I was looking at Alpaca and TradingView’s overnight bars (BOATS) for BE at the same point in time. Alpaca’s bars are overwhelmingly sparse. What’s going on?
I suspect that Alpaca’s overnight bars are much sparser because it filters certain BOATS trades—such as odd-lot and other non-price-forming trades—out of OHLC calculations. If a minute contains only filtered trades, Alpaca may produce no candle at all. TradingView, using BOATS data delivered through ICE Data Services, appears to include more of those overnight trades when constructing its bars.
@Richard_Edwards You are entirely correct. Alpaca excludes certain trades from bar calculations. Those excluded trade conditions are listed in the documentation here and reflect the SIP guidelines on how to calculate bars. One of those conditions is condition I (a trade under a full lot). Many of the BOATS trades are small trades and therefore are excluded in bar calculations. If there are no valid trades during a bar, than no bar is created.
That is probably what you are seeing?
1 Like