Automatic trading is what happens when a strategy places its own orders: the rules live in code, and when the entry condition is met the order goes to your broker without you touching anything. No hesitation, no missed alert, no talking yourself out of the setup. That is the appeal. It is also why automatic trading can empty an account faster than manual trading does — it removes the last human check between a bad idea and a filled order.
What automatic trading actually is
An automatic trading system is four links in a chain. Every link can break, and each has to be tested on its own.
Signal → alert → order routing → broker
- Signal. A coded strategy evaluates each bar as it closes and fires when its entry logic is true — a break of the opening range, or a 2-period RSI below 10 above the 200-day.
- Alert. The signal becomes an alert carrying a machine-readable payload — symbol, side, quantity, order type, stop and target. On TradingView, a webhook.
- Order routing. An execution service receives the webhook, checks it against your rules — maximum size, allowed symbols, trading hours — and turns it into a broker order.
- Broker. Your broker fills it. The fill price is the only number that ever pays you, and it is rarely the price on the chart.
How automatic trading differs from manual trading
Manual trading fails on discipline: you skip the trade after two losers, you widen the stop, you take profit early. Automated trading has none of those failure modes — and that is the problem. It replaces them with engineering ones: a missed webhook, a stale chart, a rounding error in size, a duplicate alert that opens two positions instead of one. Manual traders lose to psychology; automatic traders lose to plumbing.
The Northtape automatic trading chain, in plain terms
The base Northtape desk is a research and signal product: it surfaces the candidates and the proof, and never places an order for you. Automation is a separate, optional layer — the Loomi Auto add-on, a flat +$20/mo on top of Northtape ($8/mo at the founding rate for six months, then $20/mo). Link by link, from the pricing page:
- The strategies land on your own TradingView. Every Northtape strategy is granted to your TradingView account as an invite-only indicator tied to your username. You add it to a chart and see the exact entries and exits plotted live. The source stays closed: you rent the signal, not the code.
- You set the alerts. They run on your account, so the alerts are yours.
- Alerts route through an execution service. The invite-only strategies fire TradingView alerts, routed through a service such as TradersPost to your broker.
- Your broker executes. Trades then run hands-free. The add-on includes the setup walkthrough, plus new strategies as published.
Those strategies sit on the strategy index, each double-verified on TradingView then re-backtested in Python. Numbers are on the results page, the wider toolkit on the features page. Backtested figures are historical, not a forecast.
What you must have in place before you automate
This is the part most guides skip, and it decides whether hands-free trading is a tool or a liability. Every automated entry must carry a predefined stop loss and a take profit, position size must be capped in the software, and you must have a kill switch you can reach in seconds. Do not go live until all five are true.
| Control | What it means | Why it is non-negotiable |
|---|---|---|
| Stop AND target on every entry | The opening order attaches its exit stop and profit target — a bracket, sent at entry time. | A naked entry has no defined loss. If your connection drops, the bracket is the only thing still working. |
| Position size cap | A hard maximum per order, enforced by the execution service, not your intentions. | Protects you when a malformed payload asks for 100 contracts, not 1. |
| Max open positions | One position per instrument, plus a ceiling on concurrent positions. | Stops duplicate alerts stacking into a size you never chose. |
| Daily loss limit | A dollar or percentage loss that halts new entries for the day. | Bad regimes cluster. Turns a catastrophic day into a bad one. |
| Kill switch | One action that disables the automation and flattens open positions. | When something goes wrong you have seconds, not minutes. |
Why automatic trading fails for most people
Automating a strategy that was never profitable
Automation is a multiplier on whatever your strategy already is. If the edge is negative after costs, hands-free execution realises that negative edge perfectly, all day. Profitable automatic trading starts with a strategy that was profitable before anyone automated it — tested on real bars, with realistic costs, over enough trades that the result is not noise.
Over-optimisation
Any strategy can be tuned to look magnificent on the data it was tuned on. A 95% win rate across 18 trades is a curve fit, not an edge. Look for few parameters, a decent trade count, and results that survive data the strategy never saw. The Northtape rule: ship only when a manual TradingView test and an independent Python backtest agree.
