The ‘Heracles’ strategy – Repost (Patreon)

The Heracles strategy

This time I will be testing if the ‘Heracles’ strategy is profitable or not.

You can see the complete methodology that I use in the overview on the screen now.

You can use this as an  inspiration to develop your own methodology for testing and comparing trading strategies between each other. I urge you to always do your own research after watching these video’s.

Finally, the best results of all the tested strategies will be compared to each other in a total league. This way we can find out which strategy is has the highest probability to make money with the help of a trading bot.

You can get his strategy by visiting the freqtrade site.

• Navigate to the github repository, then click on the main github page.
• You will see a repo that’s called frektrade strategies.
• In this repository you will find a folder called user data.
• Then go to the strategies folder.
• In there, you will find the heracles strategy.

This strategy is originally created by Massoud Azizi. And if you like what he does, then let him know by paying him a visit on his github page.

To download this exact strategy to your computer, follow the commands you see on the screen. Do this from within the user data strategies folder otherwise freqtrade will not recognise the file.

wget https://raw.githubusercontent.com/freqtrade/freqtrade-strategies/master/user_data/strategies/Heracles.py

INITIAL BACKTEST RESULTS ON MULTIPLE TIMEFRAMES

The initial backtest of this strategy on multiple timeframes have the following results. As you can see the 4hour and 30 minute timeframes have the same score overall. The 30 minute timeframe has a higher winrate and bigger profits, but the 4hour has less trades to get almost similar results and an important factor, the drawdown is much lower. You wouldn’t want to experience the stress of seeing your profits dissapear.

Also, the author of this strategy preferred the 4 hour timeframe by default and who am I to doubt the expert. That is why I also stick to this timeframe.

So, I therefore will optimize the parameters on the 4 hour timeframe and backtest again.

HYPEROPTIMIZED PARAMETERS BACKTEST RESULTS

As you can see, Hyperparameter optimisation does not always result in the best parameter values.

If we focus mainly on profits and do not take risk and drawdown into consideration. Let alone the applicability of these parameter values on multiple asset pairs, then we have the most optimized strategy here.

But since all the other factors also take part in a successful trading strategy, this time optimisation of the original parameter values prove to give lesser results.

The most noticeable thing I have here is that these optimised parameters only are applicable to 11 pairs in the past. These pairs indeed make the highest gains, but you have to take into consideration that we have to have a strategy that is usable in the future and on the most pairs we want to trade as well.

How do we know which pairs will make use of this optimised strategy? We do not know. That is why I also always test to see if the strategy is usable on most of the asset pairs we have configured in out trading bot. Otherwise it will lead to opportunity costs.

Another thing is the high drawdown these parameters caused. I’d like to keep my gains and do not really want to see that my profits melt away because of drawdown. Again, we do not know how the performance will be in the future. It can easily start with causing a drawdown on your account before you even get profitable trades. Do you have the nerves to sit that out? I sure don’t…

So, to conclude this…

I will use the original strategy results, based with its original parameter values and put this in the strategy league to see at which place this will enter…

STRATEGY LEAGUE

So, where does this strategy enter the overall league?

Unfortunately for this strategy, it has not such good results and therefore it enters on the 21 position in the total list.

Just above the swing high to sky strategy, which is last.

That last strategy has much larger profits and a higher winrate. But as I said, that strategy only is applicable on 28% (that is 14 pairs of 50). Which I find too specific. We try to find a Nice generic strategy that will suit all pairs and give high profits without having to take too much risk and experience huge amounts of drawdown.

And both of them have too many issues to get higher than their current respective places.