Assume a model like this, basically explaining stock market returns with a bunch of stuff:
stockReturn(t) ~ bondReturn(t) + moneyMarketReturn(t) + inflation(t) + somethingElse(t)
Does using inflation as an independent variable bring any significant problems?
Should the results differ from using real returns instead, in this case using the inflation to calculate real stock, bond, and money market returns (assuming that the somethingElse is a non-economic variable). What would affect the difference?