all at NY time
trades value of 11.18.24
trades value at 11.19.24 03:03 AM is 437782
trades value at 11.19.24 03:31 AM is 437049
trades value at 11.20.24 03:26 AM is 437779
In simple words, the trades value of daily bar of 11.18.24 was edited twice after the bar was already closed. nice
Hi
Yes, Alpaca does alter past bars’ volume. The reason for this is canceled and corrected trades. Trades can be reported canceled or corrected a long time (possibly even days) after the bar’s end. So we regenerate the bars for a few days. This happens at around 3 AM New York time.
This can explain your experience. You can investigate these trades using the historical trades endpoint, or if you want, I can do it for you if you provide the symbol.
Thanks for the reply, you seem like a smart man, please explain to me how can I make this with alpaca:
- Fitting on all of the past data on a ticker until today (ohlcvt).
- Through each day on the future, transform the daily data grabbed today (after market closes) and make prediction for the future based on the data.
The problem here, currently, with alpaca, is that trades and volume data of the past is inconsistent with data grabbed lately. so one cannot perform this scenario on alpaca data.
Do you understand the problem here?