Based on TTP’s 6-month QFL video, the aim here was to expand the test to include all QFL types across a range of QFL percentages. The base bot is a Gordon bot, set to a single maximum active deal (MAD). Take profit is 2% (0.2% trailing), with 3 safety orders covering down to 7.5%. The plan is for zero manual intervention and to just leave the bots running. Every few days I will check in on them and update the tables below.
This isn’t a test to optimise bots, more one to assess how the different QFL flavours perform. A more complete test should presumably include varying the safety order setup according to the QFL type and the percentage.
I have duplicated these same 24 bots to have 20 maximum deals. This is to test the quantity of QFL signals. Results to be added in the next week or so - refer to the second page of this document.
Data lost from 2021-05-08 to 2021-05-09 due to a 3c glitch switching off paper trading bots. The results will reflect reduced profits as there was a small BTC dip that would have triggered QFL during this period.
I’m still updating these tables but there will be a definite slow-down or pause in most bots following the BTC dip on the 21st May (some QFL bots survived the earlier dip on the 19th very well, but the successive dips and the comparatively “light” safety order setup on these test bots has caught them out). For reference, the top performing single-deal bots prior to this were QFL D5, P5, C5 & O6. D7 & C7 were also catching up and survived the dip, so depending on the speed of the market recovery, they will likely show the best results going forwards.
After struggling with CSV exports from 3Commas for a while and not spending too much time resolving the issues I’ve finally managed to update the database. There are a few slightly suspicious results from the 3c export which I need to check and/or filter out but generally the results are strong for >230 days continuous running unedited.