For a stochastic economic analysis if we have 2017 values for most of variables but one or two variables have not been updated since 2015, is it OK to directly use them or should we convert them to 2017 values (if we have a proper indices to do so)?

Or another possible situation that we don't have access to a proper index to convert 2015 value to 2017. Is this OK to leave all values as they are (using most updated values for each variable) or we should use 2015 values for all variables to have a year zero in the analysis.

My initial guess is that it doesn't matter that much. Since most of the times we already have variables with good amount of variations and updating or not updating one variable won't affect the outcome of simulation (that much).