Topic: Let's talk about homemade bots (Read 225 times)

Ok I see what could happen to me

Yes I do reinvest, but instead of increasing the money I put on one particular coin, I just set the bot to work on another altcoin to diversify a bit. I explain the fact that the curve stays linear due to two reasons, first these new coins are less profitable. Before chosing a new market, I test it a bit and obviously, the first ones I chose were the best (although I could be wrong). Secondly, the global market of altcoin is deteriorating. After the January crisis, all the altcoins dropped, but with a really good volume nd a lot of panic sells and fomo the dip buys, that made the conditions good enough for the bot, but now, the coins are still dropping, but the volume is not here anymore.

Only time can tell me how it goes. At the moment I let the bot run as it's still profitable

I made my own bot too. It worked fine and made profits from January to July 2017. Then all alts crashed by 80%, and the bot was left with coins bought close to ATH.
I trade manually now

Looking at your curve, I take it that you don't reinvest the profits in the bot ? Otherwise I would expect an exponential shape curve, instead of linear.

You should never give your API keys to a third party as they basically give full access to your money (no password and no double authentication).

I don't plan to let people use my bot, so no worries here haha

To be accurate, it isn't Binance being hacked, but the bot that ppl used on Binance.
It could happen on any exchanges, but that bot was built for the Binance API and the ones affected exposed their API keys to the bot.
I don't think Binance should have reversed their loss at their own cost IMO, since it wasn't their fault.

Exactly, especially knowing that the cryptos are so volatile! I'm not actually putting the first bid or first ask, depending on the conditions, I go a bit deeper in the order book, I can also modulate the amounts trading to reduce my exposure. I implemented a lot of parameters to adapt as good as possible to the market.

After the latest binance hack Im staying away from any kinda things

After the hacked-bot used on the Binance platform (not Binance fault -- and they handled it superbly incurring a loss on themselves to reverse the hack, when they didn't have to), people will have to be pretty brave to use 3rd party bots? -- though not saying that this one is a "Trojan-bot" too, but just providing caution to the community.

I don't like robots very much. High-frequency trading does make money, but it is also a good choice to hold bitcoin.

So I guess it's "spread bot", what you'd call a dealer in traditional stock markets
that's interesting because I was thinking about coding one, there is plenty of not-so-much liquid market on medium market cap coins
but you still have to guess how the market will "move": will people tend more to sell at your bid or to buy your ask. how do you manage that without human supervision?

Yes exactly, don't use bots you didn't code yourself. Even if it's opensource, someone can predict how the bot is going to react by studying the code, and manipulate the market to take advantage of it. This is why my bots are only for my own trading activities.

I never use any of the bots, for me bots must be open source otherwise I can't trust them though I don't look at bots, some might be open source as well. In some situations, they are really good.

Wow, there is a lot more interest that I would have imagined

So yeah, obviously I can't share too much information about my bots, but I would be happy to talk about general problems one can encounter with bots.

I started making it mid January, after noticing that my trading strategies were profitable. The only problem is that I couldn't stay in front of the market all the time (I need to work and sleep also :p). At first I was only letting the bot run under my supervision.
The hardest moment was to let it work during a full night without looking, but then in the morning everything went fine and I'm pretty confident now.

I'm always working on it, but I'm now really close to the definitive version. Overtime I corrected a lot of bugs and coded little features, to avoid trading when I have the suspicion that someone would be manipulating the markets to take advantage of the bot's strategy.


Cool I'm not alone! Although you look more advanced than I am. How long have you been using trading bots on the crypto markets?

---

So to answer questions:


It doesn't short (this is a problem on the markets i'm working on because as I said, it means I need a little bag of the altcoin to be able to trade in both ways and it makes it relatively sensible to the market fluctuation)
It can market buy if a person puts an order that is really off my market range.
I won't share the algorithm
I cannot back test it as I'm creating and taking orders that wouldn't be there if I didn't put my bot on the market.
It can actually fail, and it has, but the idea is to use it on different altcoins and on a larger time period so that on average, it's never failing to make profits (for the moment)
I already tried some stuff to make it less sensible to market fluctuation (check the graph, on the weekend of the 5-6 of march, I wasn't home and I let it run with the objective of keeping the minimum amount of altcoins, but it was not efficient at all, so I went back to the old method)


I don't like manual trading, because if you have strategy that is not based on emotions, just on market analysis, it means you can make a procedure out of it and then code it. My bot is relatively dynamic, as it can understand trends and stop buying thousands of a dropping coin.


Yes exactly, I've coded some specific conditions to stop any trading if I have the slightest doubt about market manipulation. I talked with an algorithmic trader on the stock markets that told me about stories of manipulation.


Yep if the liquidity increases too much and the spreads are reduced to nothing, this current bot won't be able to earn so much. It's just coded to stop trade on any too liquid altcoin. I have the project of making another type of bots for really liquids markets (like USD BTC) but it will take some time to finish it.


I started programming mid January.
There is no version number, I make improvements everyday
I don't think I can release it, if it worked against itself, it would create a price war and reduce the profits drastically.
I'm making profits since the beginning, as it's just the formalization of a manual strategy I had.

How long has it been since you started programming this?

Which version is it?

Planning about releasing it or not?

Since which version can you see some solid profits?

It's hard to comment on your bot, since the algorithm is not known. You talk about liquidity for the ability to trade a large volume. It should be noted that with an increase in market liquidity, the algorithm may change. Do you consider this?

It's easy to manipulate bots with fake offers though, Using a bot could backfire BIG time.

interesting to know this .
but still i believe on manual trading bcoz strategy may change any time in such markets as waves fails many times .
so strategy is always dynamic .

Surely the most interesting thing would be to know what algorithm you are using.
Making it fully open-source would even better: sharing is caring, and like you said, it would increase (at least by a little) global liquidity, notably if it use limit orders.

Doing market makers bots for small markets with big spread would also be great and healthy for the corresponding markets. With a good strategy, and, if needed, human supervision, it can prove very profitable.

overall, congratulations for what you have achieved but I think you need to disclose a lot of more specific details: does it short? does it market buy? what's the underlying algorithm? have you back tested it? are you confident it would never fail? what are the exact markets you've used it on so far? any idea to make it more safe/successful?

Every serious trader is using own bots, or at least this is my logic. I have a dozen, for almost every popular exchange.

We can't comment much on your bot as you gave us a single profit graph. But TBH I would't disclose what my bots are doing either. Anyway, good luck - making bots and automating stuff is the right path to more profits and more free time.

Hey!
I know pHP but no idea about rubys
Do you have any reference sites? I would love to make my own bot...
I would really appreciate it...

If you want to start, you can check tutorials on how to use APIs of your favorite exchanges, most of them are using Python. You'll learn how to place an order, cancel it, read the order book etc... Then you just need to implement your strategy

Although if you don't know how to code, start by learning, it will always be useful, and you really need to get that kind of if/else logic to make trading bots.