I want to build a Crypto trading bot

I've been a software dev for years and finally decided to spend a few weeks building an automated crypto trading bot. Any resources you guys know of that could help me get started? Would like to know if there are known algorithms or economic strategies that are integrated into these bots. I know how to code, just need help with market analysis and/or pattern prediction. (Or however bots actually work in stocks/crypto)

Attached: trading-bot.jpg (750x350, 268.21K)

>asset go down
>buy
>asset go up
>sell
>a lot of positions open

Generate 0,5% profit with leverage every trade. Congratulations, you've made it

I've been working on a bot for a bit over a year part-time as a hobby... but I've spent a ton of time so far. Let me tell you what I've learned...
Building a bot is easy. Giving your bot profitable instructions is fucking almost impossible.
I build a system to test variables programatically like... indicators (for open and close), TP%, SL%, Trailing SL, differeing open and close timeframes (from 4h all the way to 5min). Using over a year's worth of historical data depending on the timeframe. Different candle types, tried "bear/bull" phases using an indicator to alternate between strategies over days/weeks...
Used several statistical calculations to try and determine what's reliable and what's not, like linear regression on balance...
All I can say is...
Unless you're going to use AI, building a bot using static rules requires that the following assumption is true:
Markets are logical/predictable and follow patterns. It's so simple but I've come to the realization that markets are just chaotic and you can't reliably profit off them by making logic rules. Seems like you should be able to but honestly I'm just about to give up.
There's a theory... i forget what it's called... but basically the more times you test something, the less reliable the results become... I've done close to 300 million tests and all I can say about the ~0.5% that were profitable was that it's just a coincidence.. that past performance is not indicitive of future performance.
I would love to be proven wrong but this is where I'm at now. There are almost no resources or communities for bot-making. Anyone that's profitable is keeping their secret, as would I.
So... before you invest a year like I have, consider yourself warned.

Check out hummingbot it’s free and open source.

trading bots will literally just make the same dumb fucking trading mistakes you do, but much faster

yep. I used it.
Pretty easy, However this user hit the nail on the head:
>>Giving your bot profitable instructions is fucking almost impossible.
Lets keep this on the down low before its ruined by tourists.

Fortune Favors the config file

>Lets keep this on the down low before its ruined by tourists.
Are you actually claiming you have a long-term profitable bot?
If so, congrats.
Can you spare ONE clue??

long term profitable? meh.
I could make bank on market flash pumps/dumps. The biggest convenience was not having to sit at the computer all day.Just set a floor and ceiling.
The second biggest thing your going to run into is the FEES on your trading.
That is going to eat away at your P&L.
Newer versions of "The Program" allowed for you to calculate and add it in your spread.
Playing with the config file was addictive.
Easiest way to make money was keep tight spreads, and run multiple instances with different floor and ceiling limits.( meaning take profit / stop loss)

Pic is what you would be playing with.
Like I said. It's addictive. Haven't been using it since Staking and Interest Bearing Accounts came around.
But with the recent economy- I think it's time to fire it up again.

Also , I have tried in house KuCoin bot... it's shit.
Funny thing is, you can use this program with KuCoin, with out them taking a 20% fee.

Attached: Disregard females.jpg (3009x1899, 1.03M)

>Unless you're going to use AI,
Yeah, the only way I see a bot truly succeeding is with AI and shit like that. Something that takes into account real-time news alongside trading knowledge and patterns... tagging news, making a dataset out of it, and getting resources for training a model would be a pain though.

With what youve learned from making a bot what would you say youve learned about the markets and trading? I know nothing about programming bots, but I try to understand how some people succeed in being profitable traders, consistently. Has all the knowledge you gained from your bot given you a better "feel" or broader pov of the market and trading itself?

During my heyday i fantasized about:
Basic real time market Info could be pulled from API calls.
Calculated within spreedshit, followed buy scripts which could update the config file(s) and adjust the bot on the fly ( bat file?)

What I learned:
The "feel" of the market - I look at my charts in heikin ashi candles.
It makes all the difference. Never looked at candles any other way.

