The Bollinger Bands Trend Indicator strategy (Patreon)

Introduction

In this post I will show you how you might get to be a profitable swing trader with the Bollinger Trend Indicator. I created 20 variants of a simple Bollinger Bands strategy and will show you which one has the best results under my backtest circumstances. I’ll also show you how you can apply this indicator on your Tradingview chart and show you the alternative indicator I created for better entry and exit signals too. 

So let’s gets started right away…

The Indicator

For this video I was recently triggered by a Tradingview feature notification, where the Bollinger Trend indicator was presented and explained. 

As many of you know, the Bollinger Bands is a very popular indicator in technical analysis and measures market volatility. 

According to the Tradingview information page, this BBTrend indicator advances traditional Bollinger Bands by providing a clearer view of market trends. 

The indicator is derived from two separate sets of Bollinger Bands — one short and one long — and displays the results in a histogram format. Values above zero indicate a bullish trend, while values below zero signify bearish conditions. The distance from zero reflects the strength of the trend.

I recommend you to watch the Tradingview Youtube video for more in depth information about this indicator. I’ll put the link to that video in the description below. https://www.youtube.com/watch?v=aG8UIRNzPJQ

The Strategy

For the development of strategies that are based on the Bollinger Band Trend indicator I searched for the internet and found the following website: https://trendspider.com/blog/bbtrend-a-new-indicator-from-john-bollinger/

This website proposes two different types of strategies, one based on momentum and one based on a ‘flip’ where the indicator goes from below to above zero or vice versa. 

Now I could develop a strategy that was based exactly on these parameters, but since I visually checked the indicator on Tradinview and found that it was already clearly lagging when detecting a trend. I decided to just go with an initial basic strategy setup. Where the entry is based on the flip from negative to positive for bullish entries and vice versa for bearish setups. 

From there I developed multiple variations of this base strategy to see which one would have the highest probabilities for profitable trading. In total I created 10 different versions based on the initial simple crossover strategy. Some of these are also based on better exit signals from another simple to use indicator, namely the MACD. 

If you have seen my earlier videos, you probably know that most of the public available strategies mostly focus on spot trading. So in this case I also created versions that can do futures trading. So in those cases trade entries can go long and short.
In total this resulted in 20 variations of this basic strategy which I have tested.

These variations are:

Bollinger Trend

Enter when the Bollinger trend crosses above the zero line, Exit when crossed below. 

Enter when the Bollinger trend crosses below the zero line, exit when crossed above (futures).

Enter as long when the Bollinger trend is above the zero line, exit - or no long trade -  when is below.

Enter as long when the Bollinger trend is below the zero line, exit – or no trade – when is above (futures).

One of the most noticable differences here is that one of these versions only trade on the crossover or crossdown of the trend indicator on the 0-line.

And the other version trades as long as the Trend indicator is above or below the 0-line. Which means that you do not have to wait for the exact cross moment but can enter as long as the trend indicates bullish or bearish momentum.

With MACD

Enter when the Bollinger trend crosses above the zero line, and MACD is above its MACD signal line, Exit when the MACD is below the MACD signal line.

Enter when the Bollinger trend crosses below the zero line, and MACD is below its MACD signal line, Exit when the MACD is above the MACD signal line (futures).

Enter as long when the Bollinger trend is above the zero line, and MACD is above its MACD signal line, Exit when the MACD is below the MACD signal line.

Enter as long when the Bollinger trend is below the zero line, and MACD is below its MACD signal line, Exit when the MACD is above the MACD signal line (futures).

I also developed my own version of the Trend indicator, since the original Tradingview indicator did not have the option to set the standard deviation level for both Bollinger Bands separately. 

By manipulating the separate Bands I could manage to make the trend indicator respond faster to trends. I also added a moving average for the detection of additional exit signals.

https://www.tradingview.com/script/rI3zj2j6-Bollinger-Bands-Fast-Trend-Indicator-DCD/

Based on this I also created the following trading signals.

Enter when the Bollinger fast trend crosses above the zero line, Exit when Bollinger Trend crosses below the zero line.

Enter when the Bollinger fast trend crosses below the zero line, exit when Bollinger Trend crosses  above the zero line.

Enter as long when the Bollinger fast trend is above the zero line, Exit when Bollinger Trend is below the zero line.

Enter as long when the Bollinger fast trend is below the zero line, exit when Bollinger Trend is above the zero line.

