So I have been running a bot for a few days (Day trading 1 min intervals) and on inspection of the market data provided by Alpaca I have noticed jumping time series.
In the first image I have 1 min intervals set and am hovering at time 20:24, the next available stick starts at 20:39 and not 20:25 , I thought maybe I was seeing shit or something but surely this cannot be q correct chart for my selected time series?
Image 1 (20:24): https://gyazo.com/b18fd1a414736c99f9416a83bad7791c
Image 2 (20:39) : https://gyazo.com/3272debc12f9b11b4fc487ce6311aa70
I dont know this was due to lack of market data or what but other brokerages provude correct 1 minute time series graphs with no missing data. Anyone able to shed some light?
I have observed the same problem.
For low volume stocks, if there was not any transaction in a particular minute, then it seems that Poly will skip and not send out a “dummy” bar. This, I take care internally in my program.
However, it happened even for high volume stocks. In one case, it sent a 1 min bar timestamped next day at a random time, wrecked havoc in my patching algorithm. I thought it could be a problem with my software configuration, although I have a 1GB fiber network connection. But then I was listening to several high volume stocks, and it happened in only 2 of them.
To support the hypothesis that Poly skips some real time 1 min bars is that there are occasional gaps in historical intraday data, called by “api.polygon.historic_agg_v2”.
Based on my experience I would not trust Poly data to do frequent intraday trading. I am looking for other paid sources.
1 Like