Ultimate Bot v1.42 Release Notes
76,000 lines of code. Over 14 months active development. Final details hopefully taken care of. Here's what's new:
Pacman logic update
Pacman no longer uses martingale, instead an averaging grid that's monitored every tick. What this means is the original Waka Waka strategy risked 100% of your account for 10pips profit, so I zero'd in on the most stable part of the strategy, AUDCAD Sells, and it now has a risk-reward profile from Infinity:1 to now 8:1. A loss wipes out 8 wins. With this averaging approach, volatility filter and others, I finally re-approve Pacman for use.
Quantum King goes gridless
King now is set to disable grids as default, switching risk to reward profiles from ~150+:1 to 1:5. The grid version can still be used, but running with Real Ticks backtesting shows a big discrepancy between 1minOHLC. Here's a side-by side of the grid version vs the non-grid. 1k -> 40k Recovery factor 4.74, 61%DD for the grid. 1k -> 80k, Recovery 7.51, 45%DD for the non-grid. Also the non-grid option can now work natively with netting accounts.
AlgoRebel Fridays has ATR as an option (default)
Friday XAUUSD Breakouts have made many correct predictions since release, but have been stopped out with its limited 10pip SL too many times. I knew there had to be a way to use the ATR to scale with volatility correctly, my first attempt failed but this time around have achieved it. This produces much better results, and also should now work with both two digits and three digit brokers. It's available as an option, enabled by default, here's the difference - more trades, better recovery rate, better profit factor, same drawdown:
Summary
Final thoughts. To begin my mission was to group strats together to diversify as much as possible.
Then it was to filter out the bad strategies or make as many safer / as safe as I could.
Then it was to come up with newer / safer ideas and to leverage AI.
Then I wanted to create a global filter so we didn't have to monitor each strategy and decide when to turn them on or off.
And finally to tidy everything up and address any doubts I had about any of the strats. I think with this update 1.42 I've crossed the line.
A lot of effort went into this. Hopefully we're all better off for it in the long run.
Here's my summary of the experimental section:
Bearclaw is doing well - recommended.
Julius had its first successful trade.
BitBull is an unknown still. It's made some trades - I felt it might be trading more narrowly than I thought.
Trade Hunter is due for a bounce (at its historical DD limit so keep an eye on it)
Yen Buster should be off until USD/GBP/EUR start to rise against the Yen again. EURJPY is my least favourite so disabled by default, as is the strategy itself.
Ultimate Bot v1.41 Release Notes
Ultimate Bot v1.41 Adds A NEW GLOBAL VOLATILITY FILTER, two new experimental strategies, and an ADX filter to Pacman.
**Global Volatility Filter!**
Great care has been taken to find a global filter that improves the statistical reliability of each strategy but only based on one value. This minimises the need to have customised settings based on the current state of the market, so strategies can be left on regardless of what the market is doing.
Emulating the VIX index, the Volatility of the XAUUSD pair is analysed and output as Low, Low Normal, Normal, High Normal and High. All strategies respond to this filter, if enabled, except for the slower or not recommended strategies:
-Handshake
-Secker Punch
-Loonatic
-Sharpe Gold
The idea of optimised parameters for each filter based on a global filter is an exciting development that should result in more accurate trades overall and less need to monitor strategies on an individual basis.
Julius Strategy
A small new EURUSD strategy found working using a new technique of AI-engine driven data analysis. A simple strategy with reasonable statistics.
BitBull Strategy
A strategy for trading Bitcoin (BTCUSD) adapted from the BearClaw strategy, trades Buys instead of Sells, working from 2018.
*WARNING* Bitcoin is a highly volatile instrument. BitBull is based on the ATR which incorporates this volatility. It uses its own risk parameters (not the global Risk Factor setting). In the input section find the parameter 'BitBull risk level (1-3)'. It's recommended to backtest with your broker and on demo accounts first to make sure operation is as intended. It's recommended to leave on the default setting, level 1 (0.01 lots per 10,000). For Yen and AUD deposits this parameter may need to be modified (Manual levels like 0.5 or lower may be needed).
ADX Pacman Filter
This is an exit filter which closes open Pacman trades if a strong ADX counter-trend is identified. A few others were tested but this came out best.
FAILED TESTS
Backtesting News was successfully unlocked and tested against King and Pacman, with no effect. Potential for news-based strategies is however now available, and this will be explored in future. Based on the lack of success as a filter for King and Pacman, the XAUUSD Global Volatility Filter was created.
Ultimate Bot v1.4 Release Notes
Ultimate Bot v1.4 marks a big update to the Ultimate Bot, with difficult months in July, August, September and October. Here's an overview of the changes.