Enter when the Bollinger fast trend crosses above the zero line, Exit when Bollinger Trend gets below its Moving Average line

Enter when the Bollinger fast trend crosses below the zero line, exit when Bollinger Trend gets above its Moving Average line

Enter as long when the Bollinger fast trend is above the zero line, Exit when Bollinger Trend gets below its Moving Average line

Enter as long when the Bollinger fast trend is below the zero line, exit when Bollinger Trend gets above its Moving Average line

Enter when the Bollinger fast trend crosses above the zero line, and MACD is above its MACD signal line, Exit when the MACD is below the MACD signal line.

Enter when the Bollinger fast trend crosses below the zero line, and MACD is below its MACD signal line, Exit when the MACD is above the MACD signal line.

So with all these different versions programmed into algorithmic trading code I could easily test all these out on multiple crypto pairs, multiple timeframes and on large datasets. 

Saving me loads of time for analyzing the results. This way I can provide you with a grand overview on which version of this Bollinger Band trading strategy will have the highest probabilities for the future to provide you with winning trades.

But always do your own research first before using this in real trading. Because my results might not reflect the performance on your preferred setup and trading pairs.

Backtest Results

Now I will not bore you with 20 different results that takes 50 minutes of your time. So this time I’ll stick to, what I think are the most remarkable results of this strategy test.

Another thing I’d like to add before I present you with these results is that this strategy is meant to be a manual trading strategy on higher timeframes. So it also contains the risk profile that mostly comes with these kind of swing trading strategies.

The highest gains (BollingerTrendMACD_is_s)

The version with the Highest gains in this test is on the one day timeframe. With the Bollinger Trend indicator with MACD exit signal. Only on spot trading, and where entries can be made as long as the trend indicator is above the 0-line. 

So here it’s not mandatory moment to wait for a crossover. Enter as long as the bollinger trend indicator is above 0 and there is no MACD exit signal.

It’s equity curve is quite jagged. But within the backtest dataset it is able to catch the high gainers quite well by looking at this curve. These moments are real steep and you can expect high gains during these explosive bullish periods.

However the Drawdowns can also be quite severe and you should take this into consideration as well if you are planning to only use these entry and exit signals.

And comparing its performance indicators with other better scoring strategies it does have quite a lot to improve. For example with the low profit percentage, win percentage ,calmar, sharpe and more. But be aware that this comparing apples with pears at the moment since this is a swing trading strategy on high timeframes for manual trading. And the other well performing algorithms are mostly for low timeframes, have high complexity and are not suited for manual trading.

Nevertheless, if you still are willing to accept this risk and you are certain that, with extra rules, you can improve this performance. Then I think it is a good baseline.

The lowest Drawdown (BollingerTrendMACD_cross_s - 1day TF)

However If you are not willing to give back your attained gains back to the market again and are mainly focusing on low drawdowns. Then you might consider the other Bollinger Trend strategy where you only trade spot markets on the 1 day timeframe, enter only on the crossovers above the 0-line ad use MACD exit signals. 

Coincidentally this is almost the exact strategy as what I showed you previous. However, the important additional rule here is that you only may enter after the crossup moment. So only after the moment where the Bollinger Trend indicator crosses up.

Since there are no entries during the Bollinger Trend above the 0-line, there are also a lot less moments where the actual cross happens. Which results in a lower amount of trades. If you don’t mind this and like the fact that you only trade at crosses then this version might be more for you.

The equity curve might not be as exiting as the previous version. But overall to me it looks to have a  lot more stable incline.

This is also visible in the drawdown curve where the drawdown on average is only 6.7 percent. 

The final result is also almost not comparable with the other automatic trading algorithms results.

But still a 1200 percent gain in my backtest is certainly no shame. There are loads of algo’s that perform way worse then this. Therefore, people that require less risk might consider starting with this manual strategy as a baseline and develop their additional rules further here.

Best Winning percentage (Bbfasttrend_is_f)

Another indicator of a successful trading strategy, that a lot of traders like to know, is the win rate or winning percentage. 

Now, a high win rate does not always say much because you can have loads of small wins, only to bust your trading account with a handful of losers. 

But for those people that want to know this anyway, the strategy variant with the highest win percentage uses the adjusted Bollinger trend on futures trading where trades can be made as long as the indicator is above for longs and below for shorts. Exits for both directions are when the indicator reverses to the other side of the 0-line. 

On this equity curve plot you can also see that it is an even match between winning and losing trades as well. 

