Trading Automation: Ideas, Needs, and Implementations

[Criptano]

Trading Automation: Ideas, Needs, and Implementations

I enjoy automating both complete trading strategies and smaller supporting processes — the kind of tools that make trading easier and more efficient.
This thread is intended as a place to collect ideas, needs, and possible implementations. If certain ideas turn out to be useful to many traders, they could eventually become real projects.

Feel free to share any "wish list" items you've ever had — things that start with, "It would be great if..." or anything else you've thought about.

• What could help you with trading or investing?
• What would make your life easier?
• Eliminate repetitive tasks?
• Watch your positions or markets state while you're away from the screen?

Here's example from my experience:

I used exchange-provided grid bots for quite a while, but eventually ran into their limitations.
They simply aren't very efficient in highly volatile markets — and crypto, as we all know, is one big volatile market.
I started looking for sideways markets and discovered that stablecoin pairs fit that description quite well,
although their price fluctuations are naturally very narrow. Examples include USDC/USDT and FDUSD/USDC.

On Binance, these pairs have zero trading fees, making it possible to capture small but relatively predictable profits from their oscillations.
However, Binance doesn't allow its built-in grid bots to operate on these specific pairs. So I wrote my own script that implements a grid strategy for FDUSD/USDC through the API.
Early in its life, this pair was surprisingly volatile. Nowadays it's much calmer, but it still occasionally throws out sharp wicks — and the script is there to catch them.

The main advantage of this approach is flexibility. You can define different position sizes for individual grid levels and even create overlapping grid ranges.

In my setup, I use three primary ranges covering the pair's normal price fluctuations and allocate the majority of capital to them.
I also have two additional ranges designed specifically to catch upside and downside spikes, with smaller allocations.

So, getting back to the main topic of this post, what would you like to automate if you had the opportunity?

[Wiwo]

Alot is at stake for traders, because exchanges never provides us with the features to execute such tasks, and one area automated feature can help improve the experience of trader's in the area of risk and portfolio management, like market monitoring automated feature that automatically adjust traders portfolio seize when market uncertainty happens.

Readjusting traders portfolio to keep them away from possible liquidations.

Stablecoin grid features is a good  example of how automated tool can solve traders portfolio limitations that exchange features will never solve, should we now say that exchange limitations are their house edge against trader's,

[Oshosondy]

People should know by now that I do not like trading automation, but I like some like TP which is directly on all the exchanges that I am using and that I have used, but let me answer your question.

Feel free to share any "wish list" items you've ever had — things that start with, "It would be great if..." or anything else you've thought about.

• What could help you with trading or investing?
• What would make your life easier?
• Eliminate repetitive tasks?
• Watch your positions or markets state while you're away from the screen?

Waiting for the bear market like this time and investment using some strategies.
Swing trading and long term holding for bitcoin.
What do you mean about repetitive task?
I can not watch my trading PnL all the time because it can make someone to be very tired and not a good way of trading if you are not scalping.

Look for good trading bots and AI, it might help you. Do you know about the new Bybit AI account? I do not know about it yet, but I open a sub account for it already. I am not likely going to use it.

[DPHOR]

I do not like using bot to trade, although I know that many people usually used that to back themselves up, but whenever the loses come they cannot control it , mistakenly if the bot has technical issues loses is very easier for that person. If I will advise person about bot is that do not use a higher amount to trade so that the loses can be regulated and control. It's better to trade personally without deploying bot that could make you lose. Although, you can also use your stop lose, and take profits so that if the market goes against you, you would be able have profits or even reduces how much you should be losing while trading.

[Alphakilo]

People should know by now that I do not like trading automation, but I like some like TP which is directly on all the exchanges that I am using and that I have used, but let me answer your question.

Feel free to share any "wish list" items you've ever had — things that start with, "It would be great if..." or anything else you've thought about.

• What could help you with trading or investing?
• What would make your life easier?
• Eliminate repetitive tasks?
• Watch your positions or markets state while you're away from the screen?

Waiting for the bear market like this time and investment using some strategies.
Swing trading and long term holding for bitcoin.
What do you mean about repetitive task?
I can not watch my trading PnL all the time because it can make someone to be very tired and not a good way of trading if you are not scalping.

Look for good trading bots and AI, it might help you. Do you know about the new Bybit AI account? I do not know about it yet, but I open a sub account for it already. I am not likely going to use it.

Here goes again the need for convenience being introduced by BYBiT AI, and seriously this is why most people would always fall victims to so many uncertainties that arise from trying to do things the easier way.
I must commend the grid bot setup by OP as a clever idea, because to think of even creating a solution with API driven trading is more easy than many would come to agree with, instead of AI bots always on task.

I would for one appreciate or rather let me use the phrase to begin my sentence which is, it would be great if such trading automation could help protect my capital from sudden market dips and liquidation.
It would be great if this setup could also alert me as to when a sudden and abrupt liquidity space surfaces in the market.
It would be great if this setup can also help trigger a command prompt to help me hedge highly leverage positions that comes up not just from current price but also from details about the way I do trade, because scalping or slot trading might be better, but a lot of profit can be made in one shot from leveraging positions..
It would be great if this trading automation can do major work while I can freely set my self free from starring at the screen all day or trying to understand candle sticks and analytics, in order to know the best decision to take.

