Topic: what trading strategies could a trading bot implement? (Read 324 times)

But if you have a possibility as a user to tweak/re-learn your algorithm on demand, would you consider to try out this kind of bot trading service?

You need a third part website to show the traders trading history, you can know his assets / trades / profit ..etc clearly, then you will know what kind of trader he is. I don't believe in any trading bot, because market are changing in every seconds, it won't have any bots writen by some codes can foreknow the markets trend.

And i know some people will say , how about quant trading? Yes, i am a quanter , i have my own trading machine but not a bot which will run by itself forever. I need to focus on market and coding to make sure the machine will do the right things, not easy like 'a bot' .

yeah, i have actually seen people use this strategy and i have to say it is one of the dumbest strategies that i have ever encountered. basically they are relying on something completely unreliable (shitty news sites, and whether people would react to the news or not). and it will blow up in their face more than them getting lucky and their expectations comes to reality.

even if news leads to hype or panic, that is something that your bot should be able to detect from the market itself by watching the volume, price, orders,... without needing any news sites.

I actually tried bots sold by big organizations to us. I bought over $ 2000 and it was the money of 5 of my friends and I put together to buy it and try it out. in the early days it worked quite well, our account kept increasing after every day, after 1 month we got over 5%. but then the bot developer left and the old algorithms once again lost our money. I still support the use of bot trading but need to choose the right organizations to be able to buy good bots

Definitely start up with a simple strategy and try to keep it that way. Anything more than 2-3 variables and it is likely to fail. Most stuff out there is junk info so try to do research in some places/sources verified by the time on markets (dozens of years preferably)

If 1,000 FOO is your maximum risk outlay, then you would want to start much smaller. An initial position of 10 FOO (1%) maybe. That would give you a cushion of 6.58 straight losses before destroying your bankroll.

I prefer to view risk in terms of BTC or USD because that's what we're really risking but you probably get the point.

A bit could use certain triggers to make trades so you don’t have to be watching the market all the time. It also helps remove emotions from trading including panic buys and sells.

There seems to be a trend in keyword trading. As in selling when you see the words "china bans..." in news items, and buying when you see the words "bitcoin approved" in news items. You get the picture. They're trading on news items and using the bots to scan for certain keywords in the text.

Bots aka automation tools can be virtually built for anything that requires algorithmic(step-by-step) structured thinking. A programmer simply creates a set of rules/conditions/steps that the machine reads and executes if the conditions are met. Depending on the developer's skills and the complexity of the trading strategy I think that traders can automate their activity and just monitor the bot. In fact that's what many big hedge funds do. Traders only monitor the algorithm and market activity and the risk management protocols.

Sure! The only solution and a "win-win" strategy is to give proper toolset for creating custom algorithms/trading bots and apply profit-sharing model.

In fact, bot trading is a hot trend on Wall Street and with big investment funds in countries around the world. bots can now handle many different charts and signals more accurately. Besides, it works very accurately and is never hindered by psychological factors. Therefore, it is being used a lot for trading in foreign exchange and stock market. But in the crypto market, you will not be able to understand how the market goes and more dangerous is that you send the wrong money to the fake bot. A lot of shoddy bot sellers just make money and then run away.

I think is someone created a really good trading bots and can make sure profits they won't sell it for commercial because they can make profits of their own. So i think if someone selling a trading bots that means they make money just from selling the bots not by winning trades with the bots

All we know the use of trading bot is not mean that you will do nothing and yet you will have income. The strategies are also in your hand. I think in using trading bot we reduce loss because we are not performing greedy trading which the common caused of loss here in trading activity.

although i may add something like this but i wouldn't add it as a strategy, i would only add it as a part of the analyzer to only generate suggestions not execute orders. because i don't think leaving order execution to bot only based on "spread" is a good idea in a small market specially when it comes to altcoins.


here is my view on one should trade:
lets say there is a high risk coin called Foo. because of it being high risk, i decide to buy 500 Foo. in other words i choose the number "500" because i calculated that it is the value that i can afford to lose based on my risk assessment.

but if i am using martingale and lose once, then i have to buy 1000 FOO and that means that i am no longer doing all that risk assessment and no longer investing what i can afford to lose. in other words i just bought a huge amount of a high risk coin. not to mention that if i am in some consecutive losses i may end up buying something like 8000 FOO,...

that's why i don't like it when it comes to trading. i think each trade needs its own individual assessment and decision making.

I wouldn't leave it up to chance and employ it randomly. The optimal way to use Martingale would be as a method of scaling in and out of positions in a larger swing trading strategy. Swing traders and position traders often enter positions the same way, with weighted dollar cost averaging.

