Set It and (Mostly) Forget It: What Automation Does for You

“Set it and forget it” is a lovely phrase. It sells rotisserie ovens, and it sells trading automation too. It’s also about 80% true, and the missing 20% is where people get hurt.

So let’s be precise about what an automation actually does for you, what it doesn’t, and how much “forgetting” is healthy.

What automation is brilliant at

Following your rules exactly, every time. You said “sell if price drops 3% from entry”. The automation sells at 3%. Not at 3.4% because it was hoping for a bounce, not at 5% because it was asleep, not at 1% because it got nervous. Just 3%.

Watching when you can’t. Crypto never closes. An automation checks your pairs on the schedule you picked, every 15 minutes or every 4 hours or whatever suits your strategy, through nights, weekends and that week you finally took a real holiday.

Staying calm. It doesn’t revenge trade after a loss. It doesn’t double down because a stranger on the internet said “this is the bottom”. It has no feelings, which in trading is a genuine competitive advantage.

Keeping score. Every trade, every result, every reason it did or didn’t enter. You get an honest record, which is more than most of us keep when trading by hand.

What automation won’t do

It won’t make a bad strategy good. An automation runs your rules faster and more consistently. If the rules lose money, it will lose money faster and more consistently. This is why we keep saying test before you trust.

It won’t notice the world changed. Your strategy might love calm, trending markets. When the market turns choppy and sideways, the automation will keep doing exactly what you told it, because that’s its whole job. Noticing that the job no longer fits is your job.

It won’t pick its own budget. You decide how much each automation can use. Give it a sensible amount and it can never lose more than that. Give it everything and, well, it will treat everything as its budget.

The healthy amount of “forgetting”

Here’s a routine that works for most people:

• Daily, 30 seconds: glance at your notifications. Anything unusual? A trade that closed on a capital stop, or an automation that stopped itself? Worth a look. Otherwise, carry on with your day.
• Weekly, 10 minutes: open the results for each automation. Is it behaving the way the backtest suggested? Roughly the same win rate, similar size of wins and losses?
• Monthly, 30 minutes: ask the uncomfortable question. Would you still start this automation today, knowing what you know now? If not, pause it. Pausing is free.

That’s it. Compared to staring at charts for hours, it’s a very good deal. It’s also much less than “never think about it again”, which is the version of the phrase that costs money.

Signs you should step in

• It’s losing more often than your testing suggested, for several weeks in a row.
• It stopped itself because its budget was used up.
• The market is doing something your strategy was never built for, like a sudden crash or a news-driven spike.
• You changed your mind about the idea behind it. That’s allowed. Nobody gets graded on consistency with a decision made in March.

Automation is a bit like a very reliable colleague who does exactly what the documentation says. Wonderful, until the documentation is wrong. Keep the documentation in good shape and let it get on with the work.

If you haven’t built your first one yet, start with a paper automation. It’s the least stressful way to see all of this for yourself.