Aitor

[Aitor]

I usually like to post updates when a cycle has ended, but in this case I will do in a middle of a cycle after a huge drop. BTC dropped from 65k to 32k and my coins dropped up to around 66% from the price the cycle started:

Worried? No, not at all. The bot keeps doing it job, lowering the average bought price with DCAs and with CT trades. And the higher the dump it is, the higher the profit will be after ending the cycle :thumbup:

On the other hand, this huge dump has been very useful to check how the strategy behaves in these kind of situations and I got important conclusions that have lead to some changes / improvements being coded to the strategy.

The cycle for USDT_BTC pair started at 650 aprox, and now the average bought price is 481. As I said, the bot keeps doing its work, and now is when you see how important is it to allocate enough funds to let the bot do its job when there is a big dump like this one. We didn’t make a calculator for fun, there was a very important reason for doing it.

So, I will be posting soon the profits after closing the cycle :wave:

[Aitor]

Nice :thumbup:

[Aitor]

Hi, in the case you set BOUGHT_PRICE on a pair which started with pre-existing QUOTE.
When Gunbot sold the pre-existing QUOTE, it can override the line with
“BOUGHT_PRICE”: null

Try removing the “BOUGHT_PRICE”: null line from config.js and it should allow gunbot to start up and cycle again.

Please ensure that BOUGHT_PRICE is either removed or overridden sensibly when QUOTE is completely sold down to 0, and it is NOT set to null.

[Aitor]

Hi, tradingview alerts only allow to change the strategy for one pair. Anyway, if something triggered the alert for BTC-LTC, and the same happens for BTC-ETH, it should be triggered too and get an alert to change it for BTC-ETH too, right? Other way would be to combine both alerts and autoconfig, or just autoconfig alone.

[Aitor]

Quick update. Things going good. BNB ends cycle with profit:

UNI got stuck for a few days after a big drop, but finally DCAs and CTs managed to lower the average bought price and exist with profit. One more time we see that reaching to deeper DCAs means higher profit when closing the cycle. In this case I got a +55,61$ profit.

More stats:

[Aitor]

Nice profit. I have seen some people saying “my bot is only using 20% of the funds, I’m going to increase TL”. I know it’s tempting, and increasing TL 2x = increasing profit 2x BUT medium-large dumps happen from time to time. And once alt seasson ends, I think a lot of alts will suffer huge drops. Just saying IMO is better to be ready to handle big dumps. Also note that after big dumps, the profit you get when closing the cycle is much higher.

[Aitor]

Thanks for sharing your results Peter. Outperform BTC is relative, I mean, if you run USDT-XXX and BTC pumps from 50k to 100k in 1 day, there is no strategy in the world that can outperform that, as it would need to buy BTC using all the USDT before the pump starts and sell it when it reachs to the top.

Same way, if you run USDT-XXX and BTC price drops, you are outperforming BTC even if you didn’t make a single trade… Why to care about something we can control? I mean, if you trade USDT-XXX, the target is to earn USDT, same way that if you trade BTC-XXX, your target is to earn BTC and you shouldn’t care too much about BTC price going up or down.

In my case, according to Binance, I got +13% while BTC trend was +2%. Anyway, I think Binance calculations are wrong. I mean, since I started testing SG Mutation v1, I went from 5160$ to 8400$ and Binance says +33% and  it’s +63%… That’s why I prefer Gunbot reports as it just states the profit you got in your base coin: USDT, BTC or whatever you use as base.

[Aitor]

Interesting to see results with BTC based pairs. I would be careful with some coins as I still have fresh memories of what happened in 2017 after Alt seasson, and some altcoins ended dropping 90%+ against BTC. I don’t mean to tell anyone what to do, just giving my personal opinion.

[Aitor]

Hi Patre, note that DCA points are dynamic and depend on several factors. Also Trailme affects a lot, and in some cases it can get much lower / better DCA points. Other important point to note is that the graphs are relative. What does it mean? It means that a graph ranging between 1000 and 900 will look the same in your Gunbot GUI than a graph ranging between 999 and 1000 as distances are relative. In the first case the bot will make a lot of trades taking advantage of a wide range (10%) and in the second case, despite you see a very swing chart, the volatility is very low.

[Aitor]

Hi Paul, Crypto Sight was integrated into Gunbot and it’s no longer a different app. Now Gunbot Telegram bot is the “new” Crypto Sight. It’s not as cool as CS used to be but it’s being improved slowly and hopefully it will be as good as CS was in a near future.

[Aitor]

Hi Paul, I have never seen an exchange dissapear from GUI after adding it and clicking “save changes”. Are you sure you saved changes? On another note, that error is related to connection error. Maybe a node between your computer and Poloniex was broken or you or Poloniex was facing connection issues. I guess this was a temporary error, wasn’t it?

[Aitor]

Last days have been very good. The bot is doing a good job grinding the charts. This is ETH graph of the last days:

More interesting info:

As I told you, I have removed USDT-BTC pair, so the profit I got with it is not measured in this report, but we can see the profit I got with the other coins:

So far the results and performance are very good. I’m testing a fix for a little bug I noticed since 24 hours . This bug only affects to v3 versions, but it’s not a big deal as even with it v3 is much better than v2 IMO. Hopefully I will be reporting soon the fix is working 100%.

On another note, I have created a new variant based on C6 but focused on being able to reach to deep DCAs with very low funds requirement. Its name is CB6 and it reduces funds requirement by 58% compared to C3, while C6’s reduction is around 39%.

[Aitor]

Changing settings dynamically can be made with Autoconfig. But it’s for advanced users and with Javascript knowledge. It needs coding…

[Aitor]

Confirmed by dev team: Automatic withdraw feature was removed to improve security.

[Aitor]

[quote quote=9872]Is it just me or is the DCA Guesstimator on the Gunbot Wiki not working. The sheet opens but just contains lots of missing reference errors?[/quote]

That tool is not an official thing, but I don’t think it even was accurate as even Backtesting Addon is not able to reproduce DCAs behavior. Also DCAs usually use trailme and that completely changes buying points.

[Author]

Posts