The drawdown curve here shows that the average drawdown of 7.5 percent with some exceptions that peak to over 20 percent. 

The end result with over 5 and a half thousand percent in my backtest situation shows that this manual trading strategy, based on the adjustments of the original indicator is not only user friendly, but also has the probability to make good profits as well. 

This could also be an excellent baseline strategy to build on, for those people that want to see regular wins and also a balanced equity curve too. Albeit with some higher drawdowns then the previous discussed version.

Overall best Bollinger Trend Strategy (BBFastTrendMACD_is_f)

And this brings me to the last, and best scoring version I want to show. 

Which Bollinger Trend Indicator version has the best balance between profits, winning trades, drawdown, but also applicability to trading pairs and good performance indicators like Sharpe, Sortino and Calmar ratio.

And according to my backtests, on my specific setup, with my preferred performance criteria, 

the adjusted Bollinger band trend indicator where the MACD is also used for additional confirmation and exit signals,  on the 1 day timeframe, AND with Futures trading, where you can go Long and Short as long as the trend indicator is above or below the 0-line, has the best score overall.

As a side note, the indicators all use their default settings too, just as with all the variants. So the Fast trend indicator has the following settings and the MACD also has its default settings too. 

I reckon that performances could be still be optimized if these settings were adjusted, but this would make the investigation too complex for this video and also prone to curve fitting. 

It can also be a good homework assignment for those viewers that want  to investigate the possibilities further. And I invite you too add your findings or remarks in the comments section below.

Now what about the hypothetical performance of this trading strategy. 

The equity curve clearly shows a nice 45 degree angle when the last bull market started. And you can clearly see the moments where the drawdown starts to kick in during the pullbacks in the real market.

I think that here lies the opportunity for people that want to build on this strategy further. To improve the detection of bearish circumstances so to prevent these drawdowns.

Especially the second drawdown has a lot of influence on the performance of this algorithm, but notice that all previous drawdown moments did not get beyond the average 8 percent here. So if these last two major drawdowns did not occur, then this strategy would perform even better.

The following box plots tell something more about the ranges in which the individual trades exist.

The winrate distribution shows a median winrate around 35%, indicating that in most cases, the strategy wins about 35% of the trades. The interquartile range (IQR) spans from approximately 0% to 55%, suggesting moderate variability in the winrates of the trades. And finally here the upper whisker extends up to 100%, showing that some periods have a perfect winrate, while the lower whisker approaches 0%, indicating some periods with very low winrates.

The profit distribution plot shows a median profit close to zero, suggesting that overall, the strategy neither gains nor loses much on average. The IQR for profits is relatively tight around zero, indicating that the majority of trades result in small profits or losses.
There are significant outliers both on the positive and negative sides, with some trades resulting in very high profits up to 15000 dollar and also significant losses down to -5000 dollar. So this means that the strategy can have large gains but also substantial risks here.

Overall, this version has the most balanced score on all its performance indicators.

It is still is not a close match for these automatic trading strategies that I tested so far. But for a manual trading strategy I think these scores are not bad at all. 

With some additional trading rules like setting manual stop losses, take profit points. Maybe use an additional strength, or volume indicator it could turn into an even better manual trading strategy here. 

As I stated earlier, it will not beat an algorithmic trading strategy that is based on numerous entry and exit points. But I can tell you that at the moment this trading strategy is the best setup for manual trading on the 1 day timeframe. Where you can go long and short, based on the signals you get.

The only direct competition on these higher timeframes is the Get Parabolic Trading strategy on the 4 hour timeframe. 

And it could well be that combining these two trading strategies will improve the probability for good trades even better. But that is for you, the viewer to find out. As a second homework assignment 😊

Ending

And with this conclusion I am at the end of this Post.

All the backtest output, plots, the files and backtesting code in which I programmed this trading strategy for automatic backtesting can be found as attachments to this post.

The Bollinger Band Fast Trend Indicator can be found on Tradingview with this link: https://www.tradingview.com/script/rI3zj2j6-Bollinger-Bands-Fast-Trend-Indicator-DCD/

And for your conveinienve I also programmed the Bollinger Trend indicator and Fast trend indicator in a Python module to that you can use this easily in your code. The module can be found here: https://pypi.org/project/bamboo-ta/

And everything is also on my Github page too at this location: https://github.com/DutchCryptoDad/bamboo-ta

So like, subscribe and comment if you like to and I’ll see you again in my next post…

Goodbye!!