To establish a base line, you could look into Some Patterns of PhDs in Mathematics Awarded Annually by Institutions of Higher Education in the United States over the Last Two Decades. This lists for each subfield of mathematics how many Ph.D.'s were awarded over a 10 or 20 year period by each major university in the US. You could then normalize that with the faculty of each department to obtain an upper bound on what you might reasonably expect as Ph.D. output per professor.
For example, in number theory a typical math department will award on average three Ph.D's every two years (table 11). These numbers are about the same in other pure math fields. I would expect the outcome of this exercise to be that tenured faculty in pure math delivers on average somewhat less than 1 Ph.D. per year.