Time-fixed effects and macroeconomic control variables

[Momentum]

Hi there,

For my master thesis I am testing (stock) momentum returns under different asset pricing models, which basically comes down to calculate whether returns generated from a momentum strategy are more than zero. That is what I tested, and my hypotheses were confirmed. However, my professor said he would like to see some Eviews regressions on the effect of cumulative past-12 months returns (independent variable) on next month stock return (dependent variable). Also, he said that I should not forget to include time- and firm-fixed effects and macroeconomic variables.

My problem is as follows. As soon as I import macroeconomic variables, these variables don't vary across the cross-section anymore, making time-fixed effects impossible (near singular matrix). I uploaded my excel for Fama-French corrected returns as an example. (return12m is accumulated past-12 months return, posdum12 is a dummy whether returns were positive or negative past 12 month, return1m is my dependent variable, DIV, TERM, DEF and TB are my macroeconomic control variables (if you filter for one month, you'll see that the values for these variables are the same for every stock), Mstate and Rstate can be ignored, and i_div, i_term, i_def and i_tb are my interaction terms).

The regression I am trying to run is then this one: Estimate equation -> return1m c posdum12 DIV TERM DEF TB i_div i_term i_def i_tb , with both firm- as time-fixed effects. However, I must choose: either macro variables but no time-fixed effects, or no macro variables (omitted variable bias) but time fixed effects. I'm afraid both will result in regression results which basically don't tell me anything.

My questions:
- do I import my data correctly?
- how could I run this regression with BOTH time-fixed effects AS macroeconomic variables?

Thanks a lot!