I am not making a bot. I am imputing ranges and triggers into a pre designed ( and secure) automated trading program.
The more your going to automated swing trade with tight spread, the more you want to find a CEX which has the lowest fees.
( this is how , at the time, i jumped into Binance - and hoarded massive amounts of BNB)

It does not really matter what the price of the coin I have bought for, it just matters the range,
( I buy 1 buttcorn for 100... i set my floor as 100... when market goes up and down- your active, swinging...if it goes below 100 , bot is shut off and i still have my 1 buttcorn)

Tax time - Learn the importance of HIFO-LIFO-FIFO. picrel most were bot trades.
Please be patient I have autism

Attached: FIFO_LIFO_HIFO.jpg (2193x728, 152.85K)

this post was for you:
this is also for you:

This snapshot is how i would (try to ) calculate.
Stuff on the left was the variables to be adjusted ( with definitions as a constant reinforcement)
Stuff in the middle were the variables based on the config file, so i could easily plug in numbers

Stuff on the right was projections on how the orders were to be placed.

I concentrated more on the back end , I wanted to "ride the waves"
I know that picture is confusing. I am a retard.

Bottom line: try it out, play with it, and use paper trading to see if your strategy works before you actually do real trades. It will defiantly keep you off the porn and monke sites. :)

Attached: Autismo.jpg (3223x1835, 1.69M)

Holy based autist. Im too dumb to do that sorry. My quest is to understand how some people can "feel" the markets. Im a noob trader but Ive been pretty good at calling tops and bottoms just based of longterm charts with a few indicators and what I see going on, like Elon on snl during the peak hype and then the following fake breakout and this deadcat bounce. But I always fuck up somehow so Im always curious to learn about how other traders do it.

I cannot stress it enough, to the point I must, by definition, be a shill:
heikin ashi candles

>urm.... so like, i can code and all but i can't figure out what kind of ways to apply it to a stream of numbers
lol
what an absolute larp

do not help him
he'll be in stack exchange five minutes from now demanding someone writes the code for him; a standard jeet approach

>if there are known algorithms or economic strategies that are integrated into these bots.
if they're known already, then they're useless. The whole point of using a bot is to discover and implement a strategy that hasn't had every cent of profit milked out of it yet by others.

Are you really so retarded you think you can take an off the shelf algo and just print money with it?

no bro i believe it.
I know a Bass Player who is a Maestro.
That nigga cannot read an imperial ruler to save his life

but yes this is also possible:
shit I forget about those CHUTYAs all the time.
I felt that user's language was that of a native white man.

I'm out.

Attached: 1621990916811.gif (396x213, 494.18K)

just setup trading view email alerts it's the same shit

So your strategy just buys and sells at % intervals within a range? Is this not the same as the Kucoin "bot"?
What's the difference between this strategy and what Kucoin does?
Or am I misunderstanding...

I had some moderate success last bear market writing a genetic trading algo. It's a 'dumb' algo and not predictive per se, it's definitely not AI, but it sure beats the fuck out of manually guessing at winning strats. Basically pull as many indicators in as you can think of, doesn't matter if you think they're good or even sensical (eg I pulled in lunar charts and even used unit bias as a valid signal), just pull them all in and let the algo start trying them at randon against historical data. Then take the must successful backtests and start breeding those indicators together, add some mutations, run the new generation through again, repeat and repeat. Add a novelty function in as well so you don't get stuck at local maxima.

I gave up on it because I wrote it in a fucked programming language and maintaining it was a chore. But it was moderately profitable while it was up and running. I could look under thr hood of the most successful trading strats that I evolved and they would be full of the weirdest fucking signals - like buying only on thursdays during high tide and only when the s&p daily closed red three days ago. Shit like that.

This is interesting, but I would predict that what you found were random coincidences, and you wouldn't have success trading with those strategies.
Did you ever try implementing a strategy into a bot to see how it works? Or did you only backtest.
It took me a long time to realize that backtesting is fairly useless for programmatic trading.