High52

[gliedtke-at-ub]

Hi, and thank you for making the replication materials available.

When constructing the 52-week-high variable using the provided scripts, I find that the resulting ratios can exceed 1 — sometimes by a large margin (e.g., >10). Given the definition as current price divided by the maximum price over the previous 12 months, this measure would typically fall within the [0, 1] range.

Could you clarify whether values greater than 1 are expected in your implementation?
If not, could this be related to using unadjusted prices in the construction?

Thanks in advance for your help!

[chenandrewy]

@gliedtke-at-ub : This looks suspicious. Yes, split adjustments are a pain in the ass. We'll look into it when we get the chance! Thanks for letting us know!