No risk caps, and nobody watching
Hands-free trading is not unsupervised trading. The system does the clicking; you still own the account. Reconcile the day’s fills against the day’s signals, and watch for orders rejected, partially filled, or filled nowhere near the alert. Silence from an automated trading system is not the same as all being well.
Latency and slippage in an automatic trading system
Between the bar close that generates your signal and the fill that costs you money, delays add up: the alert must be delivered, the execution service must validate and forward it, the broker must route and fill it. On a webhook chain that is seconds, not microseconds — fine for the strategies it suits, fatal for one whose edge lives inside a one-second window.
Slippage is what that delay costs you, plus the spread. It is largest where automation is most tempting: breakouts, the open, and news. Two consequences:
- Backtest with realistic costs. A strategy whose edge disappears when you add a tick of slippage does not have an edge.
- Match the strategy to the pipe. Bar-close signals on 5-minute, hourly and daily charts tolerate a few seconds of latency. Sub-second scalping does not.
How to start automatic trading, step by step
- Start in paper. Always. Run the entire chain — strategy, alert, execution service, paper broker — before a real dollar is at risk.
- Pick one strategy, one symbol, one timeframe, so when something breaks you know what broke.
- Read the backtest honestly: trade count, profit factor, win rate, maximum drawdown, whether costs were included. Decide now the drawdown you would accept.
- Write the risk rules down first: stop distance, target distance, size per trade, maximum open positions, daily loss limit.
- Build the chain. Webhook alert, execution service configured with your caps, paper broker at the end. Confirm the payload carries side, quantity, stop and target.
- Fire test alerts by hand. Confirm the paper order arrives correctly, then send an oversized quantity on purpose and confirm your caps reject it.
- Run paper through a full cycle of signals. Reconcile every fill against every alert; zero unexplained trades is the bar.
- Go live at minimum size. Live fills behave differently from paper fills; you are testing reality, not profitability.
- Test the kill switch on day one, while the market is open.
- Scale only on evidence. If live results diverge from paper, the cause is slippage or a bug. Find it before you add risk.
Is profitable automatic trading realistic?
It is realistic in the same sense profitable manual trading is: possible, uncommon, and entirely dependent on a genuine edge and the discipline to run it unchanged. Automation buys consistency of execution — the same rules at 3am and after three losers. It does not buy you an edge, and no honest system will claim otherwise.
The sequence that works is boring: find a strategy with evidence behind it, verify it independently, define the risk before the entry, prove the plumbing in paper, then let the machine do the clicking. For the deeper build-out, read the companion piece on building an automatic trading bot.
Key takeaways
- Automatic trading is a four-link chain: signal → alert → order routing → broker. Test every link separately.
- Every automated entry must carry a predefined stop loss and a take profit, sent at entry time.
- Cap position size and open positions in the software, set a daily loss limit, and keep a tested kill switch.
- Automation multiplies whatever edge you have, including a negative one — validate the strategy first, then paper trade the whole chain until it is boring.
Get Northtape signals before anyone else
Real backtested strategies, options flow, and a hedge fund built from the greats — join the early-access waitlist.
Automatic trading FAQ
What is automatic trading?
Automatic trading is when a coded strategy places its own orders: it detects its entry condition on a chart, fires an alert, an execution service turns that alert into a real order, and your broker fills it.
Is automatic trading profitable?
Only if the strategy has a real edge before you automate it. Automation removes hesitation and missed entries, but it does not create an edge. Automating a losing system simply loses money faster and more reliably.
Do I need to know how to code to run an automatic trading system?
Not necessarily. The common no-code path is a strategy on TradingView firing alerts, an execution service such as TradersPost receiving them, and a supported broker at the end. With the Loomi Auto add-on the strategies are granted to your TradingView username as invite-only indicators.
What risk controls should every automatic trading system have?
A predefined stop loss and a take profit on every entry, a hard cap on position size and open positions, a daily loss limit, and a kill switch you can reach in seconds.
Should I paper trade before going live with automatic trading?
Yes, always. Paper trading is the only way to prove the whole chain works: alerts firing on the right bars, orders arriving with the right size and direction, stops and targets attaching. Then go live at minimum size.
Educational tools, not investment advice. Trading involves substantial risk of loss. Past backtest performance does not guarantee future results.