SQL Server 2022

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

create table TabA (Id int, Currency varchar(5), Value int, institution varchar(5));
insert into tabA values 
  (1, 'CHF', 17, 'Fed1'), (1, 'EUR', 14, 'FedX'),
  (1, 'CHF', 14, 'Fed2'), (1, 'CHF', 15, 'FedQ'),
  (1, 'USD',  9, 'FedA'), (1, 'TRY', 19, 'FedX'),
  (1, 'CHF', 13, 'Fed3');
7 rows affected

with tmp as (
  select id, currency, institution,
         percentile_cont(0.5) within group (order by value) 
           over (partition by id, currency) AS mv
  from tabA WHERE Institution like  '%Fed%' )
SELECT Id, Currency, Count(*) AS Cnt, max(mv) as MedianValue,
       max(Institution) AS Institution 
FROM  tmp
GROUP BY Id, Currency
ORDER BY Institution


Id
Currency
Cnt
MedianValue
Institution


1
USD
1
9
FedA

1
CHF
4
14.5
FedQ

1
EUR
1
14
FedX

1
TRY
1
19
FedX


SELECT  Id, Currency, value, 
  PERCENTILE_CONT(0.5) within GROUP (order by Value) 
                       OVER (PARTITION BY id, Currency)  AS Median_Value,
  Institution 
FROM  TabA
WHERE Institution like  '%Fed%' 


Id
Currency
value
Median_Value
Institution


1
CHF
13
14.5
Fed3

1
CHF
14
14.5
Fed2

1
CHF
15
14.5
FedQ

1
CHF
17
14.5
Fed1

1
EUR
14
14
FedX

1
TRY
19
19
FedX

1
USD
9
9
FedA

Id	Currency	value	Median_Value	Institution
1	CHF	13	14.5	Fed3
1	CHF	14	14.5	Fed2
1	CHF	15	14.5	FedQ
1	CHF	17	14.5	Fed1
1	EUR	14	14	FedX
1	TRY	19	19	FedX
1	USD	9	9	FedA