what a signal is and what it is not. A signal is not an order until something sends one.
3. A grid. A grid places buys below a price and sells above it, hoping the market chops inside a band. It looks calm in a range and ugly in a trend, because the buys keep filling as price falls and the sells never come. The grid trading agent page is the Moby write-up of that shape. Treat it as a description of the rule, not as a promise that a range will continue.
4. Copying another wallet. Software repeats trades from a wallet you follow. You did not pick the token in that moment. You picked a person, or a strategy page, and accepted their timing. That is still automation. On Moby, fees are 2% ($1 minimum) on copy buys. Read copy trading before you turn it on. A copied loss hits your wallet at your size. Their win rate is not your win rate if your size, your fee, and your exit differ.
5. An AI agent. You describe a task in words. The agent can pull data, draft a rule, and, if you allow it, send an order inside caps you set. Agentic trading is the setup, and how to set up agentic trading walks the connection. The agent is still software running your instructions. Turn on Ask before every trade and your AI asks you first. Moby does not hold or approve trades. One agent per user, with its own wallet. Agents trade Solana tokens only. The agent cannot withdraw, transfer or send funds. A DCA-style plan you want the agent to watch is on the DCA agent page. Moby does not run a schedule for your agent; the schedule lives in your AI client. Your resting orders (limit, take profit, stop loss, trailing stop, DCA) keep working on Moby even when your app is closed. The schedule is yours. The resting order, once placed, does not need the chat window to stay open.
If you only want ideas, stay on the agent strategies list and do not arm an order yet. Stacking a copy, a stop, and an agent does not cancel fees. Each fill pays.
Where automated trading goes wrong
The first failure is a rule that was never specific. "Buy strength" is not a rule. A rule names a token, a price or a condition, a size, and an exit. If any of those four is missing, the software will invent one or it will wait forever. Inventing is worse. You will not recognize the fill, and you will not know which part to turn off.
The second failure is frequency. A rule that is slightly positive before fees can be negative after fees if it trades all day. Small accounts feel this first, because a minimum fee is a large percent of a small order. The worked example below is that lesson on a $50 ticket. Run your own size through the trading fee calculator before you raise the count of trades.
The third failure is an exit you did not write. Entries are exciting. Exits are the trade. A take-profit that is a wish, and a stop that you plan to "feel," is a manual trade wearing an automated costume. The risk reward calculator is there so you can see the gross win against the gross loss before fees change both numbers.
A backtest that never paid a fee is not evidence. Markets gap, and a stop is an instruction, not a guaranteed price. The next failure is letting the tool raise its own limits. You set your agent's per-trade and daily spending limits and market-cap range; Moby blocks red-rated tokens. Limits are only as tight as you set them. Sells and exits skip the market-cap and safety checks, and tokens with no safety rating are not blocked. If you set the caps wide because you wanted more action, the block on red-rated tokens will not save a bad size. Rate limits protect against runaway loops and scale with the fees your agent pays; you set the spend limits.
Worked example: a $50 rule, two ways to fill it
Assumed numbers for the math, not a quote and not a price call. Example, not a recommendation. The position is $50 of a token. The rule takes profit if the position is worth $60, which is 20% higher, and stops if the position is worth $45, which is 10% lower. No other prices are used. Onchain fees sit on top of every path below and are not in these dollar figures.
Route A uses a limit to enter and a resting exit. The entry fee is $0.375. The take-profit fee, charged on the $60 position, is $0.45. The stop-loss fee, charged on the $45 position, is $0.3375. A winning round trip costs $0.825, which is 1.65% of $50. A losing round trip costs $0.7125, which is 1.43% of $50. The gross win is $10. After the $0.825 fee the net is +$9.18. The gross loss is $5. After the $0.7125 fee the net is -$5.71.
Route B is two market swaps. The fee is $0.90 in and $0.90 out, so $1.80, which is 3.6% of $50. The same $10 gross win nets +$8.20. The same $5 gross loss nets -$6.80. Before onchain fees, the resting path keeps more of the win (+$9.18 versus +$8.20) and loses less on the stop (-$5.71 versus -$6.80).