Any bot employing a trading strategy (as opposed to market making or arbitrage) needs to be tweaked regularly to trade in the direction of the trend. A friend of mine once told me that you need to be a good trader to run a profitable bot for that reason.

Despite that though, mean reversion is a strong factor in all markets. I believe it's generally true that (outside of stocks, which can easily go to $0) Martingale will work nearly 100% of the time on a long enough timeline. It's just that most people don't have enough capital to ride out the market so they need to employ a more efficient strategy than just mean reversion.

Hi!

I think we have something to show you: we've launched an on-line toolset for building custom trading bots on Neural Network and decided to take completely different approach to automated trading. It perfectly suits for those who don't have enough experience to create or follow their own winning startegy.

The core idea is that a user only have to apply proper risk-management to the algorithms (which are created by Neural Nets specifically designed for trading needs) and set up "Bot Targets" which are as follows:

- "realized profit" - positive target you want your bot to reach and stop trading. It could be 5-10 or 20 trades. You never know.
- "realized loss" - max loss you can afford trading with this bot. It'll also stop trading when loss limit is hit. So, it's a bot overall loss trading limit (and not a limit for a given trade!)

Think of it this way: you are a commander and your bots are your soldiers. You have to manage them and set them goals. But not micro-manage..

How your "soldier" (your neural net bot) will reach the goal is up to him. Your responsibility is to train them (run a learning process), set goals and discard....and "recruit" new ones. It's an ongoing process.


Each algorithm created by Neural Network module is unique and generates its own signals. It takes approx 2 minutes to complete learning process.

You can give it a try in demo-trading and tell us what you think about our project. It's very beginning and that's why it's so important for us.

p.s. our thread: https://bitcointalksearch.org/topic/bipooncom-bipoon-build-your-custom-trading-bot-on-neural-network-5090996
our TG chat: https://t.me/bipoongroup

In implementing a bot for trading, one of the useful features of the bot trader machines is having an automatic trading system. The robot detects the range price of the graph, and if the amount of the coins is going up and suitable for investment, it will sell your coins. This kind of robot or software robot is ideal for people who want to earn money and doing nothing because robots can do. Having a good algorithm in the program of the bot it will collect all the data that can use as an advantage to the other users. Having a bot trader is not good too because if you got already an error because of the data collected, there's a chance the bot makes a wrong decision and can cause of loss of income.

Recently he had also been investigating about bots, in fact, a friend was programming one to use it in Bitmex, according to him who gives good results, but that has the virtual money version of the Exchange, which makes many improvements, he It works on it, but that is a good method, in the majority that I have seen of bots they have been for the forex market, the truth I do not know much about that market, but for Crypto a while ago I found this article: https://eng.ambcrypto.com/cryptohopper-introduces-market-making-and-exchange-arbitrage/ which redirected to this page where there is still a trail version of the bot: https://www.cryptohopper.com/, which that bot does the following:

- What the Market Maker bot will do is negotiate the difference between supply and demand or (purchase price / sale price), better known as "spread", this bot will continuously market the market of your choice and ready the bot will do the the rest.

- The Arbitration bot, what it will do is look for market inefficiencies in any Exchange to make its transactions very fast and generate small profit percentages.

I remember signing up but I didn't follow up anymore, for that moment I made a topic in my local Spanish area section in case you want to visit it too: https://bitcointalksearch.org/topic/cryptohoper-presenta-su-bot-para-market-maker-y-bot-de-arbitraje-5192971

There is also a free version that ends this Sunday of scalping based on quantitative trading, the trader who created it is of Hispanic origin, but has its section for English speakers, is sending its signals based on something that goes beyond the bots , you just have to have a Discord account and join, the rest is to wait for the signal and operate, the page is: https://scalpingjedi.com/

thanks for the pointers.
i don't really like leaving the "speculation" to the bot since it is not AI and it is just a bunch of scripts. so i prefer strategies that aren't based on this type of speculative market analysis (relying on MA, MACD,...) although i have to look more into these since i am not really an expert when it comes to TA.
that is why i prefer arbitrage since there is no speculation involved, only solid fool proof profit with risk. and since the risk is part of the deal, it is acceptable.

as for Martingale i have to say, i would never even think about implementing something like that as i find it very harmful. in my experience running after losses like this will always end horribly...

One idea commonly employed is the Martingale strategy. Your bot would double down on losses and then repair its position once the market moves in its favor. Doubling down increases the chances of a profitable entry but also necessitates significant capital requirements, to ride out bad runs.

Some other ideas:

• Simple indicator bots: Buy bullish MACD cross, Sell candle close below Bollinger Band, etc.
• Market making: Constantly set orders on both sides of the book to profit from the spread. This works especially well on exchanges like Bitmex that pay a fee rebate for maker orders.