Commentary 
It takes time for confirmation. Short run, high-frequency data is inherently noisy, so to attempt drawing any reasonable conclusion you cannot rely on such data alone. One month or even several for what might at first appear to be an emerging trend can change or vanish in whatever data with no prior warning.
However, if you aren’t starting from such economic accounts, rather using them as last-in-line evidence for what’s already proposed by other means, then high-frequency data can be useful for corroboration. The more evidence that tells a consistent story, even if not fully emergent, the higher the likelihood it’s a real one and not mere randomness or statistical noise….