Strategy list re-ordered
New order is now my most to least favourite from top to bottom
Goal of safety
Focus on moving away from grid strategies and more on 1:1 risk-reward or better, as seen with AlgoRebel, Yen Buster and Trade Hunter. 15:1 RR strategies are Secker Punch and Pacman. Higher max risk-reward ~100:1 are Loonatic, Quantum King, Handshake and Sharpe Gold.
AlgoRebel moved into main section
After a few months of live testing - the strategy is performing as expected and so has been taken out of the experimental section and moved to the main section
Yen Buster updated
Yen Buster has had its stats improved. It previously worked on the m15 and H1 chart, now it only uses the m15 chart, which improves its statistics. It also had the quiet mode removed - it was stuck in a loop after terminal restarts where it only traded in 0.01 lots. Yen Buster is still in the experimental section, as the future actions of the BOJ is yet to be seen.
Yen Buster should continue to work until there is a ground-breaking or major stimulus to the Japanese economy. With problems with its aging population - I don't see that happening soon overall - although the BOJ may take controversial action in future. USD, EUR and GBP have never been higher against the Yen, and this could trigger the BOJ to initiate a policy-based macro reversal.
Until price for USD, EUR and GBP looks to comfortably continue to rise against the Yen, this strategy will remain in the experimental section.
Order Comments addition
Order Comments now have initials after them to show which strategy has made a trade. This helps identify strategies which use overlapping pairs when trading. e.g. Order_Comment-'SP' = Secker Punch, Order_Comment-'QK' = Quantum King.
New Experimental Strategy 'Trade Hunter'
Trade Hunter uses three pairs, EURUSD, USDJPY and EURAUD. These are labelled in the Trade Hunter Pair filter as E, U and A respectively. Each pair filter option has a 1 after it (no significance). So EUA1 is all three pairs, E1 is just EURUSD, and UA1 is USDJPY and EURAUD, for example. All have been tested from 2015.
The EURUSD and USDJPY pairs use the Market Facilitation Index for entries, and EURAUD is really a separate strategy using multi-timeframe Stochastic Oscillators to predict reversals. They've been put together into one strategy for organisation purposes.
PLEASE NOTE: EURUSD and USDJPY use tick volumes for entries tested on ICMarkets - so results will vary between brokers and with some brokers these pairs will not work as intended. Please backtest these with your broker (EU1 pair filter).
SECOND NOTE: I honestly don't know why the EURAUD pair works. I know the principle of entries, using Stochastic slow downs to signal reversals, but why it works on EURAUD, on the days it does, I don't know (from a fundamental perspective). Jim Simons would say this doesn't matter, but perhaps it would be best to gather end of year data before running live. It's high was in February 2024 and hasn't got back to this peak yet 9 months later.
New Experimental Strategy 'BearClaw' added
Simple but effective strategy found working across multiple pairs, using RSI crossovers and ATR, opening Sells only. Trading EURAUD, GBPJPY, XAUUSD, GBPNZD and EURUSD, this represents a huge boost in diversification as it trades trends and volatility, the opposite of a mean-reversion grid, as well as less often used pairs. Each optimised pair can be selected individually, with a huge performance shift when combined with other strategies.
Trend strategies are difficult to find so finding one that works well with many pairs is a real boost. It is more prone to down periods as it's not a grid, it opens on many false signals, but in the right conditions the gains are very fast in a very short period of time, and handles dynamic volatility. An exciting new approach to balance against the other strategies.
New 'Not Recommended' category
Enough experimentation and new strategies have been added to created a 'Not recommended' Category. Users can still choose to use them if desired, but Loonatic and Sharpe Gold are placed there as they have the weakest Risk-Reward profiles.
Custom XAUUSD name
Users have requested option for custom XAUUSD name for testing and trading like GOLD. You can enter this in the bottom box of the input section.
TLDR Recommendations:
- No longer recommend running Loonatic and Pacman together except on lower risk. Preference to run PacMan over Loonatic
- Yen Buster remains in Experimental section. For use when BOJ is not offering major stimulus to Japan
- Trade Hunter in experimental section. Test on your own broker (tick volumes), recommend waiting until more data is in on EURAUD Stochastic strategy. In the meantime if you want to use it and your backtest works well with the EURUSD and USDJPY strategies I recommend leaving the Trade Hunter pair filter to the default 'EU1' (EURUSD|USDJPY).
- BearClaw added which should work with all brokers, simple but effective strategy with excellent RR profile across many less used pairs, trading Sells only on EURAUD, GBPJPY, XAUUSD, GBPNZD and EURUSD.