while underlying prices in Zorro-provided *.t8 file are correct (and are taken at around 16:00 ET, end of RTH), 1min bars in *.t6 files are not.
Pls have a look at 2 screenshots. One is Zorro-provided *.t8 vs Zorro *.t6 file, and the other is Zorro's *.t6 vs IB's data (which matches underlying prices in the *.t8 file).
For 2020, SPY prices in t8 and t6 files are close, but not for earlier years. Basing strike/ vol calculations on such prices is erroneous.
But even at the end of 2019, when prices in t6 are close to correct, they are shifted by 1min (i.e. reflect start of the bar).
I am, but until today thought that the adjusted prices are "forward adjusted" and represent Total Return to an investor ( and therefore adjusted prices are bigger that non-adjusted, hence my question)
As I found many vendors use "back-adjustment" (similar to continuous futures), with some deducting absolute dividend amount to adjust and some - correctly - ratio-adjust (like Quandl).
Are Zorro-supplied data ratio-adjusted?
Forward adjustment more correctly represents returns and the data do not need to be rerun with each dividend.
When you write an options trading system, you should know about adjusted and unadjusted prices because you'll often need both. You can look up on many websites how they are calculated. Basically you multiply the close with a factor (1 - dividend/ex-price). I don't think that companies pay negative dividends, so you can see that adjusted prices are always smaller than unadjusted prices, never bigger.
