ka66: Bar Range Bands

8월 23일

This tool takes a bar's range, and reflects it above the high and below the low of that bar, drawing upper and lower bands around the bar. Repeated for each bar. There's an option to then multiply that range by some multiple. Use a value greater than 1 to get wider bands, and less than one to get narrower bands.

This tool stems out of my frustration from the use of dynamic bands (like Keltner Channels, or Bollinger Bands), in particular for estimating take profit points.

Dynamic bands work great for entries and stop loss, but their dynamism is less useful for a future event like taking profit, in my experience. We can use a smaller multiple, but then we can often lose out on a bigger chunk of gains unnecessarily.

The inspiration for this came from a friend explaining an ICT/SMC concept around estimating the magnitude of a trend, by calculating the Asian Session Range, and reflecting it above or below on to the New York and London sessions. He described this as standard deviation of the Asian Range, where the range can thus be multiplied by some multiple for a wider or narrower deviation.

This, in turn, also reminded me of the Measured Move concept in Technical Analysis. We then consider that the market is fractal in nature, and this is why patterns persist in most timeframes. Traders exist across the spectrum of timeframes. Thus, a single bar on a timeframe, is made up of multiple bars on a lower timeframe. In other words, when we reflect a bar's range above or below itself, in the event that in a lower timeframe, that bar fit a pattern whose take profit target could be estimated via a Measured Move, then the band's value becomes a more valid estimate of a take profit point.

Yet another way to think about it, by way of the fractal nature above, is that it is essentially a simplified dynamic support and resistance mechanism, even simpler than say the various Pivot calculations (e.g. Classical, Camarilla, etc.).

This tool in general, can also be used by those who manually backtest setups (and certainly can be used in an automated setting too!). It is a research tool in that regard, applicable to various setups.

One of the pitfalls of manual backtesting is that it requires more discipline to really determine an exit point, because it's easy to say "oh, I'll know more or less where to exit when I go live, I just want to see that the entry tends to work". From experience, this is a bad idea, because our mind subconsciously knows that we haven't got a trained reflex on where to exit. The setup may be decent, but without an exit point, we will never have truly embraced and internalised trading it. Again, I speak from experience!

Thus, to use this to research take profit/exit points:

1. Have a setup in mind, with all the entry rules.
   
2. Plot your setup's indicators, mark your signals.
   
3. Use this indicator to get an idea of where to exit after taking an entry based on your signal.
   

Credits:

• ICT_ID for providing the idea of using ranges to estimate how far a trend move might go, in particular he used the Asian Range projected on to the London and New York market sessions.
  
• All the technicians who came up with the idea of the Measured Move.
  

10월 25일

릴리즈 노트

The following updates are implemented:

• Allow ATR-based bands, in addition to the bar range-based bands.
  
• Just use the (Settings -> Style) tab of the indicator to decide which bands should be displayed. In this chart, we have disabled (unchecked) the bar range-based bands, and only kept the ATR-based bands enabled/visible.
  

As mentioned during the initial publication of this tool, it is a research tool. A very concrete way that I use it is to estimate TP and SL levels for any set-up I am researching.

Then, we can add these bands and generally keep them hidden, turn them only when our setup is there, and we need guidance around the TP and SL as per our system's rules. This is especially handy for rules-based scalping.

Were I to name this script now, I would call it something catchier like TP/SL Estimator. But that may limit what you use the bands for. TP/SL estimation is just one use case that I have developed this for.