How to return multiple fields values in a Oracle SQL Subquery

You probably need to use keep dense_rank aggregate function like below :

SELECT
    a.GESTOR,
    a.TIPO,
    a.NUMERO,
    max(a.seq) keep (dense_rank first order by a.seq desc) as seq,
    max(a.data) keep (dense_rank first order by a.seq desc) as data,
    max(a.valor) keep (dense_rank first order by a.seq desc) as valor,
    max(a.cod_nloc) keep (dense_rank first order by a.seq desc) as cod_nloc,
    max(a.usuario) keep (dense_rank first order by a.seq desc) as usuario    
FROM
    SM_HISTPATR_SE90 a
WHERE
    a.DATA <= '01/01/2020' and
    a.MOVIMENTO = 'E' and
    numero = 66480
GROUP BY a.GESTOR, a.TIPO, a.NUMERO

You cannot return multiple columns from 1 subquery. So you syntax is wrong. Alternatively you can write your query like -

SELECT
    a.GESTOR,
    a.TIPO,
    a.NUMERO,
    b.seq,
    b.data,
    b.valor,
    b.cod_nloc,
    b.usuario
FROM
    SM_HISTPATR_SE90 a
JOIN (SELECT seq, data, valor, cod_nloc, usuario
        FROM SM_HISTPATR_SE90
         AND seq = (SELECT MAX(seq)
                      FROM SM_HISTPATR_SE90
                     WHERE b.movimento = 'E')) b ON b.numero = a.numero
WHERE
    a.DATA <= '01/01/2020' and
    a.MOVIMENTO = 'E' and
    numero = 66480
GROUP BY a.GESTOR, a.TIPO, a.NUMERO

If you use that repeating code as a CTE (Common Table Expression) a.k.a. the WITH factoring clause, query would look like this:

with temp as
  (select b.numero, 
          max(b.seq) max_seq
   from sm_histpatr_se90 b
   where b.movimento = 'E'
   group by b.numero
  )
select a.gestor, 
       a.tipo, 
       a.numero, 
       a.data, 
       a.valor, 
       a.cod_nloc, 
       a.usuario
from sm_histpatr_se90 a join temp t on t.numero = a.numero and a.seq = t.max_seq
where a.data <= date '2020-01-01'
  and a.movimento = 'E'
  and a.numero = 66480;

Also, note that you're most probably comparing date to a string, here: a.DATA <= '01/01/2020'; you should use either a date literal (as I did) or TO_DATE function with appropriate format mask, such as a.data <= to_date('01.01.2020', 'dd.mm.yyyy').

Yet another thing: you've used table alias b twice, in different context. Try to avoid that because it makes confusion.