select * from korr k
,lateral (select * from total t where t.part=k.part and t.datum<k.datum order by t.datum desc FETCH NEXT 1 ROWS ONLY)(+)
,lateral (select * from total t where t.part=k.part and t.datum>k.datum order by t.datum asc  FETCH NEXT 1 ROWS ONLY)(+)

korr
as
(
select 2 as id, 'ABC' as part,to_date('02.01.2019','dd.mm.yyyy') as datum, 10 as wert_gesamt from dual
union all
select 3 as id,'ABC' as part,to_date('03.01.2019','dd.mm.yyyy') as datum, 15 as wert_gesamt from dual
)

 47  select *
 48  from (
 49      select
 50         v.*, count((select 1 from korr k where k.part=v.part and k.datum = v.datum))
 51              over(partition by v.part order by v.datum rows between 1 preceding and 1 following) f
 52      from (
 53          select t.*
 54          from total t
 55          where exists (select 1 from korr k where k.part=t.part)
 56      ) v
 57  ) w
 58* where w.f > 0
SQL> /

SQL> /

        ID PAR DATUM    WERT_GESAMT          F
---------- --- -------- ----------- ----------
         1 ABC 01.01.19           5          1
         2 ABC 02.01.19          10          2
         3 ABC 03.01.19          15          2
         4 ABC 04.01.19          26          1

with
total
as
(select 1 as id, 'ABC' as part, to_date('01.01.2019','dd.mm.yyyy') as datum, 5 as wert_gesamt from dual
union all
select 2 as id, 'ABC' as part,to_date('02.01.2019','dd.mm.yyyy') as datum, 10 as wert_gesamt from dual
union all
select 3 as id,'ABC' as part,to_date('03.01.2019','dd.mm.yyyy') as datum, 15 as wert_gesamt from dual
union all
select 4 as id,'ABC' as part,to_date('04.01.2019','dd.mm.yyyy') as datum, 26 as wert_gesamt from dual
union all
select 5 as id,'ABC' as part,to_date('05.01.2019','dd.mm.yyyy') as datum, 30 as wert_gesamt from dual
union all
select 6 as id,'AAA' as part,to_date('05.01.2019','dd.mm.yyyy') as datum, 8 as wert_gesamt from dual
union all
select 7 as id, 'AAA' as part,to_date('06.01.2019','dd.mm.yyyy') as datum, 12 as wert_gesamt from dual
),
korr
as
(
select 2 as id, 'ABC' as part,to_date('02.01.2019','dd.mm.yyyy') as datum, 10 as wert_gesamt from dual
union all
select 3 as id,'ABC' as part,to_date('03.01.2019','dd.mm.yyyy') as datum, 15 as wert_gesamt from dual
)
,
with_ranking
as
(
    select total.id,
           total.part,
           total.datum,
           total.wert_gesamt ,
           rank() over (partition by total.part order by total.datum desc)as ranking --down
    from korr, total
    where korr.part = total.part
    and korr.datum >= total.datum
    UNION
    select total.id,
           total.part,
           total.datum,
           total.wert_gesamt ,
           rank() over (partition by total.part order by total.datum)as ranking --forward
    from korr, total
    where korr.part = total.part
    and korr.datum <= total.datum
)
    select w.*
   from (
           select t.*,
           count(decode(t2.datum,t.datum,1))
               over(partition by t.part order by t.datum rows between 1 preceding and 1 following) f
           from korr t2, total t 
           where t2.part=t.part 
   ) w
  where w.f > 0
/
SQL> /

        ID PAR DATUM    WERT_GESAMT          F
---------- --- -------- ----------- ----------
         2 ABC 02.01.19          10          1
         2 ABC 02.01.19          10          2
         3 ABC 03.01.19          15          2
         3 ABC 03.01.19          15          1

        ID PAR DATUM    WERT_GESAMT        KID          F
---------- --- -------- ----------- ---------- ----------
         1 ABC 01.01.19           5                     1
         2 ABC 02.01.19          10          2          2
         3 ABC 03.01.19          15          3          2
         4 ABC 04.01.19          26                     1