[BitMaxz]

Actually, a grid bot is risky in crypto; it is way safer to use in forex than using it in crypto.

So, getting back to the main topic of this post, what would you like to automate if you had the opportunity?

Well, if I had the opportunity, I would want the RSI and MFI to cross 50 only if the price bounces to the 200 EMA or near the 200 EMA and I can automate trade only when the market is trending.
Using the RSI and MFI is to filter out my bad trades, but for me this works manually with price action and analysis, but I do not know if it would work if I made it to automate it. Someone on the ForexFactory forum already made a bot of this but only for forex with a good result, but I do not know if this one will work on crypto, but for me, when I trade it manually, I get a good result. Maybe it would work if the price action could be included in the script, but I do not know if it's possible.

[JeffBrad12]

I would like to automate hedging and farming funding rate yield across exchanges, not limited to one market.
A bot that will find funding rate opportunity with annualized APY rate of 50%  gonna be massive, basically free money without the risk of volatility.

Tbh, you might find your luck executing those trading bot that you've made trading stablecoins on USD1/USDC pair because it fluctuates a lot, narrow but not too narrow.
Biggest risk would be USD1 depeg. This coin kept getting used by exchanges to distribute free money for some reason and rate keep fluctuating.

[Jaweria parveen]

We need all these in trading because without them we will not be able to analyze but those who are experts improve their thinking more due to which they understand the market graph and trade. But other people do such things in which they want us to trade with automation and we get good signals from which they can trade easily. But better than this we should have self-knowledge from which we can go to the market and trade anytime. If we stay in these things like this then it will remain the same for us and if ever we have to go to the market then we will do nothing but loss. Therefore we should also understand that we will improve ourselves and work hard with our knowledge.

[Criptano]

As a developer I've to say that any automation requires a clear, well-defined algorithm. Only then is it feasible.
We need a strict formalizaton of any "I wish.." statement.
If you know, how to properly define it, it becomes the doable task

automated feature that automatically adjust traders portfolio seize when market uncertainty happens
Readjusting traders portfolio to keep them away from possible liquidations.

We need clear criteria for what qualifies as “market uncertainty” and a well-defined method for how to readjust the portfolio.
Because the portfolio is dynamic and “Market uncertainty” is a very vague concept, I don’t think this kind of task can be easily formalized in a fully unambiguous way.

If I will advise person about bot is that do not use a higher amount to trade so that the loses can be regulated and control.

It’s is very simple to implement a condition in a bots algoiythm such as: "If deposit < 1000, stop trading. Send notification to Discord"
Within a trading strategy you can use both takeprofit and stoploss levels - this is entirely possible for any automation trading system.

but whenever the loses come they cannot control it , mistakenly if the bot has technical issues loses is very easier for that person

This only becomes an issue when risk control rules are not explicitly defined within the algorithm.
But in principle, what prevents us from defining those constraints upfront?

[joniboini]

We need clear criteria for what qualifies as “market uncertainty” and a well-defined method for how to readjust the portfolio.
Because the portfolio is dynamic and “Market uncertainty” is a very vague concept, I don’t think this kind of task can be easily formalized in a fully unambiguous way.

True. I think trying to define them through TA-based approach can help. For example, if the value between each indicators vary wildly from one extreme to another. If the fear index is high while the volume is low, and so on. This assume we can try and test many things. If the input is bad the bot wouldn't work properly too. You'll have to rely on in-depth analytics on this, not vague posting from social media or other sources.

[Criptano]

What do you mean about repetitive task?
I can not watch my trading PnL all the time because it can make someone to be very tired and not a good way of trading if you are not scalping.

For example, every day at 12:00 UTC, take the top-ranked symbol from Binance Futures (24h changes list)
and analyze its movement: whether it has moved up or down by 10%.

Log the result somewhere for further analysis. After collecting enough data, it may be possible to open a real futures position in the corresponding direction.
I implemented such a task for myself

What else... Twice a day, to send an account status report to Discord or Telegram.
It may include, for example, spot / futures USDT balances (both total and available funds), along with any other relevant account information.

Any numbers and parameters you see in the exchange application can be retrieved via the API, structured in a specific way,
and used to perform certain actions based on the collected data. About your account, about positions, about symbols prices etc

[EarnOnVictor]

I never knew the importance of trading automation until I started it myself. Despite trading for over 20 years in the manual way, I'm now lazy to even analyse the chart again, despite still achieving the same goal in a better way. TBH, this is what many traders should look into, and this is beyond why brokers/exchanges have put in place for us, but what we can coordinate as a combined units into one system, which entails everything possible that a trader could need, especially the entry/exit logic and the management thereof. This also helps our psychology.

[Criptano]

That's interesting ideas, Alphakilo.  Thank you for joining the discussion and sharing your wish list.

Let me try to guide you toward turning these ideas into something much closer to a fully automated system.

