It’s been over 3 months of learning, testing and coding since I first started working on a crypto trading bot that takes advantage of the Binance Announcement page to buy new coins as early as possible. The gate.io version of this trading bot looks like it’s been able to successfully make a profit using this strategy.
Context on the gate.io crypto trading bot:
Before looking at the results that this final version of the gate.io bot has generated, it’s important to understand how the idea started in the first place as well as the changes required in order to make turn this into a profitable crypto trading algorithm. You can watch the video down below for a quick recap, or continue reading:
The initial idea was to create a crypto trading bot in Python that looks at the Binance Announcement page, and places buy orders on new listings the moment they become available. The reason behind that is that a lot of new listings spike up a lot in the initial moments of the listing. Previous testing showed that no matter how early you buy a new Binance listing, it will always be too late to make an immediate profit. The fastest achievable speed is 0.1 seconds (which is simply not fast enough), due to the limits of the Binance API. Anything faster will get you a temp IP ban.
It was later observed that a actual spike in price takes place outside of the Binance exchange at the time that the Announcement is made, and not when the coin is actually listed by Binance – that’s just the peak of the iceberg.
Here’s an example of it:
See how the price of CHESS/USDT spikes at the same time as the Binance announcement is posted, check the timestamps.
Optimising the gate.io crypto trading bot
First of all, we call it the gate.io trading bot (more info linked) because it uses the Gate.io exchange in order to place trades. So instead of buying a new coin when it gets listed on Binance, it scans the announcement page and, once a new coin listing announcement is made, it will automatically buy this coin on Gate.io.
Right so with the strategy all nailed down, it took about another month of testing and optimisation before the bot would do anything interesting. But here we are, 3 months later with a functional trading bot that seems to be making decent gains on new listings (though it can be improved).
During the initial testing phase of the gate.io strategy, the main blocker has been the speed of execution. At the time, it looked like Binance is purposefully delaying announcements by 5 minutes to facilitate insider trading, however after some more investigation, that doesn’t seem to be the case. It’s most likely an issue caused by the X-cache of the endpoint that we’ve been querying. The workaround was to randomise the query sent to this endpoint, as well as replacing the current endpoint with a faster one.
I would also like to thank everyone who has contributed to the codebase on GitHub. With the discord server exploding, and so many people interested in the tool – my role unwillingly turned into a community /project manager rather than doing much coding in the last couple of weeks.
The results (so far):
So bear in mind that this only includes the latest iteration of the bot, before this update the bot was not profitable.
The gate.io bot has made 10% gains on MOVR (live mode):
… and 25% gains on ENS (test mode)
It’s still super early to tell whether the bot will be consistently profitable, but it’s looking rather promising so far. The tool comes with a test mode so you can test it out without throwing real money at it.
If you want to test this trading bot out yourself, see the guide here.