Document 4 of 7Last reviewed 14 August 2026

Risk Notice

Putting a program in charge of the orders changes nothing about how much a leveraged market is able to take from an account. This document opens with a checklist, because a checklist is what actually protects people; answers it with what we are and are not; walks the three places an automated account comes apart; and ends on the arithmetic of what that costs. Anyone who finds a line here unacceptable should close the tab rather than buy.

1.Before a live account is funded

1.1
Seven things to have done. They are in the order they are worth doing, and none of them is optional.
1. Get it confirmed by your broker, in writing if they will put it in writing, that an expert advisor may trade your account type — and satisfy yourself that the law where you live allows it too.
2. Settle on the figure you could lose outright without changing anything about your year, and fund the account with that figure and no more.
3. Read the flight manual from cover to cover, input table included, before the engine goes anywhere near a chart.
4. Run it on demo, at the broker you will actually trade with, until a flat week and a losing one have both gone by.
5. Write down the drawdown at which you would switch it off, then be honest with yourself about whether you would.
6. Sort out where the terminal is going to live: a machine that never sleeps, a connection that holds, an update schedule you control rather than one that surprises you.
7. Accept in advance that the account will spend part of the year below its own high-water mark, because every account that trades does.
Anyone unwilling to work through that list should not be running automated trading software — not ours, and not anybody else's.

2.We are not in the room

2.1
Planet Algo writes software and sells licences to it. We hold none of your money, take none of your decisions and have no way into anybody's account.
2.2
Every order leaves under credentials only you possess, on an account only you can open. Nobody here is watching it, nobody here could step in if they wanted to, and the outcome is yours from first tick to last.
2.3
The only thing sold here is a licence to a compiled program, and no regulated permission of any kind stands behind that — none to advise, none to hold money, none to place an order on somebody else's behalf. So nothing printed on this site is financial, investment, legal or tax advice, and no instrument named on it is a recommendation to trade it. Guidance shaped to your own circumstances is something to buy from a person authorised to give it, and to read next to what About Our Figures and the Purchase Terms add to it, because buying an engine accepts those two alongside this one.

3.Where the strategy can fail

3.1
Every rule in an engine was drawn from markets that have already happened. A regime the engine has never met can pull behaviour out of it that nobody designed and nobody predicted.
3.2
Compiled software also carries defects that testing has not yet reached. We find some of them; some of them find you first.
3.3
Repetition is the trait that separates a program from a person. A trader who is wrong tends to hesitate. An engine that is wrong will place the same order again, and again, until something outside it intervenes — and that something is usually the owner noticing.
3.4
A published history describes what already happened on one account. There is no arithmetic that turns it into what happens next on yours.

4.Where the wiring can fail

4.1

The engine is only ever as reliable as the machine holding it up. Five ordinary failures and what each one costs:

What breaks	What it costs you
The machine sleeps, or reboots itself for an update.	The terminal is not there at the minute an exit was due.
The line to the broker drops.	An order already sent goes unacknowledged; one not yet sent never leaves.
The broker's own platform goes down.	Nothing opens and nothing closes for as long as it lasts.
The disk runs out of room.	The terminal stops writing, and an engine that cannot log is an engine nobody should trust.
The server clock drifts.	Session filters open and shut an hour away from where they should.

Any one of them can swallow the single order that mattered most that week.

4.2
Then there is the price itself. A gap at the open, a thin book, a spread that widens on a headline, a requote, a fill several points away from the screen: each of them can turn the loss a stop was meant to cap into a bigger one.

5.Where the owner can fail

5.1
Your trading costs are not ours. Spread, commission, overnight financing and fill quality are set by your broker and land squarely in your result, and an edge that survives those costs at one broker may not survive them at another.
5.2
An input pushed outside the range the flight manual prints turns the engine into something we have never run and cannot comment on. Whatever it does from that point, it does beyond our knowledge of it.
5.3
The expensive button
The costliest control on the platform is the one that shuts the engine down halfway through a drawdown. Pressing it turns a number on a screen into a number in the account, and it gives away whatever recovery the strategy still had in it.
More owners are hurt by that single decision than by anything the software has ever done on its own.

6.The size of what can be lost

6.1
Leverage cuts in both directions. A run of bad weeks can take a slice of a deposit, the whole of a deposit, and at brokers without negative-balance protection more than was ever deposited. Money that has a job elsewhere has no business in a trading account.
6.2
The engines are built to fly at 1:30 or under. Treat that as a speed limit rather than a seat belt: a bad run inside the limit is still entirely capable of taking a serious piece out of a healthy account.
6.3
There is no drawdown figure on this site to weigh, because no public account page exists for either engine yet. That is a gap in the evidence, not a reason for comfort: every trading account spends time under water, and ours will be no exception.