According to the API documentation:
'open', 'high', 'low', and 'close' return the relevant information for the current trade bar. If there is no current trade bar, NaN is returned. These fields are never forward-filled.
Does this mean that for '1d' data, I can't rely on the current open, high, low, or close? As an example, I have taken some Quantopian code from "recording and plotting variables" and just replaced the securities with IHT. Sometimes the high/low is clearly returned as NaN, but it's not consistent. If I replace IHT with AAPL, for instance, there are no NaNs. It would seem that one should never use data.current(asset, 'high') as the value is unreliable. Is my understanding correct?