SQL Server 2017

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

CREATE TABLE t1 ( ID varchar(2), theDate DATE, AMT int ) ;
INSERT INTO t1 (ID, theDate, Amt)
VALUES
 ('A','2018-04-04',110)
,('A','2018-03-03',110)
,('A','2018-02-02',100)
,('A','2017-10-10',90)
,('B','2018-10-10',90)
,('B','2018-09-09',90)
,('B','2018-07-07',80)
;
7 rows affected

SELECT s1.ID
    , FORMAT(s1.theDate,'MM-dd-yyyy') AS theDate
    , s1.Amt
    --, s1.PrevAmt
    , CASE WHEN Amt <> prevAmt THEN FORMAT(LAG(theDate) OVER ( PARTITION BY ID ORDER BY theDate),'MM-dd-yyyy') END AS prevDate
FROM (
  SELECT ID, theDate, Amt
    , LAG(AMT) OVER ( PARTITION BY ID ORDER BY theDate) AS prevAmt 
  FROM t1 
) s1
ORDER BY ID, theDate DESC


ID
theDate
Amt
prevDate


A
10-10-2017
90
null

A
04-04-2018
110
null

A
03-03-2018
110
02-02-2018

A
02-02-2018
100
10-10-2017

B
10-10-2018
90
null

B
09-09-2018
90
07-07-2018

B
07-07-2018
80
null

ID	theDate	Amt	prevDate
A	10-10-2017	90	null
A	04-04-2018	110	null
A	03-03-2018	110	02-02-2018
A	02-02-2018	100	10-10-2017
B	10-10-2018	90	null
B	09-09-2018	90	07-07-2018
B	07-07-2018	80	null