SQL & PL/SQL

I am having trouble trying to re-write this query to avoid a full index scan:

select *
from (
      select asxsp.*
            ,row_number() over (partition by fk_asxsc_id order by last_price_date desc) rn
      from asx_share_price asxsp
      ) iq
where rn = 1
  and iq.fk_asxsc_id in (select max(id) from asx_share_company)

Explain plan:

Execution Plan
----------------------------------------------------------
0 SELECT STATEMENT Optimizer=ALL_ROWS (Cost=28487 Card=3027112
Bytes=227033400)

1 0 VIEW (Cost=28485 Card=3027112 Bytes=227033400)
2 1 WINDOW (SORT PUSHED RANK) (Cost=28485 Card=3027112 Bytes
=87786248)

3 2 INDEX (FAST FULL SCAN) OF 'ASXSP_PK' (INDEX (UNIQUE))
(Cost=4097 Card=3027112 Bytes=87786248)

4 1 SORT (AGGREGATE)
5 4 INDEX (FULL SCAN (MIN/MAX)) OF 'ASXSC_PK' (INDEX (UNIQ
UE)) (Cost=2 Card=3585 Bytes=14340)

Sorry if it is hard to read I dont know how to format it properly.

If the query is written as :

select *
from (
      select asxsp.*
            ,row_number() over (partition by fk_asxsc_id order by last_price_date desc) rn
      from asx_share_price asxsp
      ) iq
where rn = 1
  and iq.fk_asxsc_id in (23050)

I get the desired plan:

Execution Plan
----------------------------------------------------------
0 SELECT STATEMENT Optimizer=ALL_ROWS (Cost=5616 Card=396077 B
ytes=29705775)

1 0 VIEW (Cost=5616 Card=396077 Bytes=29705775)
2 1 WINDOW (SORT PUSHED RANK) (Cost=5616 Card=396077 Bytes=1
1486233)

3 2 INDEX (RANGE SCAN) OF 'ASXSP_PK' (INDEX (UNIQUE)) (Cos
t=2429 Card=396077 Bytes=11486233)

Can anyone shed some light on how I can avoid the full index scan?