When importing timeseries data from yahoo finance using pandas data reader package both "close" and "adj close" columns are returned. What is the difference between these two? And which should be used?
Hello Srungeer,
So, there are times when a company might merge its stocks, for example, merge two shares worth INR 50 each to one share worth INR 100. This is called a corporate action and there might be a variety of reasons for this.
Now, imagine a timeseries reflecting such corporate action. One on the day the price will be INR 50 and on the immediate next day, it will be INR 100. So there is some kind of structural or value break in the data. The share prices hasn't gone up INR 50. This might cause issues while modelling. It shouldn't be put across to the model that there was a 100% increase in price.
So, to go around this issue we "adjust" the raw closing prices to account for such merger/splits etc of shares. In this case, the adjustment will be dividing all prices post the merger by 2 ( because we merged to shares ). So, the closing price that became INR 100 from INR 50 will become INR 50 after adjustment.
Do get back if more explanation is needed.