some_id    type    amount
1          costs      500
1          revenue    600
2          costs      100  <-- note: no revenue for id 2
3          costs      200
3          revenue    250

some_id    type    amount
1          costs      500
1          revenue    600
1          result     100  <-- generated result row
2          costs      100
2          result    -100  <-- generated result row
3          costs      200
3          revenue    250
3          result      50  <-- generated result row

SQL> select * from t order by 1,2;

        ID TYP            AMOUNT
---------- ---------- ----------
         1 costs             500
         1 revenue           600
         2 costs             100
         3 costs             200
         3 revenue           250

5 rows selected.

SQL> select id, 
  2         decode(grouping(typ),1,'result',typ) typ,
  3         sum(decode(typ,'revenue',1,-1)*amount) amount
  4  from t
  5  group by rollup(id,typ)
  6  having grouping(id) = 0
  7  order by id, decode(typ,'result',2,1)
  8  /

        ID TYP            AMOUNT
---------- ---------- ----------
         1 costs            -500
         1 revenue           600
         1 result            100
         2 costs            -100
         2 result           -100
         3 costs            -200
         3 revenue           250
         3 result             50

8 rows selected.

SQL> with t as ( select 1 some_id, 'costs'     type,  500  amount from dual union all
  2              select 1 some_id, 'revenue'   type,  600  amount from dual union all
  3              select 2 some_id, 'costs'     type,  100  amount from dual union all
  4              select 3 some_id, 'costs'     type,  200  amount from dual union all
  5              select 3 some_id, 'revenue'   type,  250  amount from dual)
  6  --
  7  select * from t
  8   model
  9    partition by (some_id)
 10    dimension by (type)
 11    measures(amount)
 12    (amount['result']=nvl(amount['revenue'],0)-nvl(amount['costs'],0))
 13  /

   SOME_ID TYPE        AMOUNT
---------- ------- ----------
         1 costs          500
         1 revenue        600
         1 result         100
         2 costs          100
         2 result        -100
         3 costs          200
         3 revenue        250
         3 result          50

8 rows selected

SQL>