It would be great if such trading automation could help protect my capital from sudden market dips and liquidation.

Could you describe in more detail the criteria for identifying sudden market dips and liquidation risk?
Perhaps everything can be reduced to liquidation risk, since sudden market drops bring you closer to liquidation.

So, how should liquidation risk be defined?

• The asset price approaches the liquidation price, and the difference falls below a certain percentage.
• The margin ratio exceeds 70% (this would work in isolated margin mode).
• Some other criteria?

The second question is: what should the system do once liquidation risk has been detected?

• Close certain positon / all positions
• Try to add margin if there are available funds.
• Try to add margin, and if the available funds are insufficient, close the position.

And the third question relates to your next wish:

It would be great if this setup could also alert me as to when a sudden and abrupt liquidity space surfaces in the market.

How exactly do you identify a liquidity space, and what criteria make it "sudden and abrupt"?

For example, over the last hour (or some other period), by how much should the liquidity space increase? Relative to what baseline?
Which notification channel would you prefer?

In other words, all the factors and conditions that our brain takes into account when making decisions — including those
we may not even be consciously aware of — need to be identified and formalized in order to build an automated system.

[Criptano]

I never knew the importance of trading automation until I started it myself. Despite trading for over 20 years in the manual way, I'm now lazy to even analyse the chart again, despite still achieving the same goal in a better way. TBH, this is what many traders should look into, and this is beyond why brokers/exchanges have put in place for us, but what we can coordinate as a combined units into one system, which entails everything possible that a trader could need, especially the entry/exit logic and the management thereof. This also helps our psychology.

Could you share a bit more about your experience?

How did you get started?
When did the idea of automating your trading first occur to you?
Did you encounter any difficulties during the transition?
Also, does your automation script run on a server, or do you run it locally on your own PC or laptop?

[justdimin]

I never knew the importance of trading automation until I started it myself. Despite trading for over 20 years in the manual way, I'm now lazy to even analyse the chart again, despite still achieving the same goal in a better way. TBH, this is what many traders should look into, and this is beyond why brokers/exchanges have put in place for us, but what we can coordinate as a combined units into one system, which entails everything possible that a trader could need, especially the entry/exit logic and the management thereof. This also helps our psychology.

That is what happens if you ever come across a truly remarkable software that can do it for you. However, if you do not find one, then there are many "bots" out there, who just does very simple technical analysis based on some indicators and hope that it makes them a profit. Reality is that you should still be focusing on yourself and your own knowledge.

Difference between you and a newbie is huge even when using an automation. Because you know how it should be, and can stop and fix anything that goes wrong, a newbie would not know how to do that and would end up with bad results. This is the most important difference and we should be caring about this difference without that much of a problem. If you do it, it's fine, if a newbie does it, it's risky.

[Criptano]

That is what happens if you ever come across a truly remarkable software that can do it for you. However, if you do not find one, then there are many "bots" out there, who just does very simple technical analysis based on some indicators and hope that it makes them a profit. Reality is that you should still be focusing on yourself and your own knowledge.

Difference between you and a newbie is huge even when using an automation. Because you know how it should be, and can stop and fix anything that goes wrong, a newbie would not know how to do that and would end up with bad results. This is the most important difference and we should be caring about this difference without that much of a problem. If you do it, it's fine, if a newbie does it, it's risky.

This thread is created with the idea that instead of searching for some external truly remarkable software,
we can build our own system — one that follows a clear and understandable algorithm that we can fully control.
Finding a developer is actually a secondary task, and in many ways the most straightforward one.

The first and most important step is to clearly understand what exactly we need:
how the system should behave and how to formalize it into a precise and unambiguous set of rules.

The second step is highly desirable — to test this logic at least on recent historical data, in order to make sure we have accounted for all relevant nuances.

Only after that should we move on to writing a trading script. The developer’s task is to accurately implement the defined logic,
and during debugging to identify potential edge cases and properly handle them in the code.

From my experience, this process is not fast, but it is very rewarding. The time invested eventually turns into a working automated strategy that requires minimal intervention.

[EarnOnVictor]

I never knew the importance of trading automation until I started it myself. Despite trading for over 20 years in the manual way, I'm now lazy to even analyse the chart again, despite still achieving the same goal in a better way. TBH, this is what many traders should look into, and this is beyond why brokers/exchanges have put in place for us, but what we can coordinate as a combined units into one system, which entails everything possible that a trader could need, especially the entry/exit logic and the management thereof. This also helps our psychology.

That is what happens if you ever come across a truly remarkable software that can do it for you. However, if you do not find one, then there are many "bots" out there, who just does very simple technical analysis based on some indicators and hope that it makes them a profit. Reality is that you should still be focusing on yourself and your own knowledge.

The number 1 rule here is: Don't ever rely on anyone's automated system, of course, unless it has been tested to be working for at least a full year and the result must be very encouraging. Mine wasn't about relying on other people, which makes it distinct, but building mine from the scratch, based on the over 2 decades experience that I'd gathered. I just don't want that stress again. All I do now is to continue to watch out for the bugs/algorithmic flaws and improve/optimise the system.