Max function returning multiple values [SQL] - sql

Question

I have 3 tables: money, student, faculty. This query returns each faculty and highest stipend in each one of them.
select
f.name as "FACULTY_NAME",
max(stipend) as "MAX_STIPEND"
from
money m, student s
inner join
faculty f on f.id_faculty = s.faculty_id
where
m.student_id = s.id_student
group by
f.id_faculty, f.name;
Query works fine:
FACULTY_NAME | MAX_STIPEND
-----------------+---------------
IT Faculty | 50
Architecture | 60
Journalism | 40
However when I add s.name to original query to also show the name of the student who received max_stipend, query is not working like it used to - it returns all of the students
select
f.name as "FACULTY_NAME",s.name,
max(stipend) as "MAX_STIPEND"
from
money m, student s
inner join
faculty f on f.id_faculty = s.faculty_id
where
m.student_id = s.id_student
group by
f.id_faculty, f.name, s.name;
Query result:
FACULTY_NAME | s.name | MAX_STIPEND
----------------+-----------+---------------
IT Faculty | Joe | 50
IT Faculty | Lisa | 10
Architecture | Bob | 60
Journalism | Fred | 5
Architecture | Susan | 5
Journalism | Tom | 40
It does the same thing using right, left and inner joins. Can someone tell where the problem is?

Answer 1

First, you should be using proper JOIN syntax for all your joins.
Second, you can use Oracle's keep syntax:
select f.name as FACULTY_NAME,
max(stipend) as MAX_STIPEND,
max(s.name) keep (dense_rank first order by stipend desc)
from money m join
student s
on m.student_id = s.id_student join
faculty f
on f.id_faculty = s.faculty_id
group by f.id_faculty, f.name;

Answer 2

However when I add s.name to original query to also show the name of the student who received max_stipend, query is not working like it used to - it returns all of the students
When you add s.name you are looking for min value for each user.
If you want the name of user who has the MAX_STIPEND you should to move to window functions. For example Dense Rank in MS SQL Server.
with cte as
(select
f.name as "FACULTY_NAME",
s.name as "STUDENT_NAME",
stipend as "MAX_STIPEND",
DENSE_RANK() OVER
(PARTITION BY f.name, s.name ORDER BY i.stipend DESC) AS Rank
from
money m
inner join student s on m.student_id = s.id_student
inner join
faculty f on f.id_faculty = s.faculty_id
)
select "FACULTY_NAME", "STUDENT_NAME"
from cte
where rank = 1
Not all sql brands have windowed functions. Here the link for dense_rank on MySQL and also dense_Rank for Oracle