Ready to go further
Next: look at a record the hard way
Read how survivorship bias makes a list of shown results look better than the typical one, then ask of any strategy how the whole history looks.
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
- Cryptohopper was not run for this page and nothing here recommends it. The page describes how hosted bot services work in general. Read the vendor's current terms before relying on any of it.
- Crypto-assets can be financial products.
- As we understand it, Cryptohopper is a hosted service that runs trading bots for you and connects to crypto exchanges through API keys. The bot runs on the vendor's servers, so it can keep going when your own computer is off.
- Two layers carry risk: the service that hosts the bot, and the strategy that tells it what to do. Some services of this kind let other people publish strategies, which adds a second author to check.
- In the made-up year below, a four-month window reads +18.1%, the whole year +1.3%, and the four months after the window -10.7% (calculated by us on 3 Nov 2026, exact).
- Prices, plans, features, results and ratings are left out, because we have not checked them.
Two layers to check: the host and the strategy
A hosted bot has two parts that are easy to run together in your head. The first is the service, which keeps the bot alive and sends its orders. The second is the strategy, the set of rules that decides what to buy and sell. As we understand it, Cryptohopper provides the first. Services of this kind usually let you set up a bot from settings, templates or strategies, and some let other people publish strategies as well. That is where the second layer can belong to a stranger.
| The host (the service) | The strategy (the rules) | |
|---|---|---|
| What it is | The machinery that runs the bot and talks to your exchange. | The rules that choose each order: entries, exits, sizes, stops. |
| Who wrote it | The vendor. | You, the vendor, or another user who published it. |
| Typical failure | An outage, a bug, a breach, a key held somewhere you cannot see. | Rules that fit the past, a risky size setting, a record that shows only good stretches. |
| What you can check | Key permissions, stated terms, exchange statements against its figures. | Rules you can read, the method behind its record, how it behaved in bad months. |
What running on someone else's server changes
- It keeps working when you are away. That is the point of a hosted bot, and also the risk. Nobody is watching it either, unless you set an alert or look.
- The key lives with a third party. You keep the funds at the exchange and give the service a key that lets it place orders. If the service is compromised, the keys it holds are at risk, which is why a key should be limited to trading and, where offered, to listed IP addresses.
- You depend on the host staying up. If it pauses or fails, find out beforehand what state your open orders are left in.
- Your own computer stops being a point of failure for the bot, and the vendor's servers become one. That moves the risk rather than removing it.
Worked example: the window you see against the year that happened
All of this is made up. A strategy, published on a hypothetical service, made these monthly returns over a year. A listing page shows a four-month stretch, May to August, and nothing about the other eight. Nobody lied. A listing has to show some window, and a recent one that looks good is an easy choice.
| Month | Return | Value at month end (start 100) | In the listing window? |
|---|---|---|---|
| Jan | +1.0% | 101.0 | No |
| Feb | -2.5% | 98.5 | No |
| Mar | +0.5% | 99.0 | No |
| Apr | -3.0% | 96.0 | No |
| May | +4.0% | 99.8 | Yes |
| Jun | +5.0% | 104.8 | Yes |
| Jul | +3.5% | 108.5 | Yes |
| Aug | +4.5% | 113.4 | Yes |
| Sep | -4.0% | 108.8 | No |
| Oct | -6.0% | 102.3 | No |
| Nov | -2.0% | 100.3 | No |
| Dec | +1.0% | 101.3 | No |
Read the same twelve months three ways. The window a listing might show gives +18.1%. The whole year gives +1.3%. And the four months after the window, which is what a person who joined on the strength of the listing would actually have lived through, give -10.7%.
Calculated by us on 3 Nov 2026, exact. A published result is a record of the past, and the past it shows is the part somebody chose. None of this says anything about any real strategy. It shows why a result needs the whole history, the method that produced it and the dates it covers, before it means anything. The survivorship bias guide explains the wider version of the same trap, and overfitting explains why a strategy tuned to a good stretch can fail the next one.
Questions to put to a published strategy
- Who is responsible when it fails? Find out what the terms say about a strategy written by another user.
- Can you read the rules? If you cannot see them, you cannot know what you are following.
- How was the record made? Live as it happened, or built afterwards from past data, and whether costs are in it.
- How long is it? Does it cover a bad stretch, or only the weeks since it was written or listed?
- Has the strategy changed? A record can be several versions joined together, and the one you are copying may not have produced it.
- Can you inspect a trade afterwards? If you cannot trace an order back to a rule, you cannot learn from it.
Run the test somewhere that cannot hurt much
A hosted bot gets its prices and sends its orders through the exchange, so the cleanest first test is a paper mode, if the service has one. If it does not, open a separate exchange account or sub-account that holds only the test amount, and put the key on that account alone. Whatever happens, the loss is then the test amount and no more. See the crypto trading bot guide for the general ways a crypto bot fails.
What we have not claimed
This page takes no position on whether Cryptohopper works or how safe it is, and it sets no other service beside it. The year of returns is invented. Crypto-assets and leveraged products can lose all the money put into them, and past or simulated results are not a guide to future results.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What is Cryptohopper?
As we understand it, a hosted service that runs trading bots on the vendor's servers and connects to crypto exchanges through API keys. We have not run it, so the vendor's current pages are the place for what it offers.
What does hosted mean for a trading bot?
The bot runs on the vendor's computers instead of yours, so it keeps going when your machine is off. The cost is that your key sits with a third party and you depend on that party staying up.
Why does the strategy need separate checking?
Because the host and the rules are two different things. A well-run service can still run a poor strategy, and a good-looking record can belong to rules you are not allowed to see.
Why can a published record mislead without anyone lying?
Any record covers some window. If the window shown is a good stretch, the record looks better than the whole history. The example on this page reads one invented year as +18.1%, +1.3% or -10.7% depending on the months chosen.
How should I test a hosted bot cheaply?
Use a paper mode if there is one. Otherwise use a separate exchange account holding only the test amount, with a key limited to trading on that account.
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 with an invented year of returns. We have not run Cryptohopper, and the page does not suggest using it or any other service.
Know the Risk Before You Run It
The calculators and guides are free and need no account. Create a free account to log in to the Automated Trade Bots dashboard.
Free. Confirm your email, then log in. A new account shows Pending approval until our team switches it on.
Trading carries a risk of loss