Hi all,

I developed some strategies in Zorro, some of which use the ADX indicator. I then converted a couple of the strategies to Python using ta library for paper trading. However, I realised that the ADX indicator in ta library provides different numbers to Zorro even when using the same dataset meaning I cannot replicate the buy and sell signals. I also tried making my own ADX figure without using ta library but this matched the ta library figure and not the Zorro figure. I also checked tradingview and this also seems to match ta library and not the Zorro figure.

An example below for ADX 14 on SPY daily data:
Zorro Others
[193: Tue 25-05-20] 21.64 20.72
[194: Wed 25-05-21] 21.74 20.24
[195: Thu 25-05-22] 21.19 19.73
[196: Fri 25-05-23] 19.80 18.55

I had a look through the documentation but I can't find an obvious explanation. The only thing I can think of is the smoothing method used to create the PlusDI figure may be different:

"PlusDI: Plus Directional Indicator, a part of the DX indicator, ratio of a smoothed PlusDM to ATR in percent. In the first version the current asset price series is used."

When I created the indicator manually I used the formula described here:

https://chartschool.stockcharts.com...indicators/average-directional-index-adx

OR because I am a noob I am doing something completely wrong. In an ideal world I want to recreate the Zorro ADX figure in python because I have done all the backtesting in Zorro.

Any help which explains the difference appreciated!

An update to say I gave up on this one... I did find in the documentation that many of the indicators Zorro uses are based on TA-Lib, however, I found the TA-Lib ADX figure in Python is different to the Zorro one. In the end I manually made an ADX figure in Zorro and this one seemed to match other platforms and then ran backtests again. Not sure if was doing something wrong (more likely) or it is a bug (less likely)!