Wrote a trading bot

I built a trading bot that uses a combination of liquidity provision plus ta-lib statistical indicators to make about 25% return per month based on backtesting. I didn't believe it at first, so I ran a test for the last three months to see if it actually worked, and it did. I'm honestly not interested in working full time on running a crypto trading fund, so I am considering selling the program, but not the source code in case I change my mind. Would anyone on Veeky Forums be interested in buying it, and at what price range?

Some more details
>Runs in 3 modes, live, backtest, and import
>Backtest can run multitests that test 10,000's of strategies at a time, and sorts out the ones that are only positive for every month of the period selected (to prevent strategigies that simply luck out on a single price swing)
>Live is extremely robust, automatically restarts if it crashes for any reason, and will continue to attempt to reconnect through internet/exchange outages
>Currently I have Poloniex completely integrated, before I sell it will add Kraken
>Current strategies include RSI, StochRSI, and MACD. If anyone wants some other statistical indicator I can add it very easily.

Other urls found in this thread:

discord.gg/hY6j9K
twitter.com/AnonBabble

give me the source code for free faggot...

What language is it coded in?

C#, I like the language features, its a lot of code. Kind of regret it cause I use mono on Linux, if I were to do it again would write it in Python.

why would you sell it and not keeping for yourself running autopilot?

I forgot to add - its a margin trading long/short bot. 25% return is with a margin of 1:1, obviously it doubles with 2:1 margin but more risk.

I don't expect it to work for more than a year or two as the market changes. Every time the market changes you have to rerun the backtesting, which takes a few days for each currency pair. You can only put in a max of about 5 btc per currency pair (depends on the volume of the pair) before the orders in the book start to affect the market. On top of that, it does need to be checked in on at least weekly. There's no such thing as full autopilot when it comes to trading.

On top of all that, I am a poorfag and to really make big money out of this I would need to raise capital. I have no interest in going outside my apartment and talking to people.

screenshot of the UI?

What price range were you even thinking. I really can't imagine anyone paying for this.

ive also got some magic beans, anyone in on those?

If it really was 25% per month, you would just use it, modify it as needed for changing times, and shut your mouth, because you would make far more that way. Just like anyone trying to sell their strategy rather than use it to make money, it probably doesn't work the way you say it does.

...

see
I only have about 1000 USD, 25% per month for 1 year doesnt really get me much of anywhere. Nothing prevents me from running it while also selling it, and I get money now.

Also here is the bot booting when running live, i have lots more screencaps if desired

Which backtesting environment was used? How long was the live session? Sharpe ratio? Drawdown?

>Current strategies include RSI, StochRSI, and MACD
...and it's shit

One more pic, im running the bot right now with small amount as test

I wrote it all myself, including the backtesting.

I can add whatever you want, but I found those strategies work in combination with a few other tweaks.

Seems like there really isnt any interest here to buy this, which is fine. I'm abandoning the thread.

The true test will be semi-consistent profit on at least a 6 month period.

You can always fit a algo to a pre-existing data but it doesn't guarantee any results in the future.

uh you are really brave to run something that plays with money in fucking MONO!

because its not accurate implementation of .NET at all I would not use it even for non critical code

> How long was the live session? Sharpe ratio? Drawdown?

How did you implement backtests? Is it simply different data (historic vs. live) fed to the live trading engine?

rofl, runs away as soon as the interrogation begins

dont come here to LARP

>i ran a test for the last three months

Yeah... no.

There are 12 months in a year. You could have 3 winning months and 9 losing months.

Not to mention that there are going to be multiple years. Not just this one or the next.

Run the backtest as far back is your software will allow you to. Even if it takes an hour to complete the backtest.

For every bot strategy, there is a counter strategy to take advantage of that bot.
Therefore, the only theoretically successful bot is one that you keep a secret, and no one else figures out the same strategy.

Ill give you $50

I'll raise you 20 bitbens, 1 dgb token, and satoshi nakamoto's personal email.

lol no

If you’re willing to reimplement it in .NET Core, instead of Mono, I’d be very interested in buying. Please email me at [email protected]

i buy you for sharpies. contact me [email protected]

What statistical model are you using

hey big man, you should join Xaviers School, they have like 15 programmers, one being pimple who can actually help you a shit load with your project.
Also you can learn some better indicators than you listed.

all of the best traders from Veeky Forums are here now.
discord.gg/hY6j9K

Are you serious? We don't need genuine implementation for using it on calculating elementary math (statistical analysis use many simple math too). The important is will this work. FUD is everywhere and unpredictable.

Awesome

bot dev is private, useless channel

most retarded larp. no stats posted, fucking 3 month backtest, non-heteroskedastic MA model used to generate trade signals, fucking TA, 100% guaranteed spaghetti code, backtesting environment coded by hand with all of the inherent bias that comes with it.
> What could I possibly do with 25%
anybody algo trading for real over here?

Did you speak up and request to join any channels?

How the fuck do I learn about this stuff?

there is a link above to an educational chatroom

Your model is probably ridiculously overfit. I'm surprised you're making any money at all.

Did you just backtest on all available data or have proper validation and test sets?

2 Ark, full source code.

GITHUB OR DIDNT HAPPEN

Hey user

Nobody has ever heard of a backtested model that didn't work splendidly

You have no reason to believe it will work into the future and won't cost you massive amounts of money some night when something weird happens

Nice job and all but you're an idiot

hey no kidding i wrote C# api wrappers for bittrex, and bitfinex if you wanna colab im working on a framework

>C#
>Low latency trading bot

Never gonna make it

>implying python, the language that most retard baby bot makers use is faster

>Implying I don't also think python is shit
>Implying I'm not a professional C++ software engineer specializing in template metaprogramming and optimization tricks for low latency servers

Can i run this on my macbook pro
Please send link

you will NEVER get my LINK

>earns 25% return per month
>basically prints free money
>I want other people to have it

Yeah I don't believe you for a fucking second OP.

I made a telegram scrapper that finds pump and dumps and buys the coin as fast as possible, it's risky, but it werks

>Implying I'm not a professional C++ software engineer specializing in template metaprogramming and optimization tricks for low latency servers
Mah nigga

And now we.know you a fgt

How do you guys do all the calculations? The RSI, etc

fucking idiot C# is plenty fast, the bottleneck is rate limits and latency from the API

add a data visualizer mann

>Backtest can run multitests that test 10,000's of strategies at a time, and sorts out the ones that are only positive for every month

oh god

Ya is that wise?
That will weed out a good strat that experiences a bad month.

I heard there are a lot of people building trading bots as well and sharing stuff with other people in the +2500 member bitpam Veeky Forums cryptocurrency trading/mining/coding discord server

discord.gg/VpPrGmF

soon finished with mine :')

Good luck with it, OP. I'm proud of you.

That's 15k after 12 months you numbnuts

if you are interested in paying a dev to work on website and GUI hit me up

throwaway:
[email protected]

you could just have 10 000 coin flipping combinations short/long, then pick the best flipping strat. does not mean the strategy has any real edge on the market and likely will be disasterous in the future trading... jeeezzz...

Your best bet is to make a website and a UI (or work with someone who will without ripping you off) and then sell it as a product to boomers. You could even cuck some features and keep them to yourself or charge more for them.