This is my personal journey. Yours might vary, and even if you follow my steps, results might vary. And important disclaimer, none of the strats linked here are to be used for live trading. They are all for learning purpose only.
At the start, I tested some strategies. In the end, I chose this strat as the base. Over years, the strategy evolved quite a lot. There was this, this, and this among others. I tested a lot of different ideas and approaches. I looked at other strategies for inspirations. Are there new indicators I can use? Are there new logics/approaches?
This was the first form of my current stable strategy that was created after 2 years of testing things. Let me repeat, 2 years. If you look at the few last lines of my populate_indicators (dataframe[‘vol_base’] and below), I took that from another strategy (forgot which one). Others were due to my own experiments. For example, that rsi_1 and others that are missing from the file, is actually I split RSI into 4 zones. Using an overfitted RSI check, for example dataframe[‘rsi’] < 16, never works well in long-term. There are more new indicators added on later versions from similar sources, either from other people’s or from my own random creations.
As said above, it took 2 years and multiple failures to reach the basis of my stable strategy. And even after that, the strategy still developed continuously until now (2+ years and counting). Longer knowledges and experiences I gained over these past 4 years (and more) are (and will still be) written into multiple posts here. This is just a summary of my journey, and what you can expect. The journey won’t be short and easy. You will need to read all the posts here to have more insights.
If you are new to this journey, or you think you have found multiple dead-ends, I have these words for you
- Indicators aren’t everything. There is no such thing as the best indicator(s). Test anything you can find out there, make sure to use several different types of them.
- Pair selection is more important than indicator.
- Make sure you know the direction of your strategy.
- Know which stats tells better story about your strategy’s performance
- Spend times to look at charts. Try to create some custom indicators you can think of. Maybe you will notice some patterns/behaviors. Think what you would do at certain conditions if you are trading manually. See how it can be translated into codes.
- Failures are expected. Even when I (accidentally) created the stable strategy, I already expect it to fail as well like others before it. Don’t be discourages by it. Look at it as result of experiments. At least now you know that approach isn’t working. Move on to new approach/idea. Failure is our friend. Just make sure you learn something new from each failures and try not to repeat those mistakes. As wise man used to say, not even donkey fall on same hole twice. We have to be smarter than donkey.
[…] How I created my first strategy […]