No standard error in two-way fixed regression output

[shenflow]

I am estimating a two-way FE model as follows:

proc panel data = example;
id gvkey rankyear;
model y = readj_pl_vol readj_mi_vol
/ fixtwo noint hccme = 3 cluster;
run;
quit;

The output I get is as follows:

Why do I get a coefficient estimate for the second variable, but no standard error?

[PaigeMiller · Accepted solution]

Possibly scaling of the variables that causes arithmetic underflow