Hi there Botters,
Today I return to share with you that I’ve taken all configs on our community list and started composite bots on all of them to paper test their performance in 3Commas.
I’ve started the tests on 20th March, so any config added to the list after was not included in these tests, and did the sheet grow since then, awesome engagement from the community, thanks!!!
As a reminder:
If you want to access the configurations sheet, go here: https://bit.ly/TACBots
If you want to add a config to the list, go here: https://forms.gle/qxuvvsprDa3wMJB7A
As of today, the list does not have an average expected ROI column, but that is about to change...
One of our discord community members, Snardge#9572, has coded a personal backtest tool and is currently finalizing the calculation of "one year of backtests" on all our community configs, which for sure will shed some light on which configs work better on the long run, and which configs yield better results for each of the Binance top 200 coins… Awesome work right?
Once he publishes the info don’t forget to buy him a coffee in https://www.buymeacoffee.com/snurg! 🙂
For now, you can check the log files of Snardge early tests below, please note that at the end of the results the script closed all open deals, so the ROI is the reflection of that action, meaning that configs in the positive were managing funds better during those two months of backtests:
Jan&Feb 2022, on top200 coins as of 1.1.2022: https://cdn.discordapp.com/attachments/829512509798219788/955980804712968242/taCommunity_2022-01-01-2022-02-28_2022_top200.log
Log results converted to excel by IamtheOnewhoKnocks#0701: https://cdn.discordapp.com/attachments/829512509798219788/956017275880763423/SnurngBackTests.xlsx
Sellium Paper Tests
So, like I stated earlier, my new paper tests for the configs were started on 20th March, each bot configuration is a composite bot that runs 20 deals simultaneously, all using the same 20 USDT pairs, and for each the start condition is TV 1m buy.
The 20 coins/tokens were selected randomly, but also trying to keep a diverse bunch, they are:
ACA // ACH // ANC // AVAX // BNB // CHZ // EGLD // HNT // IMX // INJ // IOTX // LUNA // MANA // NEO // NMR // OXT // QTUM // RUNE // TORN // XMR
The current results
Below I leave you with the current table results for all the configs I’m testing, note that the ROI is calculated solely based on the current close deals, so all ongoing deals and it’s current state is not accounted for, and as I would have to go one by one to actually do those calculations, for now I’ll just leave that very tedious manual labor for when it is time to do the last status report.
The Bot Configuration column color codes means the following:
The other columns color code was calculated automatically and follows common sense, green is better, red is worse.
7 days results:
Sellium Paper Tests Data
If you want to access the paper tests data to confirm the results yourself, it is available here: https://docs.google.com/spreadsheets/d/1UsnopbKbbVoVurv0B3hs6kaxjYQeDZCqF6wQmNDIOGo/edit?usp=sharing
Some accumulated knowledge
Take Profit: It can increase the profitability of the config, during several community tests we’ve concluded that TP 1.5% or lower works better during bearish or a sideways market, and TP 3.0% is the preferred value for a Bullish market. Changing the TP will also influence the amount of Safety Orders consumed and the time each deal requires to close, so TP 3.0 will close deals slower and use more Safety Orders for each deal, and if you are not overextended you don’t have to worry much, but if you are, then careful when selecting a TP 3.0 for certain configs.
It would be nice to see a test of all of the configs using the same TP, 1.5 and 3.0, for a clear comparison of each of the best bot configs for Bear and Bull.
Max SO Price Deviation: This is a critical definition that you need to take close attention to, at first look it appears to mean how long will your bot continue to automatically DCA when the market is going down, but it is much more than that, and it actually can make a difference between an excellent and a terrible config during some market moves. First you need to understand that the higher it is, the more downtrend it will be able to absorb without any intervention, but at the same time more of your capital is not being used to make money, it is just sitting and waiting for when the market turns and you need to fill safety orders. At the same time, you also need to consider the gaps between each Safety Order, because that will determine how long your bot must wait to react to a market move, a gap that is too big might just be a waste of your capital in some situations. And if you’re an advanced botter, then maybe a lower price deviation capability is better for you, but don’t forget to set aside some funds that will enable you to manual DCA when the market is full bearish…
Return of Investment (ROI): It is hard to have a clear picture in the paper tests without accounting for the current P&L, but at the same time we will not close all deals at once unless we really need it to, we’ll always let the deals finish if we can endure the bags. So, it is normal to see low budget configs outperform higher budget configs based solely in the fact of the max cost per deal, but to be sure you have a winner, you also need to account the real profit you’ve made when after 2 months you decide to close all deals at once, I believe it is a great indicator/comparison of a bot performance.
What configs I’m currently running
Like stated in my previous post, from the previous tests the config I really wanted to use was Neimoidian, I think it has proven to be a good overall config, the only problem is the max price per deal, which was above my capabilities as I wanted to run 40+ deals, so I had to come up with a cheaper version, so I created Neimoidian Value, which is my current default go to config, and it has been performing good so far.
But more information is about to come to public, and my default go-to config might be changing soon, let’s see, so keep posted for more information… 😉
Thank you for your support
And it is all for today, I’m out of time to retrieve the direct links to each one of the bots, but if you want the direct link to any of the paper tests bots just post a comment below and I’ll paste it for you.
Hope this information has proven helpful to you, don’t forget to like the post and write any comments you seam fit, together we are stronger, together we’ll move forward faster and stand more to gain!
Community Power! 😊