Price inconsistency between Streak backtest & Python historical data API

sohel

Hi,

I’m experimenting using the default algo example and have a few questions.

I ran the DEMA short example on ALBK between 23/8/2017 to 22/8/2018 and noticed that on 20 September 2017:

|20 - SEP - 2017, 12:15:00 PM|ALBK NSE|SELL|-100|73.00|-76.00|EN|
|20 - SEP - 2017, 02:15:00 PM|ALBK NSE|BUY|100|73.73|-149.00|SL|

This means that 100 shares were sold at 12:15pm at Rs 73 and then bought back at 2:15pm at 73.73 since the stop loss was triggered.

However, when I fetch hourly prices via the historical data Python API, I get the following for 20 September 2017:

https://imgur.com/a/jPKaxCA

Here the times are in UTC so the relevant rows are 06:45 (12:15pm IST) and 08:45 (2:15pm IST). In neither of these rows do we see Rs 73 or Rs 73.73 respectively.

So my question is why is there an inconsistency between those prices used for Streak backtest & Python historical data API? These are not open, close, high, low, or an average of all 4 - so which prices are used, exactly?

Sorry if this is in the documentation somewhere, I looked but couldn’t find anything specific.

Thanks!

rakeshr

@sohel
Streak, also uses same Historical candle APIs.Hourly OHLC value of ALBK for 2017-09-20 12:15:00 is 73.25,73.3,72.95,72.95, as above shown by you.
Algo calculation used in Streak DEMA short, might be having different derivative calculation, not sure about this.
Can you write to streak on [email protected], with above details.

sohel

Sure, I've just sent them an email.