The numbers show the fee if those prices happen. They do not say the position will get there. Swaps and agent market trades pay 0.75% with a $0.90 minimum per swap. Limit, take-profit, stop-loss and trailing orders pay a flat 0.75% with no minimum. Onchain fees apply too. A 0.75% swap fee (minimum $0.90 per swap), plus onchain fees. Fees on your agent's trades earn up to 10% cashback in USDC. It is not a discount on the swap fee.
| Path | Win, fees | Net on a $10 gross win | Loss, fees | Net on a $5 gross loss |
|---|---|---|---|---|
| Route A, limit and resting exit | $0.825 (1.65%) | +$9.18 | $0.7125 (1.43%) | -$5.71 |
| Route B, two market swaps | $1.80 (3.6%) | +$8.20 | $1.80 (3.6%) | -$6.80 |
How to automate one rule safely
Five steps. Do them in order. Do not skip to a live size because a demo looked smooth.
- Write the rule in one sentence that names the token, the entry, the take-profit, and the stop-loss.
- Set a per-trade cap and a daily spending cap you can afford to lose before any order is armed.
- Place the entry as a limit and attach the take-profit and the stop-loss as resting orders.
- Turn on Ask before every trade so your AI asks you first, and keep that on until the log is boring.
- Run one small test, read the fill and the fee, and stop the rule if you cannot explain the result.
If the sentence needs a paragraph of exceptions, you do not have a rule yet. A cap you raise after two wins is not a cap. Step 5 is the evidence that matters: one fill, the fee you paid, and a sentence you could say out loud.
What Moby does with an automated rule
Moby is the wallet, the order types, and the limits. It is not a machine that finds a winning rule for you. One agent per user, with its own wallet. Agents trade Solana tokens only. The agent cannot withdraw, transfer or send funds. AI Trading and copy trading currently need an access pass in the app. Supported hosts: Codex, Claude Code (CLI), Gemini CLI (with a paid Gemini API key or enterprise access) and Grok CLI. ChatGPT: works on Business, Enterprise and Edu workspaces that allow custom MCP servers (web). Download Moby if you want the app. You set your agent's per-trade and daily spending limits and market-cap range; Moby blocks red-rated tokens. Limits are only as tight as you set them. Sells and exits skip the market-cap and safety checks, and tokens with no safety rating are not blocked. Use a resting order when you already know the price. Do not chase a promise of steady gains.
What to write down after the first fills
Keep a short note: the rule in one sentence, the round trips, the fees in dollars, and the result if you had done nothing. If doing nothing won, the software still ran. The rule did not.
Frequently asked questions
What is automated trading?
Automated trading means software places or manages trades from rules you set in advance. The rule might be a limit price, a stop, a bot, a copy of a wallet, or an AI agent. It removes hesitation. It does not remove the loss if the rule is wrong.
Is automated trading the same as algorithmic trading?
Not quite. Automated trading is the wider idea, including a simple limit order. Algorithmic trading usually means a precise rule written as code, sometimes a quant model. A limit is automated. A coded moving-average cross is algorithmic. The algorithmic trading page covers that narrower meaning.
How much does an automated trade cost on a small ticket?
On a $50 limit with a resting exit, the worked example costs $0.825 to win and $0.7125 to lose, or 1.65% and 1.43%. Two market swaps on the same $50 cost $1.80, or 3.6%. Onchain fees apply too. Larger tickets move toward 0.75% each way.
Can I automate a trade without a bot?
Yes. A limit to enter and a take-profit or stop-loss to exit is already automation. You can keep multiple resting orders, and they keep working when the app is closed. A bot or an agent is for a rule that still needs a fresh decision.
Is copy trading a form of automated trading?
Yes. Software repeats trades you did not click one by one. On Moby, fees are 2% ($1 minimum) on copy buys. A copied loss still hits your wallet. AI Trading and copy trading currently need an access pass in the app.
What is the main risk of automated trading?
Speed. A bad rule runs just as fast as a good one, and fees repeat on every fill. Limits are only as tight as you set them. Sells and exits skip the market-cap and safety checks, and tokens with no safety rating are not blocked.
Can an AI agent automate trades for me?
It can send orders inside limits you set, and it can ask you first. Turn on Ask before every trade and your AI asks you first. Moby does not hold or approve trades. No AI can predict markets. The agent cannot withdraw, transfer or send funds.
How should I test an automated rule?
Use a size you can lose, include fees, and read every fill. Compare the result with doing nothing. If you cannot explain the order without the model, stop. A backtest that ignores fees is not a test. One clean fill beats a busy week you cannot describe.